SAN FRANCISCO

TALL BUILDINGS
STUDY
Disclaimer: While the information presented in this report is believed to be correct, the Applied
Technology Council assumes no responsibility for its accuracy or for the opinions expressed herein. The
material presented in this publication should not be used or relied upon for any specific application
without competent examination and verification of its accuracy, suitability, and applicability by qualified
professionals. Users of information from this publication assume all liability arising from such use.
 San Francisco Tall Buildings Study

Prepared by
APPLIED TECHNOLOGY COUNCIL
201 Redwood Shores Parkway, Suite 240
Redwood City, California 94065
www.ATCouncil.org

Prepared for
City and County of San Francisco Office of Resilience and Capital Planning
Brian Strong, Chief Resilience Officer and Director
Danielle Mieler, Principal Resilience Analyst
San Francisco, California

ATC MANAGEMENT AND OVERSIGHT PROJECT TECHNICAL COMMITTEE

Ayse Hortacsu John D. Hooper (Task Leader)
Justin Moresco David Bonowitz
Gregory Deierlein
PROJECT WORKING GROUPS Shahriar Vahdani
Carlos Molina Hutt
Preetish Kakoty CITY AND COUNTY OF SAN FRANCISCO
Anne McLeod Hulsey TALL BUILDINGS EXECUTIVE PANEL
Alireza Eksir Monfared Naomi Kelly (Chair)
Max Rattie Mary Ellen Carroll
Wen-Yi Yen Kathryn How
Tom Hui
TASK REVIEW PANEL Brian Strong
Mark X. Haley
William Walton

December 2018
CITY AND COUNTY OF SAN FRANCISCO TALL BUILDINGS STAKEHOLDERS
Board of Supervisors Business
District 3 San Francisco Chamber of Commerce
District 6
Real Estate
Private Engineers and Architects Real Estate Advisory Services
DBI Code Advisory Committee
Heller Manus Architects Community and Non-profit Organizations
Building Inspection Commission SPUR
HOK BOMA
Maffei Structural Engineering The East Cut Community Benefit District
BXP
SEAONC AB 82/83 Code Advisory City and County of San Francisco Staff
Committee Department of Building Inspection
Fennie+Mehl Architects Real Estate Division
Department of Emergency Management
Developers Planning Department
San Francisco Apartment Owners Fire Department
Association SFPUC
Tishman Speyer Public Works
Boston Properties Office of Resilience and Capital Planning
 Preface

In 2010, the Applied Technology Council (ATC) concluded the Community

Action Plan for Seismic Safety (CAPSS) for the City and County of San
Francisco, identifying risks from future earthquakes and developing
recommendations to reduce the most significant earthquake impacts. As a
result, in 2011, the City published the CAPSS Earthquake Safety
Implementation Program Workplan 2012-2042 that anticipates actions for
broad groups of new and existing buildings throughout the city. In
considering this broad workplan, it became clear that special consideration
would be needed to address the unique characteristics of San Francisco’s tall
buildings. In 2017, ATC was awarded a contract to develop an inventory of
tall buildings in San Francisco and to review the impact of earthquakes on
such buildings.

ATC is indebted to the members of the Project Technical Committee who

served as the principal authors of this report, including David Bonowitz, Greg
Deierlein, John Hooper (chair), and Shariar Vahdani, and the members of the
Working Groups consisting of Carlos Molina Hutt, Anne McLeod Hulsey, and
Wen-Yi Yen. Preetish Kakoty, Alireza Eksir Monfared, and Max Rattie also
contributed to the Working Groups. ATC gratefully acknowledges Mark Haley
and Bill Walton for their review of Part 6 of this report. The names and
affiliations of those who contributed to this report are provided in the list of
Project Participants at the end of this report.

ATC also gratefully acknowledges guidance and support provided by Brian

Strong and Danielle Mieler of the City’s Office of Resilience and Capital
Planning. The Tall Buildings Executive Panel, consisting of Mary Ellen
Carroll, Kathryn How, Tom Hui, Brian Strong, and Naomi Kelly (chair)
provided advice at key stages of the work. ATC staff members Justin
Moresco and Carrie Perna provided project management support and report
production services, respectively.

Ayse Hortacsu Jon Heintz

ATC Director of Projects ATC Executive Director

ATC-119-1 Preface iii

 Summary Recommendations

In 2011, the City and County of San Francisco published the CAPSS
Earthquake Safety Implementation Program Workplan 2012-2042 as a result
of the Community Action Plan for Seismic Safety (CAPSS) in response to
Mayor Newsom’s Executive Directive 10-022. Workplan 2012-2042
anticipates programs for broad groups of new and existing buildings
throughout the City. Looking ahead, the City recognized that within each
broad group, some buildings would need special attention through
exemptions, programmatic solutions, or specific technical criteria, to make the
work feasible.

One such subgroup comprises the City’s “tall buildings.” In tall buildings,
difficult exterior access, multiple tenants and uses within a building, and their
sheer size complicate evaluation and retrofit. Their structural systems
preclude generic performance assumptions and prescriptive engineering
solutions. Damage to a tall building can pose risks well beyond its own
footprint. Their high concentration downtown poses an aggregate risk to
neighborhood and citywide recovery not presented by other building groups.
Perhaps most important, San Francisco’s new and existing tall buildings
represent a dominant portion of the City’s business sector, and increasingly
house residents as well.

For these reasons, the City initiated the first project in the nation to consider
the impact of earthquakes on tall buildings. The project conducted
investigations in seven focus areas under separate tasks. This report
documents the complete findings and recommendations of each task, and the
recommendations are summarized below.

An inventory developed for the project identified 156 buildings that are 240
feet or taller, either constructed or permitted for construction, primarily located
in San Francisco’s northeast neighborhoods (Supervisorial Districts 3 and 6).
Approximately 60% of these buildings contain primarily business uses, while
the others are predominantly residential.

The recommendations presented here flow from a study of these tall

buildings, but most are also applicable to a wider set of buildings supporting
similar functions or posing similar risks. Tall buildings, even in downtown San

ATC-119-1 Summary Recommendations v

 Francisco, are only part of a neighborhood’s building stock, and from a public
policy perspective, their earthquake performance is bound up with that of the
shorter buildings around them.

Each recommended action identifies one or more City departments to lead its
implementation. However, implementation of any new policy is assumed to
involve appropriate coordination with other City departments, outside experts
(as needed), and other stakeholders. Some recommended actions require
enactment of legislation by the Mayor and Board of Supervisors or action by
the Building Inspection Commission and can only commence after these
approvals.

Sections 1 through 4 below summarize the findings from the study in issue
statements and describe the associated recommendations to address the
issue. Section 5 presents the recommendations in a table format showing
different aspects of each recommendation, including potential implementation
timeframe, relationship to Workplan 2012-2042, City department responsible
for implementation, and the relationship of recommendations to Parts of the
project report for reference.

The Summary Recommendations were originally published by the City as

part of the Tall Buildings Safety Strategy on October 4, 2018. They have
been updated in response to feedback from the structural engineering
community. The specific changes are as follows:
 Recommendation 2A was modified to include a clarification that the repair
provisions of the San Francisco Existing Building Code should be applied
in accordance with a future Administrative Bulletin that defines
appropriate triggers and other indicators of potential damage.
 Recommendation 2B was expanded to include the addition of triggers
that apply when buildings are purchased or leases are renewed. In
addition, the issue statement was expanded to include change of
occupancy triggers.

1. Actions for Reducing Seismic Risk Prior to Earthquakes –

New Buildings
1A. Develop Regulations to Address Foundation and Geotechnical
Issues

Issue: The San Francisco Building Code sets minimum requirements for
geotechnical site investigations and foundation design. Because they are
minimum requirements, they do not fully address all of the geotechnical
conditions found in San Francisco. Over the past several decades, the San
Francisco geotechnical community has developed best practices for

vi Summary Recommendations ATC-119-1

geotechnical evaluation and foundation design, but these are not yet codified.
Many of the new tall building developments are challenging even these best
practices due to unique soil conditions, the size and weight of the new
buildings, and the sophisticated site investigation and the analysis
approaches being used to assess overall building behavior, including building
response to extreme earthquake ground motions.

Recommendation: To help reduce the risk associated with these

geotechnical challenges, the Department of Building Inspection (DBI) should
develop an Administrative Bulletin or Information Sheet (with building code
amendments as needed) with acceptable practices on topics including the
following:
 Settlement design and analysis criteria,
 Quality Assurance/Quality Control for foundation systems,
 Foundation design and other countermeasures for soil liquefaction and
lateral spreading,
 Shoring and dewatering design and analysis criteria,
 Lateral earthquake resistance of deep and shallow foundations, and
 Site characterization and exploration.

In addition, to strengthen DBI procedures for assessing the completeness of

the foundation and excavation design for tall buildings, two actions are
recommended:
 Increase DBI’s expertise on geotechnical issues related to tall buildings
through enhanced training and staffing.
 Develop a geotechnical report checklist to help ensure the completeness
of the submitted geotechnical investigation, design, and field monitoring
reports.

1B. Establish Recovery-Based Seismic Design Standards

Issue: San Francisco Building Code requirements for earthquake design,

including the performance-based requirements of Administrative Bulletin 083,
Requirements and Guidelines for the Seismic Design of New Tall Buildings
using Non-Prescriptive Seismic-Design Procedures, are primarily intended to
provide acceptable safety in extreme earthquakes. Studies conducted in this
project estimate that for a tall building designed to current standards, it might
take 2 to 6 months to mobilize for and repair damage from a major
earthquake, depending on the building location, geologic conditions, and the
structural and foundation systems. Long downtimes in tall buildings can have

ATC-119-1 Summary Recommendations vii

 disproportionate harmful effects on residents and businesses in San
Francisco. By the City’s tentative recovery goals, even three months of
downtime is unacceptably long for major employers and other recovery-
critical uses.

Recommendation: To shorten downtime in new tall buildings, DBI should

develop an Administrative Bulletin (with building code amendments as
needed) that supports the implementation of the City’s tentative recovery
goals and specifies recovery-based seismic design and construction
requirements, including tighter drift limits under expected ground motions,
enhanced design criteria for critical mechanical, electrical, plumbing, and
elevator systems, enhanced detailing requirements for exterior cladding and
interior partition walls, and measures to mitigate externalities that impede
recovery. San Francisco’s Building Occupancy Resumption Program
(BORP) is designed to address some of these externalities (see
recommendation 3B). BORP, or a program like it, should be required for all
new tall buildings.

2. Actions for Reducing Seismic Risk Prior to Earthquakes –

Existing Buildings
2A. Apply the Repair Provisions of the San Francisco Existing
Building Code with Respect to Possible Loma Prieta Damage

Issue: The 1994 Northridge earthquake revealed unexpected damage to

dozens of welded steel moment frame structures throughout the greater Los
Angeles area (some of which were tall; most of which were not). In most
cases, the structural damage did not reveal itself through evident architectural
damage or noticeable changes under everyday use. Five years earlier,
without the benefit of lessons later learned in Northridge, San Francisco’s
steel buildings were not consistently inspected after the Loma Prieta
earthquake. Ground motion data in downtown San Francisco during the
Loma Prieta earthquake are limited, and recorded data are lower than ground
motions associated with connection damage in the Northridge earthquake.
However, observational data suggest that some areas of San Francisco
experienced higher levels of shaking, potentially causing some buildings to
sustain connection damage that went undetected.

Recommendation: DBI should develop a new Administrative Bulletin to

interpret provisions of the San Francisco Existing Building Code as they apply
to post-earthquake inspection and evaluation of welded steel moment frames.
The Administrative Bulletin should be developed in a community consensus
process used to define appropriate ground motion triggers and other
indicators of potential damage. This process should consider all currently

viii Summary Recommendations ATC-119-1

available information, including FEMA 352, Recommended Postearthquake
Evaluation and Repair Criteria for Welded Steel Moment-Frame Buildings,
available recorded and observational data on ground motions from the Loma
Prieta earthquake, and information from building inspections that have
occurred since the Loma Prieta earthquake. (Development of the
Administrative Bulletin should be coordinated with related work described in
Recommendation 3F.)

Using the resulting Administrative Bulletin, DBI should apply the repair
provisions of the San Francisco Existing Building Code with respect to
possible Loma Prieta damage. Because the enforcement would be taking
place so long after the damaging earthquake, it would likely benefit from a
special program including notification, guidance to owners and tenants,
appropriate deadlines, and consideration of voluntary inspection and repair
work already performed. This Administrative Bulletin can also be used in the
implementation of Workplan 2012-2042 Task C.2.d, evaluation and retrofit of
all pre-1994 welded steel moment frame buildings.

2B. Amend the San Francisco Existing Building Code Triggers

Issue: San Francisco has its own code provisions that trigger seismic
upgrade when building alterations or changes of occupancy exceed a certain
scope. Because they apply only when two-thirds of a building’s floors are
altered or the occupant load is significantly increased, they rarely apply to tall
buildings, and are easily avoided. Therefore, even the most collapse-prone
tall buildings almost never receive the scrutiny intended by the code. That
said, since the retrofit of an occupied tall building is especially expensive and
disruptive, a more aggressive trigger provision could discourage
modernization, tenant improvement, or adaptive reuse. Thus, the San
Francisco Existing Building Code’s generic provisions are problematic for tall
and similarly large or complex facilities.
Federal and State government agencies, as well as some private institutions,
also trigger seismic evaluation, and possibly retrofit, when a facility is
purchased or leased. As contemplated in Workplan 2012-2042, San
Francisco could supplement its rarely used alteration and change of
occupancy triggers with such an acquisition trigger. Understanding that
triggered retrofits of tall buildings can be unreasonably expensive and
disruptive, acquisition-based triggers for just evaluation would at least ensure
that buildings are properly valued with respect to the risks they pose to
owners and tenants.
Recommendation: As the San Francisco Existing Building Code is
amended to coordinate with the 2019 California codes, the Building

ATC-119-1 Summary Recommendations ix

 Inspection Commission (BIC) should consider revisions that would find the
right balance between triggers and triggered scope from an alteration. This
could include setting a lower trigger, such as evaluation and disclosure, for
most of tall buildings and reserving a higher trigger, such as mandatory
retrofits, for those most prone to collapse. BIC should also consider adding
triggers that apply when buildings are purchased or leases are renewed. For
purposes of resilience and recovery planning, triggered evaluations might be
required to include an estimate of recovery time as well as safety.

2C. Recommend Minimum Levels of Earthquake Insurance or Other

Collateral to Ensure Post-Earthquake Recovery

Issue: Available information suggests that earthquake insurance availability

and market penetration for commercial and residential buildings are low.
Furthermore, when available, the insurance coverage is often limited to a
small fraction of the building replacement cost, raising questions about the
ability to repair and recover after a damaging earthquake. Insurance or other
resources to cover losses suffered by the neighbors of a tall building or costs
to the City (for debris removal or emergency protective measures) are also
unclear.

Recommendation: The City should identify potential limitations on the

availability of financial capital after a damaging earthquake and recommend
minimum levels of insurance (or other collateral) for tall building owners to
ensure recoverability of their buildings and the neighboring community.

2D. Review Requirements for Post-Earthquake Fire Suppression and

Evacuation Systems

Issue: Tall buildings rely on automatic fire suppression systems (typically

sprinklers) to inhibit fire spread and allow time for evacuation. Automatic fire
suppression will be particularly important following a significant earthquake,
when risk of fire ignitions might be higher, and the response time of fire
departments might be longer than usual. Normally, sprinkler systems in high-
rise buildings use water from the City’s municipal system, pressurized with
pumps and emergency generators. Recognizing the risk that the City water
supply may be disrupted by an earthquake, the San Francisco Building Code
requires many buildings to have an in-building secondary water supply to
operate the sprinkler system for 30 minutes.

Recommendation: The San Francisco Fire Department (SFFD) and the

Department of Emergency Management (DEM) should coordinate a study to
evaluate the adequacy of automatic fire suppression and occupant
evacuation systems in tall buildings for conditions following a significant
earthquake. The study should be coordinated with other City departments

x Summary Recommendations ATC-119-1

and within the broader context of the San Francisco Emergency Response
Plan to evaluate whether: (1) the in-building secondary water supply for
automatic fire suppression in tall buildings is sufficient to inhibit fire spread
and allow safe evacuation; and (2) the building code provisions that rely on
elevators for evacuation during a fire emergency will be effective following an
earthquake. The study should develop requirements and recommendations
for the San Francisco Building Code and Emergency Response Plan to
address any significant limitations or risks that are identified.

3. Actions for Reducing Seismic Risk Following Earthquakes

3A. Develop New Policies and Procedures for Implementing the
State’s Safety Assessment Program

Issue: The Safety Assessment Program (SAP), through which volunteer

inspectors “post” buildings with red, yellow, or green placards, is run by the
California Office of Emergency Services (Cal OES). DBI is charged with
implementing San Francisco’s participation in the program. The SAP
procedures and criteria are based on ATC-20, Procedures for
Postearthquake Safety Evaluation of Buildings, and are generic and not well-
suited to complex or recovery-critical facilities, including most tall buildings.

Recommendation: DBI should develop its own procedures suited to San

Francisco’s tall buildings (and otherwise unique building stock) regarding
such topics as limits on exterior-only inspection, limits on rapid evaluation,
damage estimates, placard use, and placard text.

In coordination with its implementation of SAP, DBI should also develop a

plan to use specially-qualified SAP volunteers to inspect pre-selected groups
of buildings, one of which might be tall buildings, especially those not covered
by BORP. More generally, the building groups of interest should be related to
the City’s adopted recovery goals. These specially trained and assigned
inspection teams would facilitate recovery of building types with fast recovery
goals.

3B. Extend and Improve the Building Occupancy Resumption

Program
Issue: BORP, created by DBI, allows building owners to arrange in advance
for post-earthquake safety inspections using their own contracted inspectors.
Participation is voluntary. DBI approves each participating building’s
application and pre-certifies the owner’s inspection team. Most of the current
BORP participants are downtown office buildings. BORP solves many of the
problems associated with applying the general Safety Assessment Program
to tall or otherwise complex or recovery-critical buildings. To enhance

ATC-119-1 Summary Recommendations xi

 BORP’s effectiveness and to derive the most value from it, DBI should
maintain and update the program.
Recommendation: DBI should enhance the BORP program with the
following:
 Conduct simulation-based training to ensure readiness of building staff,
BORP-certified inspectors, and DBI staff.
 Update the BORP instructions and procedures to improve consistency
and practicality. In particular, material required for certification by DBI
should be separate from material to be used by the BORP inspection
team in the field. The field material should be organized to align with the
ATC-20 evaluation procedures already adopted by BORP as its standard.
 Add specific criteria and pre-earthquake procedures to facilitate
implementation of FEMA 352, Recommended Postearthquake Evaluation
and Repair Criteria for Welded Steel Moment-Frame Buildings, for welded
steel moment frame structures.
 Require BORP for all new tall buildings. The program should be
extended through incentive programs, triggers, and possibly even
mandates (see recommendations 1B and 3H) for certain tall or otherwise
recovery-critical buildings. The purpose would be to require at least a
basic recovery plan, if not full BORP documentation and implementation,
for a larger set of existing buildings.
 Together with the Office of Resilience and Capital Planning (ORCP),
extend the BORP scope to support functional recovery in addition to safe
re-occupancy. For tall buildings, this might include additional procedures
for individual tenant spaces.

3C. Clarify and Update Roles and Responsibilities Associated with

Post-earthquake Emergency Response and Safety Inspection
Issue: Current procedures and practices for post-earthquake emergency
response and safety inspection are inconsistent, and sometimes out of date,
regarding the roles of certain City departments and their interaction with
state-level programs and private sector plans (including BORP).
Recommendation: DEM, in coordination with DBI and the Department of
Public Works (DPW), should update the Earthquake Annex of the San
Francisco Emergency Response Plan regarding activation of the Cal OES
Safety Assessment Program.
DBI and DPW should also update their SAP and BORP procedures regarding
the division of responsibility and criteria for establishing cordons and

xii Summary Recommendations ATC-119-1

barricades, in coordination with Lifelines Council recommendations for priority
transit routes. (See also recommendation 3G.)

3D. Update and Amend the San Francisco Existing Building Code
Triggers for Repair Projects
Issue: The San Francisco Existing Building Code triggers for seismic
upgrade based on the extent of earthquake damage have fallen out
coordination with the latest state code. In addition, while the latest
requirements are rational for most buildings, for larger structures, they can
become disproportionately expensive and disruptive; if repairs to many large
buildings are triggered, the aggregate impact can affect the City’s overall
recovery. San Francisco’s code amendments might exacerbate the problem
by not allowing use of reduced loads typical for retrofits.
Recommendation: As the 2019 California code becomes effective, DBI
should take the opportunity to update its traditional amendments and
coordinate them with the state code.
Regarding repair-triggered retrofits, ORCP, together with appropriate BIC
committees, should investigate whether San Francisco should relax its code
provisions for certain buildings, especially regarding the “substantial structural
damage” trigger. The study should consider typical San Francisco buildings,
ideally with a scenario that considers the effects of multiple buildings on
downtown recovery.

3E. Update Administrative Bulletin 099 and Clarify its Application to

Tall Concrete Structural Systems
Issue: Administrative Bulletin 099, Post-earthquake Repair and Retrofit
Requirements for Concrete Buildings, supplements the San Francisco
Existing Building Code by implementing FEMA 306, Evaluation of Earthquake
Damaged Concrete and Masonry Wall Buildings, published in 1998. Many of
AB-099’s references are due for updates. Further, while AB-099 applies to all
concrete buildings, some of its provisions might prove difficult to apply to tall
buildings.
Recommendation: DBI should update AB-099 to coordinate with newer
building code provisions and reference standards. DBI should also develop a
commentary to the updated AB-099 to guide its application to tall and
otherwise complex concrete buildings.

3F. Develop a New Administrative Bulletin for Post-earthquake

Inspection and Evaluation of Welded Steel Moment Frames
Issue: Many of San Francisco’s tall buildings have welded steel moment
frames as their structural systems. FEMA 352, Recommended

ATC-119-1 Summary Recommendations xiii

 Postearthquake Evaluation and Repair Criteria for Welded Steel Moment-
Frame Buildings, was developed after the 1994 Northridge earthquake
specifically for this structure type, but it is not fully referenced by San
Francisco codes and procedures.
Recommendation: DBI should develop an Administrative Bulletin to
implement FEMA 352. If code amendments are necessary, DBI should work
with BIC to approve them.
For post-earthquake safety evaluation, the recommended Administrative
Bulletin should clarify the procedures and criteria that DBI will accept in
coordination with Rapid and Detailed Evaluations using ATC-20. (As noted in
recommendation 3B, FEMA 352 is already allowed by BORP, but the
program procedures are incomplete regarding its use, so the recommended
Administrative Bulletin could address that shortcoming as well.)
For post-earthquake application of upgrade triggers, just as AB-099
supplements the San Francisco Existing Building Code for concrete buildings,
a similar bulletin should be developed for welded steel moment frames, which
are even more common in San Francisco’s tall buildings.

3G. Create Protocols and Procedures for Establishing Cordons

around Damaged Buildings

Issue: Cordons or barricades are often needed to protect the areas around a
damaged building. The cordoned area is generally based on the perceived
level of damage and the risks posed by potential aftershocks, wind loading,
time-dependent creep effects, or other factors. While cordons may be
required around buildings of any height, the disruptive implications of current
generic guidance for cordon distance increase dramatically with building
height, potentially leading to unnecessary closure of neighboring buildings
and infrastructure.

Recommendation: DPW, in coordination with DBI (see recommendation

3C) should develop protocols and procedures for cordoning around damaged
buildings. The procedures should be based on current practices but should
also account for tall building effects on neighborhood safety and recovery, as
well as new information regarding aftershock risks and early warning.

3H. Require Existing Buildings to File Recovery Plans

Issue: Because aggressive retrofit triggers and mandates can be

unreasonably expensive and disruptive for occupied tall buildings, thorough
recovery planning is a more feasible alternative. BORP can facilitate
recovery in a lightly damaged building, but as a voluntary program, it is not
used widely enough to improve the recovery of a whole neighborhood like

xiv Summary Recommendations ATC-119-1

San Francisco’s downtown, where the City’s tall buildings are most densely
clustered.

Recommendation: ORCP, in coordination with DBI, should identify

buildings critical to the City’s recovery and mandate the filing of building
information, and possibly a basic recovery plan, with the City. The building
information would enhance existing databases and inform the City’s response
and recovery planning, and the building-specific recovery plan would be a
way of extending the benefits of BORP without requiring ongoing expense
from owners.

4. Actions to Improve the City’s Understanding of its Tall

Building Seismic Risk
4A. Maintain and Expand the Database of Tall Buildings

Issue: This project compiled a database with information about all buildings
240 feet or taller either constructed or currently permitted for construction in
San Francisco. The database includes information on building location,
height, occupancy, age, construction material, structural system, year of
retrofit, and foundation type. Prior to the creation of this database, the City
had no centralized, searchable repository with this information about all tall
buildings in its jurisdiction. Following the completion of this project, the City
will need to develop a mechanism for maintaining or expanding the database.

Recommendation: As the owner of the database, ORCP should work with

DataSF and DBI to maintain and expand the tall building database. Specific
recommended actions include:
 Develop mechanisms for harvesting building data from building permits or
inspections administered by DBI or other agencies. For example, San
Francisco Existing Building Code Chapter 4E will yield information on
building façades as the provisions are implemented over the coming
years.
 Make the database available for public review and use. This will
encourage updates by building owners and will facilitate use of the data in
studies to promote recovery and resilience planning.
 Expand the database to include the following:
o All buildings taller than 75 feet. As described in the project report, the
240-foot height criterion for the initial database was somewhat
arbitrary. To the extent that the San Francisco Building Code
imposes elevator, fire safety, and other requirements on high-rise

ATC-119-1 Summary Recommendations xv

 buildings defined as those taller than 75 feet, it would be useful to
expand the database to include at least all buildings above this height.
o Specific building uses by industry or employment sector. The data
should be coordinated with development of the City’s overall recovery
goals. Ideally, similar data for non-tall buildings serving those same
recovery functions would also be compiled, allowing a better
understanding of the tall building effects in terms of neighborhood and
citywide recovery.
o Foundation type of any building on a site mapped as susceptible to
liquefaction.

4B. Develop a Comprehensive Recovery Plan for the Financial

District and Adjacent Neighborhoods
Issue: The present study addresses the effect of tall building damage on the
tall buildings themselves and, to a lesser extent, on the downtown
neighborhoods where tall buildings are densely clustered and on the City
overall. But it does not explicitly address the likely interactions between the
tall buildings, the non-tall buildings that still comprise most of downtown, and
the critical infrastructure that serves the neighborhood. Nor does it explicitly
consider resource demands and capacities of the businesses, residents,
workers, and other stakeholders.
Recommendation: A separate recovery plan, drawing on the present
study’s findings, would bring these ideas together in a practical way to
support a neighborhood and its functions, as opposed to just individual
buildings with certain characteristics. An interim recovery plan, which should
be developed by the City Administrator, would need to make many
assumptions. This should be followed by a more thorough recovery study to
confirm or correct those assumptions and to fill in the most critical knowledge
gaps. Such a study would address the combined effects of tall buildings,
non-tall buildings, and infrastructure, including liquefaction effects. The study
would develop a recovery curve estimating the level of immediate functional
loss and the extent of re-occupancy and recovery over time following one or
more scenario earthquakes.

The recovery plan should consider developing alternative habitability

standards for tall buildings. In 2012, SPUR recommended relaxing normal
habitability standards during post-earthquake recovery, gradually returning to
a state of normalcy. But the SPUR recommendations were developed
primarily for houses and wood-frame apartment buildings. Tall buildings, by
contrast, rely on sophisticated systems for heating, ventilation,
air-conditioning, elevators, and fire suppression, so the SPUR

xvi Summary Recommendations ATC-119-1

recommendations will not apply. DEM, together with ORCP, should develop
alternative recovery-phase habitability standards for tall buildings, considering
minimum requirements for fire barriers, suppression and safety systems,
vertical transportation, water services, and electricity.

5. Summary Tables

Table S-1 presents the recommendations in terms of potential

implementation timeframes. Short-term actions can be started with
essentially no additional study and completed without a legislative process.
Typical examples include administrative updates, development of
Administrative Bulletins, and initiatives requiring only nominal
interdepartmental coordination. Mid-term actions can normally be completed
without substantial additional study and without a legislative process but are
expected to involve substantial input from multiple stakeholder groups. Long-
term actions are expected to require substantial additional technical study or
a legislative process.

Table S-2 relates the recommendations to relevant to tasks Workplan

2012-2042. In some cases, this tall building study identified additional
programs and initiatives not detailed in Workplan 2012-2042. In others, the
tall building recommendations identify specific technical criteria, exemptions,
or programmatic solutions suitable for tall and similarly complex buildings,
even relaxing some of the Workplan 2012-2042’s broad directives.

Table S-1 Implementation Timeframe

Short- Mid- Long-
Recommended Action Term Term Term
1. Actions for Reducing Seismic Risk Prior to Earthquakes – New Buildings
1A. Develop Regulations to Address Foundation and Geotechnical Issues
Training and checklist X
Develop geotechnical regulations X
1B. Establish Recovery-Based Seismic Design Standards X
2. Actions for Reducing Seismic Risk Prior to Earthquakes – Existing Buildings
2A. Apply the Repair Provisions of the San Francisco Existing Building Code with Respect
X
to Possible Loma Prieta damage
2B. Amend the San Francisco Existing Building Code Triggers
Alteration and change of occupancy triggers X
Acquisition triggers X
2C. Recommend Minimum Levels of Earthquake Insurance or Other Collateral to Ensure
X
Post-Earthquake Recovery
2D. Review Requirements for Post-Earthquake Fire Suppression and Evacuation Systems X

ATC-119-1 Summary Recommendations xvii

Table S-1 Implementation Timeframe (continued)
Short- Mid- Long-
Recommended Action Term Term Term
3. Actions for Reducing Seismic Risk Following Earthquakes
3A. Develop New Policies and Procedures for Implementing the State’s Safety
X
Assessment Program
3B. Extend and Improve the Building Occupancy Resumption Program
Conduct simulation-based training X
Update procedures X
Extend program X
3C. Clarify and Update Roles and Responsibilities Associated with Post-earthquake
X
Emergency Response and Safety Inspection
3D. Update and Amend the San Francisco Existing Building Code Triggers for Repair
X
Projects
3E. Update Administrative Bulletin 099 and Clarify its Application to Tall Concrete
X
Structural Systems
3F. Develop a New Administrative Bulletin for Post-Earthquake Inspection and Evaluation
X
of Welded Steel Moment Frames
3G. Create Protocols and Procedures for Establishing Cordons around Damaged Buildings X
3H. Require Existing Buildings to File Recovery Plans X
4. Actions to Improve the City’s Understanding of its Tall Building Seismic Risk
4A. Maintain and Expand the Database of Tall Buildings X
4B. Develop a Comprehensive Recovery Plan for the Financial District and Adjacent
X
Neighborhoods

Table S-2 Relationship of Recommended Actions to Workplan 2012-2042 Tasks

Recommended Action Workplan 2012-2042 Task
1. Actions for Reducing Seismic Risk Prior to Earthquakes – New Buildings
1A. Develop Regulations to Address Foundation and New action
Geotechnical Issues
1B. Establish Recovery-Based Seismic Design Standards B.6.a Update code for new buildings to reflect desired
performance goals
2. Actions for Reducing Seismic Risk Prior to Earthquakes – Existing Buildings
2A. Apply the Repair Provisions of the San Francisco Existing B.4.b Develop post-earthquake repair and retrofit standards
Building Code with Respect to Possible Loma Prieta
damage

xviii Summary Recommendations ATC-119-1

Table S-2 Relationship of Recommended Actions to Workplan 2012-2042 Tasks (continued)
Recommended Action Workplan 2012-2042 Task
2. Actions for Reducing Seismic Risk Prior to Earthquakes – Existing Buildings (continued)
2B. Amend the San Francisco Existing Building Code Triggers C.1.a Mandatory evaluation on sale or by deadline
C.1.b Evaluation of buildings retrofitted prior to 1994 or building
to non-conforming performance standards
C.2.a Mandatory retrofit of older non-ductile concrete residential
buildings
C.2.d Mandatory evaluation and retrofit of pre-1994 welded
steel moment frame buildings
C.2.e Mandatory evaluation and retrofit of other low-
performance buildings
2C. Recommend Minimum Levels of Earthquake Insurance or A.1.b Provide information and assistance about insurance
Other Collateral to Ensure Post-earthquake Recovery
2D. Review Requirements for Post-Earthquake Fire Suppression A.6.i Study fire-related earthquake resilience topics
and Evacuation Systems
3. Actions for Reducing Seismic Risk Following Earthquakes
3A. Develop New Policies and Procedures for Implementing the A.4.f Update post-earthquake inspection (ATC-20) policies and
State’s Safety Assessment Program procedures
3B. Extend and Improve the Building Occupancy Resumption B.1.b Develop non-structural upgrade program for businesses
Program
3C. Clarify and Update Roles and Responsibilities Associated Procedural update
with Post-Earthquake Emergency Response and Safety
Inspection
3D. Update and Amend the San Francisco Existing Building B.4.b Develop post-earthquake repair and retrofit standards
Code Triggers for Repair Projects
3E. Update Administrative Bulletin 099 and Clarify its Application A.4.d Adopt disproportionate damage trigger
to Tall Concrete Structural Systems B.4.b Develop post-earthquake repair and retrofit standards
3F. Develop a New Administrative Bulletin for Post-Earthquake A.4.d Adopt disproportionate damage trigger
Inspection and Evaluation of Welded Steel Moment Frames B.4.b Develop post-earthquake repair and retrofit standards
3G. Create Protocols and Procedures for Establishing Cordons Program update, new action
around Damaged Buildings
3H. Require Existing Buildings to File Recovery Plans B.1.b Develop non-structural upgrade program for businesses
4. Actions to Improve the City’s Understanding of its Tall Building Seismic Risk
4A. Maintain and Expand the Database of Tall Buildings A.2.b Adopt façade maintenance regulations
4B. Develop a Comprehensive Recovery Plan for the Financial A.4.a Develop and adopt Shelter-in-Place policies and
District and Adjacent Neighborhoods procedures
B.2.b Mandatory evaluation of 5+ dwelling unit residential
buildings and hotels/motels
C.2.b Mandatory evaluation and retrofit of critical stores,
suppliers, and service providers
C.2.c Mandatory evaluation and retrofit of larger (over 300
occupants) assembly buildings

ATC-119-1 Summary Recommendations xix

 Table S-3 presents the recommendations in terms of City departments
responsible for implementation. The departments are identified as follows:
 BIC Building Inspection Commission
 DataSF
 DBI Department of Building Inspection
 DEM Department of Emergency Management
 DPW Department of Public Works
 ORCP Office of Resilience and Capital Planning
 SFFD San Francisco Fire Department
Table S-4 presents the relationship of Summary Recommendations to Parts
of this report for reference.

xx Summary Recommendations ATC-119-1

Table S-3 Responsible Department
Data
Recommended Action BIC SF DBI DEM DPW ORCP SFFD

1A. Develop Regulations to Address Foundation and

X
Geotechnical Issues
1B. Establish Recovery-Based Seismic Design
X
Standards
2A. Apply the Repair Provisions of the San Francisco
Existing Building Code with Respect to Possible X
Loma Prieta damage
2B. Amend the San Francisco Existing Building Code
X
Triggers
2C. Recommend Minimum Levels of Earthquake
Insurance or Other Collateral to Ensure Post- X
Earthquake Recovery
2D. Review Requirements for Post-earthquake Fire
X X
Suppression and Evacuation Systems
3A. Develop New Policies and Procedures for
Implementing the State’s Safety Assessment X
Program
3B. Extend and Improve the Building Occupancy
X X
Resumption Program
3C. Clarify and Update Roles and Responsibilities
Associated with Post-Earthquake Emergency X X X
Response and Safety Inspection
3D. Update and Amend the San Francisco Existing
X X X
Building Code Triggers for Repair Projects
3E. Update Administrative Bulletin 099 and Clarify its
X
Application to Tall Concrete Structural Systems
3F. Develop a New Administrative Bulletin for Post-
Earthquake Inspection and Evaluation of Welded X X
Steel Moment Frames
3G. Create Protocols and Procedures for Establishing
X X X
Cordons around Damaged Buildings
3H. Require Existing Buildings to File Recovery Plans X X

4A. Maintain and Expand the Database of Tall

X X X
Buildings
4B. Develop a Comprehensive Recovery Plan for the
X X
Financial District and Adjacent Neighborhoods

ATC-119-1 Summary Recommendations xxi

Table S-4 Recommended Actions by Part
Recommended Action Part 1 Part 2 Part 3 Part 4 Part 5 Part 6 Part 7
Actions for Reducing Seismic Risk Prior to Earthquakes – New Buildings
1A. Develop an AB and Code Amendment that
X
Addresses Foundation and Geotechnical Issues
1B. Establish Performance-Based Seismic Design
X
Standards
Actions for Reducing Seismic Risk Prior to Earthquakes – Existing Buildings
2A. Apply the Repair Provisions of the SFEBC with
X
Respect to Possible Loma Prieta Damage
2B. Amend the SFEBC Triggers X
2C. Require Minimum Levels of Earthquake
X
Insurance to Ensure Recovery
2D. Increase Local Water Supply for automatic Fire
X
Suppression Systems in Tall Buildings
Actions for Reducing Seismic Risk Following Earthquakes
3A. Develop New Policies and Procedures for
X X
Implementing the State’s SAP
3B. Extend and improve BORP X
3C. Clarify and Update Roles and Responsibilities
Associated with Post-earthquake Emergency X X
Response and Safety Inspection
3D. Update and Amend the SFEBC X
3E. Update AB-099 and Clarify its Application to Tall
X
Concrete Structural Systems
3F. Develop a New AB to Implement FEMA 352 for
Post-earthquake Inspection and Evaluation of X X
Welded Steel Moment Frames
3G. Create Protocols and Procedures for Establishing
X
Cordons Around Damaged Buildings
3H. Require Existing Buildings to File Recovery Plans X
Actions to Improve the City’s Understanding of its Tall Building Seismic Risk
4A. Maintain and Expand the Database of Tall
X
Buildings
4B. Develop a Comprehensive Recovery Plan for the
X
Financial District and Adjacent Neighborhoods

xxii Summary Recommendations ATC-119-1

 Table of Contents

Preface ......................................................................................................... iii

Summary Recommendations ...................................................................... v

List of Figures ......................................................................................... xxix

List of Tables ........................................................................................... xxxi

List of Abbreviations ............................................................................ xxxiii

Introduction to Tall Buildings Study .........................................................I-1

I.1 Background...........................................................................I-1
I.2 Project Objectives and Scope ...............................................I-1
I.3 Intended Audience and Next Steps .......................................I-2
I.4 Report Organization ..............................................................I-3

Part 1: Inventory of Tall Buildings

1. Introduction .....................................................................................1-1
1.1 Background..........................................................................1-1
1.2 Inventory Overview ..............................................................1-2
1.3 Data Fields and Sources ......................................................1-4
1.4 Maintaining an Updated Inventory........................................ 1-6
1.5 Organization ........................................................................1-6

2. Characterization of Building Inventory Features .............................. 1-9

2.1 Inventory of Buildings Taller than 240 Feet .......................... 1-9
2.2 A Closer Look: Financial District Building Inventory ........... 1-13
2.3 Evolution of Seismic Building Codes .................................. 1-16
2.4 Other Considerations ......................................................... 1-19

3. Tall Building Cohorts ..................................................................... 1-23

3.1 Identification of Tall Building Cohorts ................................. 1-23
3.2 Older Steel Buildings ......................................................... 1-23
3.3 Older Concrete Buildings ................................................... 1-25
3.4 Modern Concrete Buildings ................................................ 1-26

4. Summary and Recommendations.................................................. 1-29

A. Tall Building Inventory Data Fields, Definitions, and Notes ............ 1-33

B. Building Permit Supplement for Inventory Maintenance ................. 1-35

ATC-119-1 Table of Contents xxiii

 C. Reference for Bedrock Depth Data ................................................1-39

D. References ....................................................................................1-41

Part 2: Geotechnical Engineering for Tall Buildings

1. Introduction......................................................................................2-1
1.1 Background ..........................................................................2-1
1.2 Intended Audience and Use of this Report ...........................2-1
1.3 San Francisco Conditions ....................................................2-2
1.4 Organization.........................................................................2-4

2. Total and Differential Settlement ......................................................2-5

3. New Tall Buildings on Deep Foundations ........................................2-7

3.1 Deep Foundation Types .......................................................2-7
3.1.1 Drilled Shafts .........................................................2-7
3.1.2 Driven Concrete or Steel Piles ...............................2-9
3.1.3 Auger Cast Piles ..................................................2-10
3.1.4 Other Deep Foundation Types .............................2-11
3.2 Piles Going Through Soft or Liquefiable Soils and Firm
Soil Interface ......................................................................2-12
3.3 Downdrag Loads on Piles ..................................................2-12
3.4 Integrity Testing of Deep Foundations ................................2-13
3.5 Allowable Tolerance for Installation of Deep
Foundations .......................................................................2-13
3.6 Pile Load Test Procedure and Interpretation of Results .....2-14
3.7 Soil Liquefaction-Related Design Issues ............................2-14
3.7.1 Temporary Loss of Axial Support .........................2-15
3.7.2 Reduction of Lateral Support ...............................2-15
3.7.3 Downdrag Loads .................................................2-16
3.7.4 Effect on Ground Motion ......................................2-16
3.7.5 Liquefaction Related Ground Settlement .............2-17
3.8 Floating Deep Foundations ................................................2-17
3.9 Pile Group Effects ..............................................................2-18
3.10 Modeling of Pile-Supported Structures and Soil-Structure
Interaction ..........................................................................2-19
3.11 Expected Performance of Piles During the Design
Seismic Event ....................................................................2-22
3.12 Use of Battered Piles .........................................................2-22

4. New Tall Buildings on Shallow Foundations ..................................2-23

4.1 Bearing Capacity and Short or Long Term Settlement .......2-23
4.2 Foundation-Soil Interaction and Foundation Design
Against Lateral Loads ........................................................2-25
4.3 Effects of Foundation Flexibility on Dynamic Response
of a Tall Building ................................................................2-25

5. Evaluation and Retrofit of Foundations of Existing Tall

Buildings ........................................................................................2-27

xxiv Table of Contents ATC-119-1

 5.1 Geotechnical Investigation Methods Operating from
Basement of an Existing Tall Building ................................ 2-28
5.2 Use of Micropiles for Strengthening the Foundation of an
Existing Tall Building .......................................................... 2-28

6. Foundation Design of Tall Buildings Near Shoreline ...................... 2-31

7. Shoring and Dewatering ................................................................ 2-33

8. Sea Level Rise ..............................................................................2-37

9. Recommendations......................................................................... 2-39

A. Case Histories: Settlement of Tall Buildings .................................. 2-45

Part 3: Performance Expectation for New Buildings

1. Introduction .....................................................................................3-1
1.1 Background..........................................................................3-1
1.2 Seismic Performance Objectives ......................................... 3-1
1.3 Organization ........................................................................3-3

2. Archetype Tall Buildings ..................................................................3-5

2.1 Design Basis and Objectives ...............................................3-5
2.2 Reinforced Concrete Shear Wall Residential Building .......... 3-7
2.3 Steel Buckling Restrained Braced Frame Office Building ..... 3-9
2.4 Assessment Methodology .................................................. 3-10
2.5 Loss and Downtime Results ............................................... 3-21
2.6 Variations in Archetype Tall Building Design Criteria .......... 3-27

3. Other Tall Building Studies ............................................................ 3-31

3.1 Recent Tall Building Projects ............................................. 3-31
3.2 Research Publications ....................................................... 3-31
3.3 Mid-Rise Building Loss Studies .......................................... 3-33

4. Summary, Recommendations, and Cost Implications .................... 3-35

4.1 Summary ........................................................................... 3-35
4.2 Recommendations ............................................................. 3-37
4.3 Cost Implications................................................................ 3-37

A. References ....................................................................................3-41

Part 4: Post-earthquake Structural Evaluation

1. Introduction .....................................................................................4-1
1.1 Background..........................................................................4-1
1.2 Organization ........................................................................4-1

2. Current Policy ..................................................................................4-3

2.1 Evaluation, Repair, and Triggered Retrofit ........................... 4-3
2.1.1 2016 SFEBC Provisions ........................................ 4-3

ATC-119-1 Table of Contents xxv

 2.1.2 Administrative Bulletins ..........................................4-5
2.2 Relation to Post-Earthquake Safety Inspection ....................4-7
2.3 Relation to Other Code-Regulated Work ..............................4-8

3. Issues and Recommendations .......................................................4-11

3.1 Revisions and Amendments to the SFEBC ........................4-11
3.1.1 Policy on Repair-Triggered Retrofit ......................4-11
3.1.2 Routine Code Development .................................4-13
3.2 Administrative Bulletins and Reference Documents ...........4-14
3.2.1 AB-099 ................................................................4-14
3.2.2 New Administrative Bulletins................................4-15

A. Annotated Building Code Provisions ..............................................4-17

B. References ....................................................................................4-37

Part 5: Tall Building Effects on Post-earthquake Recovery

1. Introduction......................................................................................5-1
1.1 Background ..........................................................................5-1
1.2 Acknowledgements ..............................................................5-1
1.3 Organization.........................................................................5-2

2. Tall Building Effects .........................................................................5-3

2.1 Overview ..............................................................................5-3
2.2 Characterization of Building Tallness ...................................5-4
2.2.1 Building Period ......................................................5-5
2.2.2 High Mode Effects .................................................5-5
2.2.3 Aspect Ratio ..........................................................5-6
2.2.4 Story Drift Ratio .....................................................5-6
2.2.5 System Ductility Requirements ..............................5-6
2.2.6 Collapse-Prone Existing Buildings .........................5-7
2.2.7 Importance or Risk Category .................................5-7
2.2.8 Structural Integrity .................................................5-8
2.2.9 Falling Hazards......................................................5-9
2.2.10 Fire Safety .............................................................5-9
2.2.11 Planning and Zoning ............................................5-10
2.2.12 Summary Table ...................................................5-11

3. Safety and Emergency Response..................................................5-13

3.1 Cordoning ..........................................................................5-13
3.1.1 Current Guidance ................................................5-13
3.2 Evacuation and Fire Following Earthquake ........................5-17
3.3 Search and Rescue............................................................5-18

4. Consequences from Damage ........................................................5-21

4.1 Direct Costs to Owners and Tenants ..................................5-21
4.2 Indirect Costs to Owners and Tenants ...............................5-23
4.3 Consequences to City ........................................................5-24

5. Accelerating Recovery ...................................................................5-27

xxvi Table of Contents ATC-119-1

 5.1 Inspection and Tagging ...................................................... 5-27
5.2 Reoccupancy and Functional Recovery ............................. 5-29
5.3 Insurance ........................................................................... 5-30

6. Issues and Recommendations ...................................................... 5-35

A. References ....................................................................................5-39

Part 6: Post-earthquake Occupancy and Safety Inspection

1. Introduction .....................................................................................6-1
1.1 Background..........................................................................6-1
1.2 Organization ........................................................................6-1

2. Background .....................................................................................6-3
2.1 Current San Francisco Practice ........................................... 6-3
2.2 Safety, Reoccupancy, Functional Recovery, and City
Resilience ............................................................................6-3
2.3 Related Issues .....................................................................6-6

3. The Cal OES Safety Assessment Program ..................................... 6-9

3.1 SAP as Applied in San Francisco......................................... 6-9
3.2 Recommendations Regarding SAP .................................... 6-10
3.2.1 Activation ............................................................. 6-10
3.2.2 Early Recovery for Critical Building Groups ......... 6-11
3.2.3 Indicator Buildings for Aftershock Tracking .......... 6-11
3.2.4 Consistency ......................................................... 6-12

4. Building Occupancy Resumption Program .................................... 6-13

4.1 Background........................................................................ 6-13
4.2 Issues and Recommendations Regarding BORP ............... 6-14
4.2.1 Training Simulation .............................................. 6-14
4.2.2 Program Extension .............................................. 6-14
4.2.3 Personnel and Financing ..................................... 6-15
4.2.4 BORP Report Organization ................................. 6-16
4.2.5 BORP Report Content ......................................... 6-17
4.2.6 Use of Instrumentation ........................................ 6-18
4.2.7 Cordons and Barricades ...................................... 6-19
4.2.8 Use of FEMA 352 ................................................ 6-19
4.2.9 Concrete Structures............................................. 6-20
4.2.10 Functional Recovery ............................................ 6-20

5. ATC-20 ..........................................................................................6-23
5.1 ATC-20 as Criteria for SAP and BORP .............................. 6-23
5.2 Issues and Recommendations Regarding ATC-20............. 6-27
5.2.1 Limits on Exterior-Only Inspection ....................... 6-27
5.2.2 Limits on Rapid Evaluation .................................. 6-27
5.2.3 Damage Estimates .............................................. 6-28
5.2.4 Placard Use and Tracking ................................... 6-29
5.2.5 Placard Text ........................................................ 6-29

ATC-119-1 Table of Contents xxvii

 6. FEMA 352 .....................................................................................6-31
6.1 Background ........................................................................6-31
6.2 Issues and Recommendations Regarding FEMA 352 ........6-34
6.2.1 Implementation of FEMA 352 ..............................6-34
6.2.2 Application of FEMA 352 to Loma Prieta
Damage ...............................................................6-35

7. Inspection Technologies ................................................................6-39

A. References ....................................................................................6-41

Part 7: Pre-earthquake Evaluation

1. Introduction......................................................................................7-1
1.1 Organization.........................................................................7-1

2. Current Policy for Pre-earthquake Evaluation ..................................7-3

2.1 2016 SFEBC Alteration Triggers ..........................................7-3
2.2 2016 SFEBC Change of Occupancy Triggers ......................7-6

3. Alternative Approaches....................................................................7-9
3.1 Mandatory Retrofit Ordinances ............................................7-9
3.2 Aggressive Triggers with Focused Targets ........................7-10
3.3 Aggressive Triggers with Reduced Scope ..........................7-14
3.4 Acquisition Triggers............................................................7-17
3.4.1 The PML Market as an Alternative to Public
Policy ...................................................................7-20
3.5 Incentives for Voluntary Evaluation or Retrofit ....................7-21

4. Issues and Recommendations .......................................................7-23

4.1 Revisions and Amendments to the SFEBC ........................7-23
4.2 New Legislation ..................................................................7-24

A. Annotated Building Code Provisions ..............................................7-27

B. Seismic Retrofit as an Alteration ....................................................7-47

C. References ....................................................................................7-49

Project Participants ................................................................................. P-1

xxviii Table of Contents ATC-119-1

 List of Figures

Figure 1-1 Spatial distribution of San Francisco’s downtown building

inventory mapped as a function of building height ................ 1-3
Figure 1-2 Spatial distribution of San Francisco’s tall building
inventory ..............................................................................1-3
Figure 1-3 Figure showing 156 San Francisco buildings tall than
240 ft by percentage of buildings by SFRS type,
breakdown of SFRS by decade, and distribution of SFRS
by number of stories .......................................................... 1-11
Figure 1-4 Figure showing 156 San Francisco buildings tall than
240 ft by percentage of buildings by occupancy type,
distribution of occupancies by decade, and distribution of
occupancies by number of stories ...................................... 1-12
Figure 1-5 Figure showing distribution of the number of stories by
decade ...............................................................................1-13
Figure 1-6 Figure showing percentage of buildings by foundation
type, and breakdown of foundation types by depth to
bedrock ..............................................................................1-13
Figure 1-7 Map showing 243 buildings in a 31-block downtown San
Francisco study area.......................................................... 1-14
Figure 1-8 Joint distribution of pre-1994 steel moment-resisting
frame tall buildings ............................................................. 1-24
Figure 1-9 Proposed permit supplement form ..................................... 1-36
Figure 1-10 Proposed permit supplement form ..................................... 1-37
Figure 1-11 USGS map used for determining the depth to bedrock ...... 1-40
Figure 2-1 Map showing foundation types of tall buildings on a map
showing liquefaction potential of the soil .............................. 2-2
Figure 2-2 Typical soil profile for a downtown San Francisco site ......... 2-3
Figure 3-1 Reinforced concrete shear wall residential building .............. 3-8
Figure 3-2 Buckling-restrained braced frame office building .................. 3-9
Figure 3-3 Site locations of archetype buildings in downtown San
Francisco ........................................................................... 3-11

ATC-119-1 List of Figures xxix

 Figure 3-4 Maximum Considered Earthquake (MCE) target spectrum
superimposed with the average and individual ground
motion spectra ...................................................................3-13
Figure 3-5 Earthquake demands from non-linear response history
analysis of the reinforced concrete shear wall residential
building under the design earthquake.................................3-17
Figure 3-6 Earthquake demands from non-linear response history
analysis of the steel buckling-restrained braced frame
building under the design earthquake.................................3-19
Figure 3-7 Loss and downtime assessment methodology ...................3-20
Figure 3-8 Summary of median and 90th percentile loss results
under DE shaking ..............................................................3-22
Figure 3-9 Contribution of different building components to overall
loss at DE for Site Class D .................................................3-23
Figure 3-10 Summary of median and 90th percentile downtime to
functional recovery considering the baseline case and
recovery planning under DE shaking ..................................3-24
Figure 3-11 RCSW functional recovery time deaggregation at Site
Class D under DE ..............................................................3-26
Figure 3-12 Steel BRB functional recovery time deaggregation at Site
Class D under DE ..............................................................3-26
Figure 3-13 Mean story drift demand versus repair cost losses .............3-28
Figure 3-14 Mean story drift demand versus repair times ......................3-28
Figure 3-15 Mean story drift demand versus functional recovery times .3-29
Figure 3-16 Mean MCE-level story drift demands from recently
completed tall buildings designed using performance-
based methods ..................................................................3-31
Figure 3-17 Loss versus drift data for different structural systems .........3-32
Figure 3-18 FEMA P-58 PET drift versus loss sensitivity results ...........3-34
Figure 3-19 FEMA P-58 PET drift versus repair time sensitivity results .3-34
Figure 5-1 Initial cordoning safe distances ..........................................5-14
Figure 5-2 Illustration of a tall building cordon with a radius equal to
1.5 times the building height...............................................5-15
Figure 5-3 Damage types by Schweier and Markus (2004) .................5-19
Figure 5-4 Guidelines for barricading ..................................................5-28

xxx List of Figures ATC-119-1

 List of Tables

Table S-1 Implementation Timeframe ..................................................xvii

Table S-2 Relationship of Recommended Actions to Workplan 2012-
2042 Tasks ......................................................................... xviii
Table S-3 Responsible Department...................................................... xxi
Table S-4 Recommended Actions by Part ............................................ xxii
Table I-1 Implementation Timeframe ...................................................I-4
Table I-2 Relationship of Recommended Actions to Workplan 2012-
2042 Tasks ...........................................................................I-5
Table I-3 Responsible Department.......................................................I-7
Table I-4 Recommended Actions by Part ............................................. I-8
Table 1-1 Height and Age of All Buildings in Study Area .................... 1-15
Table 1-2 Predominant Features of Tall Building Cohorts of Special
Interest...............................................................................1-23
Table 1-3 Tall Building Inventory Data Fields ..................................... 1-33
Table 3-1 Performance-Based Design Criteria ..................................... 3-7
Table 3-2 Dynamic Vibration Periods of the RCSW Building ................ 3-8
Table 3-3 Dynamic Vibration Periods of the Steel BRBF.................... 3-10
Table 3-4 Site Class B Selected Ground Motions .............................. 3-14
Table 3-5 Site Class D Selected Ground Motions .............................. 3-14
Table 3-6 Engineering Demand Parameters for Evaluation of the
Reinforced Concrete Shear Wall Residential Building ........ 3-16
Table 3-7 Maximum Engineering Demand Parameters for
Evaluation of the Steel Buckling-Restrained Braced
Frame Office Building ........................................................3-18
Table 3-8 Median Impeding Factor Delay Estimates for the
Baseline Case and with Recovery-Planning Measures
in Place ..............................................................................3-20
Table 3-9 Summary of Median Loss Results and Associated Drifts ... 3-22

ATC-119-1 List of Tables xxxi

 Table 3-10 Summary of Median Repair Times and Functional
Recovery Time ...................................................................3-24
Table 4-1 Dates Required to Demonstrate Building Compliance ........4-25
Table 5-1 Height Precedents in Past and Current Codes and
Policies ..............................................................................5-11
Table 6-1 ATC-20 Posting Classifications ..........................................6-25
Table 6-2 Concrete and Steel Damage Patterns Associated with an
ATC-20 UNSAFE Posting ..................................................6-26
Table 6-3 Comparison of ATC-20 and FEMA 352 Terminology and
Evaluation Methods............................................................6-33
Table 6-4 Comparison of Minimum Scope of ATC-20 Rapid and
FEMA 352 Preliminary Methods .........................................6-33
Table 7-1 Tabular Format for Identifying Mitigation Priorities
Triggered by Alterations .....................................................7-13
Table 7-2 Dates Required to Demonstrate Building Compliance ........7-34

xxxii List of Tables ATC-119-1

 List of Abbreviations

AB administrative bulletin
ATC Applied Technology Council
BIC Building Inspection Commission
CAPSS Community Action Plan for Seismic Safety
CEBC California Existing Building Code
DBI Department of Building Inspection
DD disproportionate damage
DEM Department of Emergency Management
DPW Department of Public Works
FEMA Federal Emergency Management Agency
IEBC International Existing Building Code
ORCP Office of Resilience and Capital Planning
SEAONC Structural Engineers Association of Northern California
SFEBC San Francisco Existing Building Code
SFFD San Francisco Fire Department
SSD substantial structural damage

ATC-119-1 List of Abbreviations xxxiii

 Introduction to Tall Buildings Study

I.1 Background

In 2017, the City and County of San Francisco initiated the first project in the
nation to consider the impact of earthquakes on a city’s tall buildings. The
project conducted investigations in seven focus areas under separate tasks.
This project supplements the CAPSS Earthquake Safety Implementation
Program Workplan 2012-2042 published by the City in 2011 as a result of the
Community Action Plan for Seismic Safety (CAPSS) in response to Mayor
Newsom’s Executive Directive 10-022. The Workplan 2012-2042 anticipates
programs for broad groups of new and existing buildings throughout the city.

In 2018, the City recognized that tall buildings may need special attention
through exemptions, programmatic solutions, or specific technical criteria, to
make the work described in the Workplan 2012-2042 feasible. This is
because difficult exterior access, multiple tenants and uses within a building,
and their sheer size complicate evaluation and retrofit; their structural
systems preclude generic performance assumptions and prescriptive
engineering solutions; and San Francisco’s new and existing tall buildings
represent a dominant portion of the City’s business sector. In addition, their
high concentration downtown poses an aggregate risk to neighborhood and
citywide recovery not presented by other building groups.

I.2 Project Objectives and Scope

The primary objective of this study is to examine the earthquake performance

of San Francisco’s tall buildings and develop recommendations to address
building code requirements, policies and practices for the design of new
buildings, assessment and retrofit of existing buildings, and post-earthquake
inspection and response to promote the earthquake resilience of San
Francisco. To achive this objective, work was conducted in seven focus
areas under separate tasks. This report documents the findings and
recommendations resulting from each of the focus areas. It is noted that the
findings and recommendations here flow from a study of these tall buildings,
but most are also applicable to a wider set of buildings supporting similar
functions or posing similar risks. Tall buildings, even in downtown San
Francisco, are only part of a neighborhood’s building stock, and from a public

ATC-119-1 Introduction to Tall Buildings Study I-1

 policy perspective, their earthquake performance is bound up with that of the
shorter buildings around them.

The initial task of the project was to quantify the tall buildings in the City in
terms of height, age, usage, and structural system characteristics. The
building inventory database is developed in a Geographical Information
Systems (GIS) format to facilitate analysis and visualization of the tall building
inventory.

The project team served as principal authors of the findings and

recommendations of the focus areas. The findings were developed by
means of a literature review and consultation with other professionals, where
available. The City provided guidance and feedback through regular
meetings with the Tall Buildings Executive Panel (TBEP). In addition, key
stakeholders representing elected officials, private engineers and architects,
developers, community organizations, and city officials were convened by the
City to receive updates on the study’s progress and provide feedback.

I.3 Intended Audience and Next Steps

This report presents the project team’s findings and recommendations. While
the Summary Recommendations of this report can be used for big picture
implementation, a thorough review of the discussion presented in this report
is critical for the success of any implementation activity.

Table I-1 presents the recommendations in terms of potential implementation

timeframes. Short-term actions can be started with essentially no additional
study and completed without a legislative process. Typical examples include
administrative updates, development of Administrative Bulletins, and
initiatives requiring only nominal inter-departmental coordination. Mid-term
actions can normally be completed without substantial additional study and
without a legislative process but are expected to involve substantial input
from multiple stakeholder groups. Long-term actions are expected to require
substantial additional technical study or a legislative process.

Table I-2 relates the recommendations to relevant to tasks in Workplan 2012-

2042. This connection will ensure that the recommendations fit in the City’s
existing programs. In some cases, this study identified additional programs
and initiatives not detailed in Workplan 2012-2042. In others, the tall building
recommendations identify specific technical criteria, exemptions, or
programmatic solutions suitable for tall and similarly complex buildings, even
relaxing some of the Workplan’s broad directives.

I-2 Introduction to Tall Buildings Study ATC-119-1

Table I-3 presents the summary recommendations in terms of City
departments responsible for implementation. Implementation of any new
policy is assumed to involve appropriate coordination with other City
departments, outside experts (as needed), and other stakeholders. Some
recommended actions require enactment of legislation by the Mayor and
Board of Supervisors or action by the Building Inspection Commission and
can only commence after these approvals. In Table I-2, the departments are
identified as follows:
• BIC Building Inspection Commission
• DataSF DataSF
• DBI Department of Building Inspection
• DEM Department of Emergency Management
• DPW Department of Public Works
• ORCP Office of Resilience and Capital Planning
• SFFD San Francisco Fire Department
I.4 Report Organization

This report documents findings of seven focus areas and related

recommendations. Each focus area is covered in a “Part” of this report.
Each Part consists of multiple chapters that document findings and
recommendations, as well as necessary appendices and references.
Although the focus areas are documented distinctly, where substantial
overlaps are identified, references are made to relevant Parts.
• Part 1 presents the tall building inventory
• Part 2 summarizes best practices for geotechnical engineering
• Part 3 studies performance expectations for new buildings
• Part 4 studies post-earthquake structural evaluation
• Part 5 studies tall building effects on post-earthquake recovery
• Part 6 studies post-earthquake safety inspection of buildings
• Part 7 studies pre-earthquake evaluation of buildings

Table I-4 presents the relationship of Summary Recommendations to Parts of

this report for reference.

ATC-119-1 Introduction to Tall Buildings Study I-3

Table I-1 Implementation Timeframe
Short- Mid- Long-
Recommended Action Term Term Term
1. Actions for Reducing Seismic Risk Prior to Earthquakes – New Buildings
1A. Develop Regulations to Address Foundation and Geotechnical Issues
Training and checklist X
Develop geotechnical regulations X
1B. Establish Recovery-Based Seismic Design Standards X
2. Actions for Reducing Seismic Risk Prior to Earthquakes – Existing Buildings
2A. Apply the Repair Provisions of the San Francisco Existing Building Code with Respect
X
to Possible Loma Prieta damage
2B. Amend the San Francisco Existing Building Code Triggers
Alteration and change of occupancy triggers X
Acquisition triggers X
2C. Recommend Minimum Levels of Earthquake Insurance or Other Collateral to Ensure
X
Post-Earthquake Recovery
2D. Review Requirements for Post-earthquake Fire Suppression and Evacuation Systems X
3. Actions for Reducing Seismic Risk Following Earthquakes
3A. Develop New Policies and Procedures for Implementing the State’s Safety
X
Assessment Program
3B. Extend and Improve the Building Occupancy Resumption Program
Conduct simulation-based training X
Update procedures X
Extend program X
3C. Clarify and Update Roles and Responsibilities Associated with Post-Earthquake
X
Emergency Response and Safety Inspection
3D. Update and Amend the San Francisco Existing Building Code Triggers for Repair
X
Projects
3E. Update Administrative Bulletin 099 and Clarify its Application to Tall Concrete
X
Structural Systems
3F. Develop a New Administrative Bulletin for Post-Earthquake Inspection and Evaluation
X
of Welded Steel Moment Frames
3G. Create Protocols and Procedures for Establishing Cordons around Damaged Buildings X
3H. Require Existing Buildings to File Recovery Plans X
4. Actions to Improve the City’s Understanding of its Tall Building Seismic Risk
4A. Maintain and Expand the Database of Tall Buildings X
4B. Develop a Comprehensive Recovery Plan for the Financial District and Adjacent
X
Neighborhoods

I-4 Introduction to Tall Buildings Study ATC-119-1

Table I-2 Relationship of Recommended Actions to Workplan 2012-2042 Tasks
Recommended Action Workplan 2012-2042 Task
1. Actions for Reducing Seismic Risk Prior to Earthquakes – New Buildings
1A. Develop Regulations to Address Foundation and New action
Geotechnical Issues
1B. Establish Recovery-Based Seismic Design Standards B.6.a Update code for new buildings to reflect desired
performance goals
2. Actions for Reducing Seismic Risk Prior to Earthquakes – Existing Buildings
2A. Apply the Repair Provisions of the San Francisco Existing B.4.b Develop post-earthquake repair and retrofit standards
Building Code with Respect to Possible Loma Prieta
damage
2B. Amend the San Francisco Existing Building Code Triggers C.1.a Mandatory evaluation on sale or by deadline
C.1.b Evaluation of buildings retrofitted prior to 1994 or building
to non-conforming performance standards
C.2.a Mandatory retrofit of older non-ductile concrete residential
buildings
C.2.d Mandatory evaluation and retrofit of pre-1994 welded
steel moment frame buildings
C.2.e Mandatory evaluation and retrofit of other low-
performance buildings
2C. Recommend Minimum Levels of Earthquake Insurance or A.1.b Provide information and assistance about insurance
Other Collateral to Ensure Post-earthquake Recovery
2D. Review Requirements for Post-earthquake Fire Suppression A.6.i Study fire-related earthquake resilience topics
and Evacuation Systems
3. Actions for Reducing Seismic Risk Following Earthquakes
3A. Develop New Policies and Procedures for Implementing the A.4.f Update post-earthquake inspection (ATC-20) policies and
State’s Safety Assessment Program procedures
3B. Extend and Improve the Building Occupancy Resumption B.1.b Develop non-structural upgrade program for businesses
Program
3C. Clarify and Update Roles and Responsibilities Associated Procedural update
with Post-earthquake Emergency Response and Safety
Inspection
3D. Update and Amend the San Francisco Existing Building B.4.b Develop post-earthquake repair and retrofit standards
Code Triggers for Repair Projects
3E. Update Administrative Bulletin 099 and Clarify its Application A.4.d Adopt disproportionate damage trigger
to Tall Concrete Structural Systems B.4.b Develop post-earthquake repair and retrofit standards
3F. Develop a New Administrative Bulletin for Post-earthquake A.4.d Adopt disproportionate damage trigger
Inspection and Evaluation of Welded Steel Moment Frames B.4.b Develop post-earthquake repair and retrofit standards
3G. Create Protocols and Procedures for Establishing Cordons Program update, new action
around Damaged Buildings
3H. Require Existing Buildings to File Recovery Plans B.1.b Develop non-structural upgrade program for businesses
Note: The Earthquake Safety Implementation Program (ESIP) Workplan 2012-2042 was published by the City and County of San
Francisco in 2011 as a result of the Community Action Plan for Seismic Safety (CAPSS) in response to Mayor Newsom’s
Executive Directive 10-02.

ATC-119-1 Introduction to Tall Buildings Study I-5

Table I-2 Relationship of Recommended Actions to Workplan 2012-2042 Tasks(continued)
Recommended Action Workplan 2012-2042 Task
4. Actions to Improve the City’s Understanding of its Tall Building Seismic Risk
4A. Maintain and Expand the Database of Tall Buildings A.2.b Adopt façade maintenance regulations
4B. Develop a Comprehensive Recovery Plan for the Financial A.4.a Develop and adopt Shelter-in-Place policies and
District and Adjacent Neighborhoods procedures
B.2.b Mandatory evaluation of 5+ dwelling unit residential
buildings and hotels/motels
C.2.b Mandatory evaluation and retrofit of critical stores,
suppliers, and service providers
C.2.c Mandatory evaluation and retrofit of larger (over 300
occupants) assembly buildings

I-6 Introduction to Tall Buildings Study ATC-119-1

Table I-3 Responsible Department
Data
Recommended Action BIC SF DBI DEM DPW ORCP SFFD
1A. Develop Regulations to Address Foundation and
X
Geotechnical Issues
1B. Establish Recovery-Based Seismic Design
X
Standards
2A. Apply the Repair Provisions of the San Francisco
Existing Building Code with Respect to Possible X
Loma Prieta damage
2B. Amend the San Francisco Existing Building Code
X
Triggers
2C. Recommend Minimum Levels of Earthquake
Insurance or Other Collateral to Ensure Post- X
Earthquake Recovery
2D. Review Requirements for Post-earthquake Fire
X X
Suppression and Evacuation Systems
3A. Develop New Policies and Procedures for
Implementing the State’s Safety Assessment X
Program
3B. Extend and Improve the Building Occupancy
X X
Resumption Program
3C. Clarify and Update Roles and Responsibilities
Associated with Post-Earthquake Emergency X X X
Response and Safety Inspection
3D. Update and Amend the San Francisco Existing
X X X
Building Code Triggers for Repair Projects
3E. Update Administrative Bulletin 099 and Clarify its
X
Application to Tall Concrete Structural Systems
3F. Develop a New Administrative Bulletin for Post-
Earthquake Inspection and Evaluation of Welded X X
Steel Moment Frames
3G. Create Protocols and Procedures for Establishing
X X X
Cordons around Damaged Buildings
3H. Require Existing Buildings to File Recovery Plans X X

4A. Maintain and Expand the Database of Tall

X X X
Buildings
4B. Develop a Comprehensive Recovery Plan for the
X X
Financial District and Adjacent Neighborhoods

ATC-119-1 Introduction to Tall Buildings Study I-7

Table I-4 Recommended Actions by Part
Recommended Action Part 1 Part 2 Part 3 Part 4 Part 5 Part 6 Part 7
Actions for Reducing Seismic Risk Prior to Earthquakes – New Buildings
1A. Develop an AB and Code Amendment that
X
Addresses Foundation and Geotechnical Issues
1B. Establish Performance-Based Seismic Design
X
Standards
Actions for Reducing Seismic Risk Prior to Earthquakes – Existing Buildings
2A. Apply the Repair Provisions of the SFEBC with
X
Respect to Possible Loma Prieta Damage
2B. Amend the SFEBC Triggers X
2C. Require Minimum Levels of Earthquake
X
Insurance to Ensure Recovery
2D. Increase Local Water Supply for automatic Fire
X
Suppression Systems in Tall Buildings
Actions for Reducing Seismic Risk Following Earthquakes
3A. Develop New Policies and Procedures for
X X
Implementing the State’s SAP
3B. Extend and improve BORP X
3C. Clarify and Update Roles and Responsibilities
Associated with Post-earthquake Emergency X X
Response and Safety Inspection
3D. Update and Amend the SFEBC X
3E. Update AB-099 and Clarify its Application to Tall
X
Concrete Structural Systems
3F. Develop a New AB to Implement FEMA 352 for
Post-earthquake Inspection and Evaluation of X X
Welded Steel Moment Frames
3G. Create Protocols and Procedures for Establishing
X
Cordons Around Damaged Buildings
3H. Require Existing Buildings to File Recovery Plans X
Actions to Improve the City’s Understanding of its Tall Building Seismic Risk
4A. Maintain and Expand the Database of Tall
X
Buildings
4B. Develop a Comprehensive Recovery Plan for the
X
Financial District and Adjacent Neighborhoods

I-8 Introduction to Tall Buildings Study ATC-119-1

 PART 1:
Inventory of Tall Buildings
Chapter 1

Introduction

1.1 Background

San Francisco’s building inventory has grown significantly over the past
century, with much of the current development rooted in the rebuilding
following the 1906 earthquake and fire. This rebuilding included significant
land development along the San Francisco waterfront. Since the late 1950s,
San Francisco’s skyline has changed dramatically with the construction of tall
buildings in the downtown financial district. Significant growth over the past
fifteen years has included high-rise residential and mixed-use buildings
concentrated in the region south of Market Street. Among the multi-faceted
earthquake risks facing the city, the concentration of tall buildings and
infrastructure in the densely populated downtown neighborhoods is raising
questions about the risks to life, property, and recovery from large
earthquakes in San Francisco.

As a first step towards addressing these questions, this report describes the
development of an inventory of tall buildings in San Francisco. The primary
motivation is to quantify the tall buildings in terms of height, age, usage, and
structural system characteristics of the buildings. Although not the only
earthquake risk, tall buildings are of special concern due to their size and
large occupant loads, where earthquake damage to one tall building can have
disproportionate effects on its occupants, its neighbors, and the community at
large. The inventory is intended to help inform policies and practices to
manage risks associated with existing buildings and to improve the planning
and design of future buildings. The inventory can further serve to inform
planning for emergency response and recovery.

One of the key decisions in establishing the tall building inventory is to

establish a definition of “tall.” While the immediate emphasis of the tall
building inventory development is on buildings taller than 240 feet (roughly 18
to 24 stories tall), this height is somewhat arbitrary insofar as all buildings are
vulnerable to earthquake damage and have implications on the seismic
resilience of San Francisco. The 240-foot value is based on a threshold for
certain seismic design requirements for buildings in national standards that
are adopted by the San Francisco Building Code (CCSF, 2016). Further
discussion is provided in Part 5. While 240 ft is a convenient threshold to

ATC 119-1 Part 1: Inventory of Tall Buildings 1-1

 define the initial scope of this inventory, apart from this practical aspect, the
240-foot height does not have any empirical or scientific basis in how
buildings respond to earthquakes. In this regard, any planning, policy, or
other recommendations developed on the basis of the inventory would be
equally relevant for shorter buildings.

This Part summarizes the development of the inventory and presents

recommendations for maintenance and update of the inventory. This Part
corresponds to Recommendation 4A presented in Summary
Recommendations.

This Part is not specifically covered in CAPSS Earthquake Safety

Implementation Program’s Workplan 2012-2042 (CCSF, 2011), but
addresses aspects of the following task:
• Task A.2.b Adopt façade maintenance regulations

1.2 Inventory Overview

For this study, two databases were assembled. The first is a general
database of all buildings in the downtown neighborhoods of San Francisco,
which was compiled from building information available from DataSF
(https://datasf.org/), the San Francisco Property Information Map (CCSF,
2017a), and other sources. Shown in Figure 1-1 is a map of buildings from
this database in downtown San Francisco, including the Financial District and
the South of Market neighborhoods. This map shows the concentration of tall
buildings in the downtown region, distinguished between height groupings
with thresholds of 75 ft (6 to 8 stories), 160 ft (12 to 16 stories) and 240 ft (18
to 24 stories). The 75-foot threshold is based on the fire safety code in
Section 403 of the California Building Code (CBSC, 2016), which requires
special provisions for “high-rises” over 75 ft. The 160-foot and 240-foot limits
are based on building code requirements for seismic design. A subset of
these buildings covering the blocks with the highest concentration of tall
buildings was also produced and is discussed in Part 2 Section 2.2.

The second database is a more detailed inventory of buildings over 240 ft tall
that includes information on building location, height, occupancy, age of
construction, construction material, structural system, foundation type, façade
system, and other relevant design information. This was compiled from the
general data sources listed above, combined with surveys of building permit
drawings for all buildings taller than 240 ft. The locations and schematic
geometries of these tall buildings are shown in Figure 1-2. The locations are
superimposed on a map where beige shading identifies areas that are
susceptible to soil liquefaction during strong earthquakes, which is one of

1-2 Part 1: Inventory of Tall Buildings ATC 119-1

many risk considerations addressed in their design. The database includes
156 buildings taller than 240 ft, based on currently available data. The
database also includes some additional buildings between 160 ft to 240 ft tall,
although this portion of the list is incomplete and does not reflect the
complete inventory in this height range.

Figure 1-1 Spatial distribution of San Francisco’s downtown building

inventory mapped as a function of building height.

Figure 1-2 Spatial distribution of San Francisco’s tall building inventory.

Footprints of buildings not included in the inventory are shown
for reference. Beige denotes liquefaction zone (per DataSF).

ATC 119-1 Part 1: Inventory of Tall Buildings 1-3

 Further details and statistics from both databases are discussed later in
Section 2. The two databases are compiled in an industry standard
Geographic Information System (GIS) Shapefiles, which can be viewed and
interrogated using a variety of open source and proprietary software systems.

1.3 Data Fields and Source

Damage to tall buildings has the potential to affect a large number of people
and can have significant consequences on surrounding areas. Susceptibility
to earthquake damage is attributed to several factors including: intensity of
ground motion shaking, soil characteristics at the site, age of construction,
building material, structural system, cladding, interior finishes, and
mechanical and electrical components. Many of these features, which are
identified in the database, can help understand the expected seismic
performance of these building under varying intensities of ground motion
shaking.

The data fields contained within the tall building inventory can be classified
into the following broad categories:
• Building name and location
• Height (by feet and number of stories)
• Primary occupancy
• Date (permit or completion, depending on source)
• Retrofit information (if applicable/available)
• Construction material
• Structural system (i.e., lateral force resisting system)
• Foundation system
• Geotechnical information (e.g., soil class, depth to rock)
• Massing (e.g., overall dimensions, square footage)
• Irregularities (e.g., podiums, setbacks)
• Façade composition (e.g., pre-cast concrete panels, glass)
• Fire-suppression information (i.e., system type and rating)
• Presence of strong motion instrumentation
• Other remarks and observations

A detailed list of all the variables included within the database is included in
Appendix A. Although not all parameters are available for all buildings, the

1-4 Part 1: Inventory of Tall Buildings ATC 119-1

available information provides a fairly comprehensive description of buildings
taller than 240 ft and allows for the expansion of the database as additional
information becomes available.

Numerous data sources enabled the development of the tall building

inventory. Review of available information found that because these data
sources are developed for different purposes, information for a single building
parameter can be conflicting across sources. For instance, when reporting
square footage, some sources may report gross square footage, while others
may indicate usable square footage. When conflicting parameters were
identified from different source documents, the most representative values
were selected based on judgement for inclusion in the database. The
following are the key sources of data used in the compilation of the tall
building inventory:
• Construction permit documents at the San Francisco Department of
Building Inspection (DBI)
• Public data from the City and County of San Francisco’s Planning
Department (CCSF, 2017)
• San Francisco Fire Department data
• Building Occupancy Resumption Program (BORP) reports (discussed
further in Part 6)
• Existing literature, including a previous database developed by the
Structural Engineers Association of Northern California (Molina Hutt et al.
2016, Almufti et al. 2018, USGS, 2018)
• Interviews and surveys with local professional engineers.
• Emporis (https://www.emporis.com/)

Appendix A identifies the major source of data for each data field included in
the tall building inventory.

In addition to data fields discussed above, the GIS shape file also includes
spatial location polygons representing the footprint of each building. Polygon
footprint data for buildings that existed during a LIDAR flight, conducted on
behalf of the City in 2010, are taken from the LIDAR dataset available on
DataSF. For buildings constructed since 2010, the footprint is approximated.
Note that these polygons are based on the footprints at ground level and
include, for example, podium structures that are not as tall as the main tower
of the building.

ATC 119-1 Part 1: Inventory of Tall Buildings 1-5

 1.4 Maintaining an Updated Inventory

Up-to-date and comprehensive building databases are an important resource

to better understand and quantify vulnerability of the city to earthquakes, fire
conflagrations, bayshore flooding, and other hazards. When coupled with
modern computational and information technology software, detailed
databases enable reliable assessment and implementation of policies and
other measures to reduce vulnerability and increase resilience of the city. By
incorporating information on seismic retrofit of buildings, BORP participation,
and other measures to improve resilience, the database can provide City
administrators data on building owner behavior and the effectiveness of
voluntary and mandatory policies. Chapter 3 of this Part includes information
and suggestions of three categories of tall buildings where tracking and
assessment of policy implementations could be particularly worthwhile.

To help facilitate maintenance and updating of the tall building inventory,

Appendix B provides a recommended form for consideration by San
Francisco DBI as a supplement to building permit applications (retrofit and/or
new construction). The proposed supplement would enable maintenance of
the database with information on new or retrofitted buildings, including the
following key parameters: building name and location, height, occupancy,
construction date, construction material, structural system, foundation system
and façade composition. It is further recommended that San Francisco DBI
consider developing mechanisms to collect information from other inspections
and reports, such as those required by San Francisco’s façade ordinance
(CCSF, 2017b), to supplement the database. To the extent that all of the
information in the database was obtained from publicly available sources, the
database can be made publicly available.

1.5 Organization

Chapter 2 describes the characterization of building inventory features,

presents a study of all buildings in a limited study area in downtown San
Francisco for context-setting, and presents an overview of seismic codes and
other considerations.

Chapter 3 describes tall building cohorts identified when reviewing the

database of tall buildings.

Chapter 4 presents a summary of the inventory and recommendations.

Appendix A presents attributes present in the tall building inventory and

descriptions.

1-6 Part 1: Inventory of Tall Buildings ATC 119-1

Appendix B sets forth a supplementary form to be used during the building
permit process to enhance inventory maintenance and expansion.

Appendix C presents the background for bedrock depth data in San

Francisco.

A list of references is provided at the end of this Part.

ATC 119-1 Part 1: Inventory of Tall Buildings 1-7

Chapter 2

Characterization of Building
Inventory Features

2.1 Inventory of Buildings Taller than 240 Feet

The tall building inventory includes detailed information on 156 buildings over
240-feet tall, which currently exist or have been permitted for construction in
San Francisco. The specifics of the structural materials and systems for
these buildings were identified by reviewing structural drawings on file in the
building permit and Building Occupancy Resumption Program (BORP) report
databases maintained by San Francisco Department of Building Inspection
(DBI) (more discussion on the BORP program is in Part 6). The tall building
inventory also includes information on an additional 22 buildings between 160
ft to 240-ft tall, which have been collected from other sources. The list of
buildings below 240-feet tall, and details of their structural systems and
materials, are not complete since the first priority for the database was to
focus on buildings above 240 ft.

Summary statistics from the database of buildings 240-ft tall and above are
provided below according to height, occupancy, construction date,
construction material and structural configuration.
• Height: Roughly 10% of the inventory is below 20 stories, 70% of the
inventory is in the 20- to 40-story range, 20% is above 40 stories.
• Occupancy: Approximately 60% of the buildings are commercial and
just under 40% are residential or hotel.
• Construction date: Over 55% of the tall building inventory was
constructed between 1960 and 1990. Tall building construction slowed
down in the 1990s, but has resurged since then, with almost 25% of the
inventory constructed since 2000.
• Construction material: Structural steel systems account for about 65%
of the inventory, reinforced concrete systems account for about 20% of
the buildings, and the remainder either incorporate mixed steel-concrete
systems or, in a few instances, the systems were unidentified.
• Structural configuration: Approximately 50% of the buildings employ
steel moment-resisting frames as the seismic-force-resisting system

ATC 119-1 Part 1: Inventory of Tall Buildings 1-9

 (SFRS), 10% have steel braced frame systems, and about 30% have
reinforced concrete shear wall systems. Consistent with the building
codes and engineering practice of the 1960-1990 period, most of the steel
braced frame and the older concrete shear wall systems are combined
with a moment resisting frame, forming a dual system. Many concrete
shear wall buildings constructed since the introduction of performance-
based seismic design procedures around 2007 are shear wall only
systems. About 10% of the older buildings in the inventory did not have
enough available information to identify a structural system.

Figure 1-3 illustrates the fraction of buildings according to their SFRS, decade
of construction, and number of stories. Superimposed on the figure are notes
indicating two significant earthquake events (San Fernando in 1971 and
Northridge 1994) and the advent of performance-based design procedures
(2007) that relate to significant changes building code requirements and
design practices, which are discussed further in the next section.

Figure 1-4 presents the inventory distinguished by building occupancy.

Whereas the structural system and construction date are relative indicators of
expected seismic performance, the occupancy type provides insights into the
services that tall buildings within the inventory provide to the community.

Three noteworthy trends in the data in Figures 1-3 and 1-4 are: (1) the
predominance of steel moment-resisting frame construction during the 1960s,
70s and 80s; (2) a transition from steel moment-resisting frames to steel
braced frame dual systems in the 1990s; and (3) the emergence of concrete
shear wall systems since 2000. The trend away from steel moment-resisting
frame to braced frames is due to a combination of architectural preferences
towards taller buildings with more irregular geometries, combined with
advancements in computer methods that facilitated their analyses. The post-
2000 emergence of concrete shear walls is in large part due to the
construction of high-rise residential buildings, whose architectural
configurations (smaller floor plate sizes and room layouts, and smaller
occupant densities with lower elevator and HVAC requirements) make them
more amenable to concrete construction.

Figure 1-5 presents an alternate view to illustrate the number of buildings as

a function of construction date and number of stories. The red shaded
portion reveals that about 35% of the tall building inventory is in the range of
15 to 29 stories, built in the 1960s through 1980s. The blue shaded portion
highlights more recent construction with a clear trend toward modern
buildings of 40 stories, representing just over 12% of the inventory, built since
2000.

1-10 Part 1: Inventory of Tall Buildings ATC 119-1

 (a)

Performance-Based (2007)
Northridge (1994)
San Fernando (1971)

(b)

(c)
Figure 1-3 Figure showing 156 San Francisco buildings tall than 240 ft by: (a)
percentage of buildings by SFRS type, (b) breakdown of SFRS by
decade, and (c) distribution of SFRS by number of stories.

ATC 119-1 Part 1: Inventory of Tall Buildings 1-11

 (a)

(b)

(c)
Figure 1-4 Figure showing 156 San Francisco buildings taller than 240 ft by: (a)
percentage of buildings by occupancy type, (b) distribution of
occupancies by decade, and (c) distribution of occupancies by
number of stories.

1-12 Part 1: Inventory of Tall Buildings ATC 119-1

Figure 1-5 Figure showing distribution of the number of stories by decade.

Figure 1-6 provides a distribution of the foundation systems used, with over
half of the buildings on pile foundations and about a quarter on shallow mat
foundations. The remainder of foundations consist of shallow footings, pile
supported mats, or drilled shafts. Drilled shaft foundation techniques were
introduced relatively recently to San Francisco, and used in several tall
buildings constructed in sites with depth to bedrock greater than about 200 ft.

(a) (b)
Figure 1-6 Figure showing: (a) percentage of buildings by foundation type, and (b) breakdown of
foundation types by depth to bedrock.

2.2 A Closer Look: Financial District Building Inventory

To understand San Francisco’s tall buildings in context, a limited study was

conducted considering every building, of any height, in a 31-block portion of
downtown, shown in Figure 1-7a.

ATC 119-1 Part 1: Inventory of Tall Buildings 1-13

 The study area does not include all of San Francisco’s tallest or most iconic
buildings. The Transamerica Tower, Embarcadero Center, Millennium
Tower, and Salesforce Tower are all outside the study area. However, the
study area does cover the downtown neighborhoods with the highest
concentration of tall buildings. Were the study area to be extended one block
in each direction, the basic findings would not change significantly.

The study area contains 243 distinct buildings. The structural system, retrofit
status, use/occupancy, and occupant load of each building are of interest to a
comprehensive resilience-based inventory, but for this context study, only the
most basic building data – height, building area, and age – were compiled.
Figure 1-7b shows the distribution of height and dates of the buildings.

(a) (b)
Figure 1-7 Map showing 243 buildings in a 31-block downtown San Francisco
study area: (a) the study area, focused on the blocks with the highest
concentrations of tall buildings; and (b) the height and dates of
buildings in the study area, with the percent in each height category.

The height, area, and age data were compiled from Datasf.org, San
Francisco Property Information Map (CCSF, 2017a), Sanborn maps, and
satellite and street level imagery. In addition, limited information for the 243
buildings were found or confirmed from online searches by address, with data
coming from websites produced by building owners, leasing agents,
contractors, architects, or building historians. The resulting dataset, while not
perfect, is more complete and reliable than the data from any single source.

Table 1-1 summarizes the data from Figure 1-7 in bins by height and age in
count normalized to the total number of buildings. The height bins are based

1-14 Part 1: Inventory of Tall Buildings ATC 119-1

on the precedent building code limits (75 ft, 160 ft, and 240 ft) as well as an
arbitrary distinction at 400 ft. The age bins are based on San Francisco’s
development history, as evident from Figure 1-7, and on the date of a key
change in steel building design that was implemented following the 1994
Northridge earthquake.

Table 1-1 Height and Age of All Buildings in Study Area

Number of Buildings
Height [ft] Count Pre-1915 1915-1950 1951-1994 Post-1994
401+ 18 0 1 15 2
241 – 400 33 0 5 25 3
161 – 240 28 3 9 14 2
75 – 160 71 41 14 14 2
Up to 74 93 54 19 18 2
All 243 98 48 86 11
Normalized to total of 243
Height [ft] All Pre-1915 1915-1950 1951-1994 Post-1994
401+ 7% 0% 0% 6% 1%
241 – 400 14% 0% 2% 10% 1%
161 – 240 12% 1% 4% 6% 1%
75 – 160 29% 17% 6% 6% 1%
Up to 74 38% 22% 8% 7% 1%
All 100% 40% 20% 35% 5%

While the age bins in Table 1-1 are continuous, Figure 1-7 reveals trends
within each bin:
• The Pre-1915 bin captures the decade of rebuilding that followed the
1906 earthquake and fire, but there is an obvious concentration in the first
few years. Sixty-two of the 98 buildings in this group date from 1906 to
1908.
• The 1915-1950 bin includes the building boom of the 1920s, in which
much of San Francisco’s housing stock was built. Forty-three of the 48
buildings in this group were built before 1930.
• The 1915-1950 bin also shows the gap in development during the Great
Depression and World War II, during which San Francisco added only a
handful of buildings to its downtown.

ATC 119-1 Part 1: Inventory of Tall Buildings 1-15

 • The 1951-1994 bin includes the mini-boom of tall buildings in the 1970s
and 1980s. Within this 44-year period, 54 of the 86 buildings were built in
those two decades.

Table 1-1 and Figure 1-7 show, unsurprisingly, a relationship between height
and age. Very few of San Francisco’s tallest buildings were built before
1960, and most of the shorter buildings (even up to 160 ft) were built before
1930. This trend will become stronger as each new tall building replaces
several older and shorter ones.

Available data did not support a study of building area relative to height or
age. Even so, it is obvious that the taller buildings have much larger areas;
the 51 buildings taller than 240 ft likely contain more total square footage than
the 192 shorter buildings combined. Therefore, a structurally deficient tall
building represents a greater potential loss than a similarly deficient shorter
building.

On the other hand, the total street frontage presented by buildings in different
height categories is likely to include a higher percentage of lower buildings.
Just as the tall buildings have greater volume, the more numerous short
buildings front a larger share of downtown streets and sidewalks. Cladding
failure and other nonstructural falling hazards can be as dangerous from 50
feet as from 250 feet. So where falling hazards or leaning buildings are of
concern, it is possible that the shorter buildings, as a group, may pose
greater risks to more blocks and more public spaces than the tall buildings,
especially considering their age and obsolete materials. San Francisco’s
façade inspection requirements are likely to increase understanding on this
issue.

2.3 Evolution of Seismic Building Codes

Changes to the seismic and other design requirements that have been made
to the San Francisco Building Code over the past century offer important
clues to interpreting the building databases. Building codes are living
documents that are revised and updated over time, often in response to
observations from damaging earthquakes, development of new building
technologies, and most recently to advancements in building performance
assessment methods. Building code requirements also evolve as the risk or
the expected performance associated with such requirements is deemed
unacceptable (FEMA, 2006). Therefore, older buildings may include
deficiencies that were unrecognized by the building codes and practices
under which they were designed and constructed but have since been
acknowledged in newer building codes. This section provides a brief review

1-16 Part 1: Inventory of Tall Buildings ATC 119-1

of the evolution in seismic building codes as related to implications on the
safety and performance of tall buildings.

The Uniform Building Code (UBC) was first published in 1927 to help promote
public safety and standardized construction. The code was updated every
three years until 1997, which was the final version of the code before the
introduction of the International Building Code (IBC) in 2000. One significant
change over the course of the UBC revisions was the inclusion of seismic
design code provisions. From 1927 to 1961, the UBC had a non-mandatory
appendix, until mandatory seismic design requirements were introduced via
recommendations published in 1959 by the Structural Engineers Association
of California (SEAOC) in Recommended Lateral Force Requirements and
Commentary, or the “SEAOC Blue Book.” This established performance
goals that still underlie today’s building codes. The code’s minimum design
provisions were intended to provide earthquake safety. Most were based on
fundamental engineering principles, but many, including height limits and
height-related details, were set by consensus engineering judgment. The
requirements included minimum lateral strength provisions, simplified design
methods and other prescriptive requirements to help avoid deficiencies that
led to severe building damage and collapse in past earthquakes.

Over the course of time, the definitions for seismic requirements have been
updated. In the 1970s, Cornell’s (1968) principles of seismic hazard analysis
were implemented to develop hazard maps that represented the intensity of
earthquake ground shaking with a 10% chance of being exceeded occurring
during the [assumed] 50-year design life of a building. Statistically, this
probability of exceedance is also characterized as a ground motion with a
475-year return period (the average time over which the ground motion is
typically exceeded). For many years, this level of intensity was the basis of
the design basis earthquake (DBE) in building codes. As more data became
available to characterize earthquake ground motion hazard and recognizing
the nonlinear relationship between ground motion risk and intensity, the
Maximum Considered Earthquake (MCE) was introduced into building codes
in the mid-1990s. In downtown San Francisco, the MCE ground motion has
roughly a 2% to 5% chance of being exceeded in 50 years, depending on the
soil conditions (about 2% in rock sites and 5% in soft-soil sites). This level of
ground motion is on the order of what could be expected by a magnitude-8
earthquake on a nearby segment of the San Andreas fault.

The ground motion hazard is one component of design, which defines the
expected ground shaking. The other component involves many assumptions
inherent in building code design and analysis requirements, which vary by
structural materials and systems. The combined risk is based on the

ATC 119-1 Part 1: Inventory of Tall Buildings 1-17

 combined probability of experiencing strong ground motions and the chance
of damage and collapse to those ground motions. The most recent building
codes, based on ASCE/SEI 7-16 (ASCE, 2017), are more explicit as to the
implied life safety risk to buildings, stating that a code-designed building
should have less than 10% probability of collapse under the Maximum
Considered Earthquake (MCE) ground motion. While this is a maximum
permissible risk, it is generally accepted that the actual risk of collapse in
modern code-designed tall buildings is less than the 10% rate under MCE
ground motions, although this risk is difficult to rigorously quantify.

For most buildings, design methodologies are still based on simplified

analysis methods and prescriptive requirements for proportioning structural
members and connections. The predominant method of design entails
calculating the lateral forces based on weight of the building modified by
coefficients to account for soil conditions, building height, the type of
structural system and other factors. Though these prescriptive
methodologies do not explicitly measure the seismic performance of the
design, they are generally accepted as providing the minimum requirement of
life safety for conventional buildings. However, these types of prescriptive
rules are limited to structural systems and configurations based on traditional
design and construction practices and observations of building performance
in past earthquakes. More recently, advanced performance-based analysis
and design approaches have been introduced for tall buildings.

From a structural response perspective, tall buildings have unique seismic

response characteristics including fundamental translation periods of
vibration well in excess of 1 second, significant mass participation and lateral
response in higher modes of vibration, as well as seismic resisting systems
with slender aspect ratios (PEER, 2017). These characteristics result in
distinct seismic behavior when compared to low- and mid-rise construction.
The limitations of code prescriptive requirements to tall building design are
apparent through height limitations associated with certain SFRS. For
example, the standard building code does not allow prescriptively designed
steel braced frame or concrete shear wall systems to be used in buildings
over 160 ft tall (or with certain restrictions 240 ft), unless they are combined
with moment-resisting frames (ASCE, 2017).

In response to the shortcomings and limitations of standard prescriptive

seismic design requirements, over the past decade, performance-based
seismic design guidelines have been developed and applied for tall building
design. In 2008, San Francisco issued Administrative Bulletin 083 (CCSF,
2008), which formally permitted the use of performance-based design
procedures and outlined several key design requirements. In 2010, the first

1-18 Part 1: Inventory of Tall Buildings ATC 119-1

edition of the Guidelines for Performance-based Seismic Design of Tall
Buildings was published (PEER, 2010), which was updated in 2017 and is
routinely employed for the design of tall buildings (greater than 240 ft tall) in
San Francisco. The Los Angeles Tall Building Structural Design Council has
published similar guidelines (LATBSDC, 2017), which have been adopted for
the design of new tall buildings in Los Angeles.

The so-called performance-based seismic design procedures employ

advanced nonlinear analysis to evaluate the expected building response to
large earthquakes and, thereby, provide greater assurance that buildings
meet the performance intended in current building codes to minimize life
safety risks under extreme earthquakes, and help control damage under
small frequent earthquakes. Moreover, by permitting exceptions to certain
prescriptive design requirements, the performance-based design procedures
allow opportunities to design more economical structural systems.

Further discussion of seismic design and performance for three predominant

types of structural systems (older steel framed buildings, older non-ductile
concrete buildings, and modern high-rise residential buildings) are included in
Chapter 3 of this Part.

2.4 Other Considerations

The inventory process also revealed the following about San Francisco’s tall
buildings:

Geotechnical Site Conditions: Much of San Francisco’s downtown is built

on soil fill placed between the late 1800s to mid-1900s. Liquefiable soils are
also within this area, usually in the top 30 ft of soil with denser soil (sand and
clay layers) below. The soil type and foundation system type can influence
the seismic response of tall buildings (see Part 2). The tall building
foundation types, summarized in Figure 1-6, and reported in the inventory are
selected based on the soil conditions at the site, combined with other design
parameters, including:
• Depth to rock (typically 0 to > 250 ft in San Francisco)
• Soil type or stiffness (typical San Francisco soil profiles include recent
marine deposits, Colma sands, and old bay clay).
• Building height and weight
• Number of basement levels
• Slope of soil surface or proximity to shoreline
• Ground water level

ATC 119-1 Part 1: Inventory of Tall Buildings 1-19

 • Proximity to adjacent or underlying structures
• Available construction technologies, logistics, and economics

Fire Regulations: Since 1976, San Francisco has required fire sprinklers in
all new buildings 75 feet high or taller. This threshold is based on limitations
of firefighters reaching stories about this height. In 1993, San Francisco
required retrofitting high-rise commercial buildings and hotels with sprinklers,
but the requirement excluded residential buildings. Nevertheless, even many
residential buildings have been retrofitted such that most buildings have at
least some percentage of their interior sprinklered. The database reports the
percentage of sprinklers that are installed in the 135 buildings that could be
cross-referenced with the Fire Department’s database. Of these buildings
with known information, 84% of them are over 95% sprinklered (i.e.,
essentially fully sprinklered), 7% are reported to be less than 30%
sprinklered, and 9% are in between 30% to 95% sprinklered.

Zoning Ordinances and Urban Planning: Zoning ordinances and urban

planning can influence the development of high-rise buildings. For instance,
prior to the mid-80s few tall buildings were located south of Market Street,
whereas the majority of post-2000 tall buildings built in this area of the city.
The Downtown Area Plan (CCSF, 1995), adopted in 1985, introduced a set of
policies to guide land use decisions that created today’s downtown. Urban
planning and zoning changes can continue to shape future developments.

BORP: Following the 1989 Loma Prieta earthquake, San Francisco

introduced the Building Occupancy Resumption Program (BORP) to facilitate
inspection and recovery following an earthquake. Participation in the
program is voluntary and requires an up-front investment by the building
owner to underwrite the preparation of an engineering report of the building
system. Fifty seven of the 156 buildings over 240-feet tall (37% of the
inventory) currently participate in BORP.

Building Instrumentation: As specified in San Francisco’s Administrative

Bulletin 058, Procedures for Seismic Instrumentation of New Buildings,
(CCSF, 2014) and the California Building Code (CBSC, 2016), all new
buildings over 10 stories tall constructed since about 2008 are required to
have seismic strong motion instrumentation. This instrumentation provides
data on how the building responds to earthquakes, which can facilitate post-
earthquake inspection and provide data on the performance of buildings,
which can inform earthquake safety policies and building code provisions and
ultimately lead to improved design and construction practices. Based on
information from the California Strong Motion Instrumentation Program
(CSMIP), an organization that records strong motion data, 8 of the 156

1-20 Part 1: Inventory of Tall Buildings ATC 119-1

buildings over 240-ft tall have strong motion instrumentation. This number is
contrasted with about 20 buildings constructed (or permitted) since 2010,
suggesting that either some buildings are instrumented but not included in the
CSMIP database or are otherwise out of compliance.

ATC 119-1 Part 1: Inventory of Tall Buildings 1-21

Chapter 3

Tall Building Cohorts

3.1 Identification of Tall Building Cohorts

The tall building inventory enables the identification of building clusters or

“cohorts” that best represent the City’s existing tall building stock. This
section identifies the following three cohorts which have relevance to
earthquake damage and recovery in San Francisco: older steel-framed
buildings, older nonductile concrete buildings, and modern concrete
residential buildings.

Table 1-2 summarizes the height range, occupancy type, construction dates,
construction material and structural system for the three cohorts of interest.
Sections 3.2 to 3.4 describe the rationale for the selection of each cohort
(e.g., high risk, prominent building type, disproportionate consequences
associated with poor performance, or unique design approach) and further
observations.

Table 1-2 Predominant Features of Tall Building Cohorts of Special Interest

Story Occupancy Construction Construction
Cohort Name Range Type Dates Material Structural System
Older Steel Buildings 20-40 Commercial Before mid- 1990s Steel Moment-resisting frame
Older Concrete Buildings Low 20s Commercial Before mid- 1970s Reinforced Concrete Shear Wall & Dual System
Modern Concrete Building >40 Residential Post 2000 Reinforced Concrete Shear Wall

3.2 Older Steel Buildings

Approximately 50% of the buildings within the inventory have steel moment-
resisting frame seismic-force-resisting systems (SFRS), and almost 90% of
these were constructed in the 1960s through the 1980s. These are identified
as a building cohort of primary interest owing to: (1) their prominence as the
most common system type in the tall building inventory; and (2) concern
regarding deficiencies in their design, including the potential for fracture
prone welded connections, which came to light following the 1994 Northridge
earthquake.

The heatmaps in Figure 1-8 illustrate the number of steel moment-resisting

frame buildings as related to construction date, number of stories, and

ATC 119-1 Part 1: Inventory of Tall Buildings 1-23

 occupancy type. Almost 70% of the steel framed structures are office
buildings, followed by about 15% that are hotels, with the remainder being
residential or mixed use. The predominance of the steel moment-resisting
frames of 20 to 39 stories is shown in Figure 1-8a, and the predominance of
office buildings constructed during the 1960-80s is evident in Figure 1-8b.

(a) (b)
Figure 1-8 Joint distribution of pre-1994 steel moment-resisting frame tall
buildings for (a) building date versus number of stories, and (b) date
versus occupancy type (values and color intensity denote the
number of buildings in each category).

Since the 1906 earthquake, structural engineers have generally regarded

steel moment-resisting frame systems as being among the most ductile and
reliable SFRS for buildings (FEMA, 2000a). The common view was that
when subjected to earthquake shaking, moment-resisting frames would
experience only localized damage due to ductile yielding of members and
connections. This led to widespread construction of this system, particularly
in the high seismic regions of the western United States (FEMA, 2000a). The
1994 Northridge earthquake dramatically changed existing perceptions, when
post-earthquake inspections revealed cracking in the beam-to-column joint
welds in several dozens of low- and mid-rise buildings.

In 1995, the Los Angeles City Council passed an ordinance mandating

connection inspections and repairs in steel moment-resisting frame buildings
that may have experienced damage. The ordinance only applied to regions
that experienced strong ground shaking in the Northridge earthquake,
excluding downtown Los Angeles and other parts of the city (FEMA, 2000b).
The LA ordinance was primarily targeted at detecting and repairing damage
that had occurred in the Northridge earthquake, as opposed to proactively
identifying and addressing buildings that may be at risk to future earthquakes.

Outside of Southern California, the Northridge earthquake damage prompted

investigations of some steel moment-resisting frame buildings in San
Francisco that had been subject to the 1989 Loma Prieta earthquake (FEMA,
2000b). Although the building code would normally have required repair of

1-24 Part 1: Inventory of Tall Buildings ATC 119-1

any Loma Prieta damage, the City did not retroactively require additional
investigation five years later, in part because the Loma Prieta ground motions
in downtown San Francisco were relative low (see Part 6 Section 6.2.2, for
additional discussion). Even where Loma Prieta damage did not occur or
was nominally repaired, the risk of damage in future earthquakes remains.
As illustrated in Figure 1-8b, the majority of ‘older steel buildings’ within the
tall building inventory were constructed prior to the Northridge earthquake.
Furthermore, typical construction details of these buildings, as observed
during the compilation of the database (and previous studies), are similar to
the beam-to-column connections that fractured in the Northridge earthquake.
There are also concerns about the integrity of welded column splices in these
systems.

When first introduced into construction practice, it was customary for steel
moment-resisting frame systems to be configured as “space frame systems,”
whereby all of the beams and columns in the building are engaged to resist
lateral loads by providing rigid moment-resisting connections at all beam-to-
column intersections. Motivated by improved construction economics and
architectural configurations, this practice later evolved to one where the
moment-resisting frames are concentrated in a subset of building frame. For
example, a common system is to concentrate the seismic-force-resisting
frames at the building perimeter of the building. Both the space frame and
perimeter frame configurations are observed within this building cohort with
roughly a 50-50 split between them.

While steel moment-resisting frames are clearly more vulnerable than

originally envisioned, the significance of the risk in these buildings is likely to
be highly variable, depending on the specific characteristics of the building. A
recent study by Molina Hutt et al. (2018) suggests that the older (pre-1994)
steel moment-resisting frames with fracture prone connections could have a
mean annual frequency of risk of collapse 25 times higher than new code-
conforming buildings and up to 2 times higher average annual risk of
economic loss and downtime. A currently ongoing study funded by the
National Institute of Standards and Technology (NIST) is further exploring
these issues to help inform policy decisions based on the risks posed by
existing tall steel buildings, including those associated with cordons around
damaged tall buildings.

3.3 Older Concrete Buildings

Damage and collapses observed to concrete buildings during the 1971 San
Fernando earthquake and other earthquakes have raised serious concern
about the safety of older non-ductile reinforced concrete buildings. In

ATC 119-1 Part 1: Inventory of Tall Buildings 1-25

 response to the San Fernando earthquake experience, major changes to
building codes were instituted in the late 1970s for seismic design of
reinforced concrete structures. Prior to these changes, seismic design
requirements did not require sufficient steel reinforcement or capacity design
provisions to resist story collapse mechanisms and other types of failure.
While it is generally recognized that older concrete buildings do not provide
the same level of safety and damage control as modern buildings, there are
continuing debates as to what policies should be adopted to implement
detailed risk assessment and mitigation (e.g., Liel and Deierlein 2013 and
http://www.concretecoalition.org).

In Southern California, cities such as Los Angeles, Santa Monica and West
Hollywood have recently mandated a retrofit policy for non-ductile reinforced
concrete frame buildings. In San Francisco, there are an estimated 3,300
concrete buildings constructed prior to 1980, which may pose high collapse
risks (http://www.concretecoalition.org).

As indicated in Figure 1-3b, older concrete buildings account for only about
10% of the buildings in the tall building database, the majority of which were
constructed in the 1960s and are in the lower (20-story) height range. Most
of these are either concrete shear wall or dual wall-frame systems. Further
work is needed to better understand the risk and develop appropriate policies
for this tall building cohort in San Francisco, as well as with the vast majority
of low and mid-rise older concrete buildings. New approaches identified by
FEMA can enable screening assessments to identify the most vulnerable
nonductile concrete buildings.

3.4 Modern Concrete Buildings

Many of the recently constructed (post-2000) tall buildings are residential

concrete shear wall systems, typically over 40 stories in height. These
comprise over 10% of the tall building inventory, a number that is likely to
continue to increase to address San Francisco’s housing needs. Roughly
half of this cohort is composed of shear wall only systems and half are dual
shear wall and moment-resisting frame systems. As discussed in Chapter 2
of this Part and noted in Figure 1-3b, many modern (post-2000) tall buildings
are engineered following performance-based seismic design guidelines. It is
estimated that about half of this cohort is designed using the performance-
based approach, which permits use of shear wall only systems. Otherwise,
for regular shaped buildings over 240 ft, the prescriptive building code
requirements would mandate a dual wall-frame configuration.

1-26 Part 1: Inventory of Tall Buildings ATC 119-1

This cohort is identified as one of special interest because of the prevalence
of these tall buildings for residential occupancies, which pose risks to San
Francisco’s residents that are quite different from office buildings. In
particular, while the modern high-rise residential towers are generally
considered to pose acceptably small life safety risks, damage from moderate
to large earthquakes may result in extended repair times and building
closures that displace residents. The Guidelines for Performance-based
Seismic Design of Tall Buildings (PEER, 2017) explicitly state that they are
“written with the intent that a building properly designed in accordance with
these Guidelines should be capable of achieving the seismic performance
objectives intended by ASCE 7”. In other words, the current design
requirements for tall buildings are intended to achieve parity with the
performance implied by the current San Francisco Building Code. As an
example of what might occur, the recent Haywired Scenario (USGS, 2018)
included an assessment of a modern 42-story concrete residential tower
located in Oakland. Under the magnitude-7.0 Hayward scenario earthquake,
the building is estimated to incur damage resulting in median times of 4
months for re-occupancy and 8 months for functional recovery. While the
risks do not carry over directly to buildings in San Francisco, the potential
clearly exists for large scale displacement of residents that can severely
impact the city.

Further study of this cohort will highlight the expected seismic performance of
newer tall buildings and cost implications associated with enhanced seismic
design performance objectives (see Part 3).

ATC 119-1 Part 1: Inventory of Tall Buildings 1-27

Chapter 4

Summary and Recommendations

In an effort to understand the impact that tall building performance may have
on the recovery of the City after a major earthquake, an inventory of the city’s
tall buildings was developed. The building inventory database is developed
and implemented in a Geographical Information Systems (GIS) format to
facilitate analysis and visualization of the tall building inventory and
identification of locations where building damage is most likely to have
disproportionate impacts on the urban community, and the broader socio-
economic factors that affect resilience (e.g., services provided to the
community by the affected buildings).

The database tabulates building characteristics by location, height,

occupancy, age of construction, construction material, structural system,
foundation type, façade system and other relevant design information, as
available (such as geometry and irregularities). An overview of the inventory
highlights the following statistics:
• The inventory includes 156 buildings with heights 240 ft or taller and 22
buildings with heights between 160 ft to 240 ft, either constructed or
permitted for construction in San Francisco.
• Buildings constructed in the 1960s, 1970s, and 1980s represent roughly
55% of the tall building inventory, and buildings constructed since 2000
represent about 25% of the inventory.
• The breakdown of structural materials and systems types in the tall
building inventory are as follows:

o 50% - steel moment resisting frame systems

o 10% - steel dual braced frame-moment frame systems

o 30% - reinforced concrete shear wall systems

o About 10% of the older buildings in the inventory did not have
sufficient documentation to identify a structural system
• 10% of the inventory is below 20 stories, 70% of the inventory is in the 20
to 40 story range, 20% is above 40 stories.

ATC 119-1 Part 1: Inventory of Tall Buildings 1-29

 • Approximately 60% of the buildings house commercial occupancies and
just under 40% are residential (multi-family housing) and hotel
occupancies.

Based on this review of the tall building inventory and considering the impact
of the evolution of seismic design, past earthquake observations, and other
considerations, the following tall building cohorts of special interest are
identified as worthy of further study:
• Older Steel Buildings: Steel-framed buildings constructed in the 1960s
through the early 1990s have many known deficiencies, relative to current
seismic building codes, including welded connection details similar to
those that fractured in buildings subjected to strong shaking during the
1994 Northridge Earthquake. The vulnerability of these buildings
warrants further study of the risks they pose to their occupants and
neighboring buildings.
• Older Concrete Buildings: Concrete buildings constructed prior to 1980
are likely to have non-ductile detailing and other deficiencies that resulted
in building collapses during the 1971 San Fernando Earthquake.
Although the number of older, tall concrete buildings is small relative to
those with other structural systems, the vulnerability of these systems
poses a significant risk. Vulnerable older concrete buildings represent a
much larger percentage of low- and mid-rise buildings and are well
recognized to pose a significant risk. This has prompted mandatory
assessment and retrofit requirements in other jurisdictions in California,
which should be evaluated for implementation in San Francisco.
• Modern Concrete Buildings: Newer concrete shear wall buildings that
meet modern (post-2000) building code requirements are expected to
have low life-safety risks from falling hazards and collapse. However,
modern building codes do not provide minimum requirements for
controlling earthquake damage that may require extensive repair with
extended downtime. Given the emergence of modern high-rise
residential buildings in San Francisco, it is recommended that the
implications of adopting seismic performance objectives beyond the code-
minimum be evaluated.

The appendix to this report summarizes the metadata fields of the digital GIS
database. To help maintain and expand the database, it is recommended to
develop a checklist of information that could be collected from building permit
applications for new and/or retrofitted buildings, façade inspection reports,
and other similar programs managed by the City of San Francisco.

30 Part 1: Inventory of Tall Buildings ATC 119-1

Given the relatively small number of older concrete buildings in the tall
building database (6 concrete and 7 mixed concrete-steel older buildings in
the 17 to 25 story range), the non-ductile concrete issue is less driven by tall
building concerns as compared to the overall high risk posed to the large
number (estimated 3,300) of low- to mid-rise older concrete buildings in San
Francisco. The modern (post-2000) residential high-rise buildings are
generally considered to be safe. The concern with these relates to their
potential risk to damage and extended downtime, causing displacement of
residents, in a moderate to severe earthquake.

The continued maintenance and expansion of this database is considered to

be an essential resource for San Francisco to assess and manage risks to
the large building stock that houses residents and businesses of the city.

ATC 119-1 Part 1: Inventory of Tall Buildings 1-31

Appendix A

Tall Building Inventory Data Fields,

Definitions, and Notes
This appendix lists all variables included within the tall buildings database.
For all basic attributes, such as date, height, number of stories, occupancy, if
permit drawings located at the San Francisco Department Building Inspection
(DBI) or Building Occupancy Resumption Program (BORP) reports were
unavailable or insufficient, the data were taken from the Emporis building
information database (https://www.emporis.com/). Sources of data for the
various data fields in Table 1-3 are indicated through the footnotes to the field
descriptions in the table.

Table 1-3 Tall Building Inventory Data Fields

Field Field Type Descriptions
OBJECTID Numeric Object ID
Shape Geometry: Polygon Footprint of the building at ground leveli
Name Text Name of the building
Address Text Street address of the building
MapBlockLot Text The block and lot number of the representative parcel for the buildingii
MBL_Unique Text The MapBlockLot (MBL) number, modified with a cardinal direction if
multiple buildings share the same MBL
Date Numeric Building's year of constructioniii
Retrofit_Date Numeric Year the building was seismically retrofit, if knowniv
Description Text Description of basic building characteristics
Height_ft Numeric Height of the building, in feetv
Stories_Above_Grade Numeric Number of stories above ground levelv
Stories_Below_Grade Numeric Number of stories below ground levelv
Occupancy Text Occupancy type of building, consistent with San Francisco's Land Use
categorizationvi
Structural_Material Text Structural material of the building (steel, reinforced concrete, or mixed)v
Structural_System Text Lateral Force Resisting System (LFRS) of the buildingvii
Structural_Types Text Broad categorization of the LFRS (steel moment frame, steel braced
frame, reinforced concrete shear wall)
Façade_Material Text Façade material, classified by visual identification
Foundation_System Text Type of foundation system usedvii
BORP_Report Text Is the building part of BORP?

ATC 119-1 Part 1: Inventory of Tall Buildings 1-33

Table 1-3 Tall Building Inventory Data Fields (continued)
Field Field Type Descriptions
Instrumented Text Is the building seismically instrumented to record earthquake motion? viii
Square_Footage Numeric Square footage of the buildingix
Fire_Resistence_Type Text Fire resistance of building materialsix
Percent_Sprinklered Numeric Percentage of building with sprinklers installedix
Liquefaction_Potential Text Classification of the liquefaction potential at the building's centroidx
BedrockDepth_MEAN Numeric Mean depth to bedrock over the building's footprintxi
BedrockDepth_MAX Numeric Maximum depth to bedrock over the building's footprintxi
BedrockDepth_MIN Numeric Minimum depth to bedrock over the building's footprintxi
Permit_Date_1 Numeric Year of any date stamps found on the permit documents
Permit_Date_2 Numeric Year of any date stamps found on the permit documents
Permit_Date_3 Numeric Year of any date stamps found on the permit documents
Completion_Date Numeric Year of building completionxii
Building_Code_Year Numeric The date of the applicable design code, if noted in permit documentation
Base_Plan_Size Text Dimensions of the plan at ground levelxiii
Tower_Plan_Size Text Dimensions of the plan for the towerxiii
Typ_Story_Height Text Height of typical storyxiii
Attyp_Story_Height Text Any unique story heightsxiii
Architectural_Notes Text Relevant architectural features, such as setbacks or plan shapexiii
Atrium_Location Text Story on which an atrium is locatedxiii
MEP_levels Text Story on which mechanical, electrical, or plumbing (MEP) equipment is
locatedxiii
Site_Class Text Site class, if noted in permit documentation
Foundation_Info Text Additional description of the foundationvii
Collected_on Text Date the data was collected
Primary_Source Text Primary source the data was collected from (BORP reports, DBI permits,
Emporis, or Peer Reviewers)
Shape_Length Numeric Footprint geometry info (in feet)
Shape_Area Numeric Footprint geometry info (in sqft)
i Extracted from DataSF Building Footprints or hand-drawn for newer buildings
ii Identified based on SF Property Information Map
iii In order of preference: certificate of final completion, latest permit date, or Emporis
iv Discussions with Engineer of Record or design peer reviewers
v Permit application drawings or Emporis (Emporis’s height may be measured directly or calculated as ~12.5ft × number of stories)
vi Informed by “main usage” field in Emporis
vii Only from permit applications or discussions with engineers/peer reviewers (not Emporis)
viii Inferred from locations on the Strong Motion Center map (https://www.strongmotioncenter.org/LoaderNC.html). Anecdotal evidence suggests more
buildings are instrumented but how/where they are managed is unknown.
ix From SF Fire Department records (February 2018). Not all buildings have a match.
x From USGS maps (https://pubs.usgs.gov/of/2000/of00-444/)
xi Based on 1961 USGS bedrock map and current day topographic data (Part 1, Appendix C)
xii Certificate of Final Completion (CFC) or Emporis (only if no permit dates are available)
xiii Only from permit applications

1-34 Part 1: Inventory of Tall Buildings ATC 119-1

Appendix B

Building Permit Supplement for

Inventory Maintenance

In order to maintain and expand the tall building inventory, it is recommended

that the form shown in Figure 1-9 is used as a starting point for the
development of a supplemental data collection form required in the process
of permit applications. The fields in Figure 1-9 are based directly on the main
attributes of the digital database prepared in this study and are identified in
Figure 1-10. Further development of the supplemental form should consider
coordination with other data sets and feasibility of implementation and
maintenance.

In addition to the list of information items, it is recommended to request from

the building owner/design team, an electronic submission of a .shp, or .kml
file that contains the building footprint. This information can be incorporated
in the database to facilitate visualization of the building location and
geometry.

ATC 119-1 Part 1: Inventory of Tall Buildings 1-35

Figure 1-9 Proposed permit supplement form (page 1).

1-36 Part 1: Inventory of Tall Buildings ATC-119-1

Figure 1-10 Proposed permit supplement form (page 2).

ATC 119-1 Part 1: Inventory of Tall Buildings 1-37

Appendix C

Reference for Bedrock Depth Data

The tall building inventory includes three fields related to the depth of the
bedrock beneath the buildings: the average, maximum, and minimum depth
across the footprint. These values were calculated based on two sources:
one for the elevation at ground level and one for the elevation at bedrock.

The elevation at ground level is provided as a raster dataset (a single

elevation value for every square in a grid across the location of interest,
similar to pixels in an image) by the U.S. Geological Survey (USGS). The
dataset was published in 2000 and can be downloaded at https://pubs.usgs
.gov/of/2000/of00-444/.

The elevation at bedrock also comes from the USGS in the form of a contour
map from 1961 (Figure 1-11). The contours were traced and spatially
interpolated to obtain a smooth surface for comparison with the ground level
data. Using the footprint location of the buildings as a template, the bedrock
elevation was then subtracted from the ground level elevation for the
average, maximum, and minimum depth to bedrock for each building.

ATC 119-1 Part 1: Inventory of Tall Buildings 1-39

Figure 1-11 USGS map (1961) used for determining the depth to bedrock.

ATC 119-1 Part 1: Inventory of Tall Buildings 1-40

Appendix D

References

Almufti, I., Molina Hutt, C., Mieler, M.W., Paul, N.A., and Fusco, C.R., 2018,
The HayWired Earthquake Scenario - Engineering Implications: Case
Studies of Tall-Building Structural Analyses, Downtime and Loss
Assessment for the HayWired Scenario Mainshock, United States
Geological Survey, Menlo Park, California.
ASCE, 2017, Minimum Design Loads and Associated Criteria for Buildings
and Other Structures, ASCE/SEI 7-16, American Society of Civil
Engineers Structural Engineering Institute, Reston, Virginia.
CBSC, 2016, 2016 California Building Code, California Code of Regulations,
Title 24, California Building Standards Commission, Sacramento,
California.
CCSF, 1995, Downtown Area Plan, Planning Department, the City and
County of San Francisco, California.
CCSF, 2008, Requirements and Guidelines for the Seismic Design of Tall
Buildings Using Non-Prescriptive Seismic-Design Procedures,
Administrative Bulletin 083, Department of Building Inspection, the
City and County of San Francisco, California.
CCSF, 2011, CAPSS Earthquake Safety Implementation Program Workplan
2012-2042, the City and County of San Francisco, California.
CCSF, 2014, Procedures for Seismic Instrumentation of New Buildings,
Administrative Bulletin 058, Department of Building Inspection, the
City and County of San Francisco, California.
CCSF, 2016, The San Francisco Building Code, the City and County of San
Francisco, California.
CCSF, 2017a, San Francisco Property Information Map, Planning
Department, the City and County of San Francisco, California.
Available at: http://propertymap.sfplanning.org/, last accessed: August
2018.
CCSF, 2017b, Building Façade Inspection and Maintenance, Administrative
Bulletin 110, Department of Building Inspection, the City and County
of San Francisco, California.

ATC 119-1 Part 1: Inventory of Tall Buildings 1-41

 Cornell A., 1968, “Engineering seismic risk analysis,” Bulletin of the
Seismological Society of America, Vol. 58, No. 5, pp. 1583-1606.
FEMA, 2000a, State of the Art Report on Systems Performance of Steel
Moment Frames Subject to Earthquake Ground Shaking, FEMA
355-C, prepared by the Structural Engineers Association of California,
Applied Technology Council, and Consortium of Universities for
Research in Earthquake Engineering (SAC) Joint Venture for the
Federal Emergency Management Agency, Washington, D.C.
FEMA, 2000b, State of the Art Report on Past Performance of Steel Moment-
Frame Buildings in Earthquakes, FEMA 355-E, prepared by the
Structural Engineers Association of California, Applied Technology
Council, and Consortium of Universities for Research in Earthquake
Engineering (SAC) Joint Venture for the Federal Emergency
Management Agency, Washington, D.C.
FEMA, 2006, Next-generation Performance-based Seismic Design
Guidelines, FEMA 445, prepared by the Applied Technology Council
for the Federal Emergency Management Agency, Washington, D.C.
LATBSDC, 2017, An Alternative Procedure for Seismic Analysis and Design
of Tall Buildings Located in the Los Angeles Region, Los Angeles Tall
Buildings Structural Design Council, Los Angeles, California.
Liel, A., and Deierlein, G.G., 2013, “Cost-benefit evaluation of seismic
mitigation alternatives for older reinforced concrete frame buildings,”
Earthquake Spectra, Vol. 29, No. 4, pp. 1391-1411.
Molina Hutt, C., Rossetto, T., and Deierlein, G.G., 2018, “Comparative risk-
based seismic performance assessment of 1970s vs modern tall steel
moment resisting frames,” manuscript submitted for publication.
PEER, 2010, Guidelines for Performance-based Seismic Design of Tall
Buildings, PEER Report 2010/05, Pacific Earthquake Engineering
Research Center, University of California, Berkeley, California.
PEER, 2017, Guidelines for Performance-based Seismic Design of Tall
Buildings, PEER Report 2017/06, Pacific Earthquake Engineering
Research Center, University of California, Berkeley, California.
SEAOC, 1959, Recommended Lateral Force Requirements and
Commentary, Seismology Committee, Structural Engineers
Association of California, Sacramento, California.
USGS, 2018, The HayWired Earthquake Scenario - Engineering Implications,
USGS Scientific Investigations Report 2017-5013-I-Q, US Geological
Survey, Menlo Park, California.

1-42 Part 1: Inventory of Tall Buildings ATC 119-1

 PART 2:
Geotechnical Engineering
for Tall Buildings
Chapter 1

Introduction

1.1 Background

In San Francisco, geotechnical engineering and foundation design varies

considerably among practitioners because best practices are not yet fully
codified. This Part summarizes the state-of-the art and state-of practice for
geotechnical engineering for tall buildings and corresponds to
Recommendation 1A presented in Summary Recommendations.

The purpose of this Part is to put forward recommendations that can be

established as guidelines to bring consistency and uniformity to geotechnical
practice for foundation design. Accordingly, this Part presents the following
information for each type of foundation considered for support of a tall
building:
• What are the key technical issues?
• How should key technical issues be addressed?
• What are the reference documents?
• What is the state-of-practice and what is the state-of-the-art?
• Considering the best practices at the national and international levels,
what are the shortcomings of the state-of-practice in our region and how
can it be advanced to the state-of-the-art?

1.2 Intended Audience and Use of this Report

This Part summarizes the state-of-the art and state-of practice for
geotechnical engineering for tall buildings based on experience in San
Francisco and cities with tall buildings built on similar soils (Chicago and
Boston), as well as guidance from New Zealand. The material was
developed for review by persons with advanced geotechnical engineering
knowledge and understanding. References are provided within each section
to allow for further review of technical details.

Administrative Bulletin (AB) 083, Requirements and Guidelines for the

Seismic Design of New Tall Buildings using Non-Prescriptive Seismic Design
Procedures, (CCSF, 2014) sets requirements for structural design of tall
buildings, and it is recommended that the recommendations of this report be

ATC-119-1 Part 2: Geotechnical Engineering for Tall Buildings 2-1

 implemented in a similar AB for geotechnical design following a thorough
review of relevant sections by local practitioners.

In addition, to strengthen the Department of Building Inspection’s (DBI)

procedures for assessing the completeness of the foundation and excavation
design for tall buildings, two additional actions are recommended:
• Increase DBI’s expertise on geotechnical issues related to tall buildings
through enhanced training and staffing.
• Develop a geotechnical report checklist to help ensure that submitted
geotechnical investigation, design, and field monitoring reports are
complete.

Geotechnical design of tall buildings was not covered in the ESIP Workplan
(CCSF, 2011).

1.3 San Francisco Conditions

Figure 2-1 (developed as part of the inventory described in Part 1) illustrates

the types of foundations for tall buildings in downtown San Francisco,
superimposed on a map that identifies regions that are susceptible to soil
liquefaction during strong earthquakes. Figure 2-2 shows a typical soil profile
illustrating the varying soil layers.

Figure 2-1 Map showing foundation types of tall buildings superimposed where beige shading
identifies regions that are susceptible to soil liquefaction during strong earthquakes (soil
information from https://datasf.org/).

2-2 Part 2: Geotechnical Engineering for Tall Buildings ATC-119-1

Figure 2-2 Typical soil profile for a downtown San Francisco site.

ATC-119-1 Part 2: Geotechnical Engineering for Tall Buildings 2-3

 Foundation types of tall buildings vary due to the following factors:
• Depth to rock: This may range from 0 ft to more than 250 ft. Figure 2-2
shows a typical soil profile from a downtown site with the Franciscan
bedrock formation located at 250 ft depth.
• Soil type and stiffness: The stiffness of the soil may vary depending on
the site’s soil profile, as illustrated in Figure 2-2.
• Building height and weight
• Number of basement levels
• Slope of site
• Proximity to shoreline
• Adjacent and underlying structures
• Ground water level
• Available construction technologies, logistics, and economics

1.4 Organization

Chapter 2 presents a discussion on acceptable limits of total and differential

settlement for tall buildings.

Chapters 3 and 4 present discussions on geotechnical engineering for new

tall buildings on deep foundations and shallow foundations, respectively.

Chapter 5 presents a discussion on evaluation and retrofit of foundations of

existing tall buildings.

Chapter 6 presents a discussion on foundation design of tall buildings near

shoreline.

Chapter 7 presents a discussion on shoring and dewatering considerations.

Chapter 8 presents a discussion on sea level rise and its effects on

foundation design.

Chapter 9 presents summary recommendations for evaluating the adequacy

of geotechnical reports.

Appendix A presents recommendations regarding settlement considerations

based on case histories.

References and full citations are incorporated into each chapter for ease of
use.

2-4 Part 2: Geotechnical Engineering for Tall Buildings ATC-119-1

Chapter 2

Total and Differential Settlement

The current California Building Code does not provide criteria for maximum
allowable limits for total and differential settlement of foundations supporting
commercial or residential tall buildings. However, many practitioners and
officials with building departments of various cities consider acceptable a total
settlement of less than 4 inches and a differential settlement limited to an
angular distortion of 1/500 for composite steel or concrete core buildings and
1/750 for reinforced concrete towers (rigid structures) acceptable. The
structural system, cladding system, partitions, and other nonstructural
components may also control the acceptable differential settlement.

It is noted that some buildings, such as base-isolated structures, are more

sensitive to differential settlement than other conventional buildings and
would require a tighter limit on differential settlement. However, base-
isolation systems are seldom used in tall buildings.

Building settlement should be monitored during construction and to at least

10 years after construction. Monitoring and reporting requirements should
follow guidelines presented on Information Sheet S-18, Interim Guidelines
and Procedures for Structural, Geotechnical, and Seismic Hazard
Engineering Design Review for New Tall Buildings, issued by the San
Francisco Department of Building Inspection (SFDBI) on December 2017.

ATC-119-1 Part 2: Geotechnical Engineering for Tall Buildings 2-5

Chapter 3

New Tall Buildings on Deep

Foundations

When ground conditions at a site are not suitable for shallow spread or raft
foundation systems, especially for high-rise buildings where the vertical and
lateral loads imposed on the foundation are significant, it is necessary to
support the building on deep foundation or piles. Drilled shaft, driven pile, or
auger cast pile foundations can be used either as a single unit or in groups
and are generally located beneath columns and load bearing walls.

Deep foundations provide support through skin friction, end bearing, or

combination of skin friction, and end bearing. In the latter case, displacement
compatibility between mobilized ultimate capacity in friction and end bearing
should carefully be evaluated.

Due to the presence of liquefiable fill and soft Bay mud, the majority of the
existing tall buildings in downtown San Francisco are supported either on
drilled shafts extending to Franciscan bedrock formation or on driven 12-inch
or 14-inch precast, prestressed concrete piles gaining support through end
bearing within the dense Colma sand layer below Young Bay mud.

3.1 Deep Foundation Types

Most common types of deep foundation used in the San Francisco Bay Area
for support of tall buildings are drilled shafts, driven concrete or steel piles,
and auger cast piles. An alternative deep foundation used frequently in
Europe is Barrettes, or known on the east coast of the United States as LBEs
(Load Bearing Elements, which is an isolated slurry vertical diaphragm with
rectangular elements 3 feet to 4 feet wide by 9 feet to 10 feet long. The
application of this foundation in the western United States has been very
limited with only one tall building in San Francisco supported on Barrettes.
However, on the east coast, they have been used on many tall buildings
since the late 1980s.

3.1.1 Drilled Shafts

Diameter of drilled shafts or caissons commonly ranges from 30 inches to 72

inches; however, larger diameter drilled shafts have also been successfully

ATC-119-1 Part 2: Geotechnical Engineering for Tall Buildings 2-7

 constructed and tested in the San Francisco Bay Area (e.g., 300 feet deep
drilled shafts with a diameter as large as 10 feet were constructed and
successfully tested for support of the high viaduct at Presidio Parkway in San
Francisco in early 2012).

Depending on the local soil conditions, drilling fluid (slurry) or steel casing is
used to stabilize the drilled hole before placement of rebar cage and tremie
placed structural concrete. High capacity drilled shafts (axial capacity in the
range of 20,000 kips) with rock socket into Franciscan bedrock formation
have been successfully installed and tested in the recent past.

Due to the high groundwater table in the San Francisco Bay Area, drilled
shafts are commonly constructed using the slurry method, and concrete is
placed using the tremie method for concrete mixes having a slump of 7 to 9
inches. Tremie pipes should be at least 10 inches in diameter for pumped
and gravity-fed methods, respectively. If used for stability of the drilled hole,
to avoid contamination of concrete through caving soils, the temporary casing
is pulled out with concrete surface being several feet above the bottom of
casing at all times.

It is not practical to perform pile load test on high capacity (large diameter
and very long) drilled shafts using conventional pile load test conducted from
ground surface. In these cases, pile load tests are performed by placing and
activating hydraulic jacks placed near the bottom of drilled shafts. Osterberg
(O-Cell) or equivalent testing devices have successfully been used to
measure ultimate skin friction and end bearing of high capacity drilled shafts
in soil and rock.

Ultimate skin friction capacity obtained through O-Cell testing in bedrock of

Franciscan bedrock formation ranges from 4 kips per square feet (ksf) to 20
ksf. Ultimate end bearing in bedrock of Franciscan bedrock formation greatly
varies, depending on the contractor’s procedure and the effort in cleaning of
the bottom of the shaft prior to placement of concrete. It varies from almost
zero (a recent test at a project site in downtown San Francisco has
encountered no end bearing resistance for vertical shaft movement at the tip
of up to 2 inches) to ultimate end bearing of 100 ksf and greater.

There are various methods for cleaning the bottom of drilled shafts. These
include the use of cleanout bucket or air lifting. Inspection of high capacity
drilled shaft is performed by various methods including measurement of
drilled shaft height immediately before placement of concrete using a
weighted tape measure (least reliable method), a Mini-SID (Shaft Inspection
Device), or DID (Ding Inspection Device). For smaller diameter drilled shafts

2-8 Part 2: Geotechnical Engineering for Tall Buildings ATC-119-1

(3 feet or less) where the bottom of the shaft could not be inspected / tested,
end bearing capacity is normally ignored. Shaft verticality is checked with
automated monitoring equipment on the drill rig by using Kelly bar or cutting
auger, downhole sonic caliper, or stringed plumb bobs from a centered hole
frame.

For drilled shafts socketed in bedrock, the minimum rock socket length
should generally be 15 feet or twice the shaft diameter, whichever is greater.
The minimum rock socket length requirement could be relaxed for sites
where the depth to bedrock is greater than 25 times the shaft diameter and
hard rock with substantial lateral extension and depth (i.e., the hard rock
boulder is the size of a house and not a compact car) is encountered below
the bottom of drilled shafts.

Typically, for sound rock, 50% of axial capacity of rock socket is due to skin
friction. For softer rocks, the contribution of skin friction increases to 75%.

Until concrete has gained sufficient strength (i.e., set for 48 to 72 hours),
newly installed drilled shafts should be protected against construction related
vibration (e.g., pile driving) or installation of other drilled shafts within six shaft
diameters or 25 feet, whichever is less. The above criteria should be revised
if admixtures (retarders) are used.

The most relevant references are the following:

• FHWA, 2010, Drilled Shafts: Construction Procedures and LRFD Design
Methods, FHWA Publication No. FHWA-NHI-10-016, U.S. Department of
Transportation Federal Highway Administration.
• For installation guidelines: ACI, 2014, Report on Design and Construction
of Drilled Piers, report by by ACI Committee 336, American Concrete
Institute, Farmington Hills, Michigan.

3.1.2 Driven Concrete or Steel Piles

When precast, prestressed concrete piles are used, the typical allowable
axial capacity of 12-inch-square and 14-inch-square piles driven to refusal
within Colma sand layer are 100 and 150 tonnes, respectively.

For sites where either Colma sand does not exist, the layer is not thick
enough to prevent punching through while pile driving, or the Franciscan
bedrock formation is within 100 to 200 feet below ground surface, steel H
piles with driving shoes are used. The spatial variation of rock quality and
variation in bedrock surface elevation makes steel H piles an attractive
alternative as the length of pile could be increased through field welding or
adjusted by cutting extra length of pile if refusal occurs before the pile is fully

ATC-119-1 Part 2: Geotechnical Engineering for Tall Buildings 2-9

 driven to the design cut-off evaluation. If steel pile splices are required, they
should be full penetration welds with no flange or web plates allowed in
seismic regions.

The main disadvantages of driven concrete piles are human factors, such as
objection to noise and vibration that are problematic in congested and well-
developed areas of the City. In addition, due to presence of debris in the
surficial fill layer, predrilling of pile locations would normally be required. The
main disadvantage of steel pipe or H piles is the potential for corrosion.

The axial capacity of steel piles driven to refusal within Franciscan bedrock
formation is controlled by structural capacity of the pile; however, the ultimate
capacity may be controlled by buckling failure mode if soft Young Bay mud or
liquefiable soils are present.

There are three alternatives for addressing corrosion issues related to steel
piles. These include the following:
• Adding a sacrificial thickness to the pile section (normally 1/8-inch or
1/16-inch, depending on soil corrosivity). This option is the least reliable
as reduction in thickness of pile may not occur uniformly along the pile as
assumed.
• Applying coating material (factory applied epoxy coating) to the top
portion of piles to a depth where there is no potential for oxygen to reach
the pile (this depth is normally taken at 10 feet below the surface of
Young Bay mud). This option is more reliable than adding sacrificial
thickness to pile, but the coating may be damaged during pile driving and
if a section of coating is removed during pile driving resulting in scratches
on the pile, very weak spots vulnerable to corrosion are created.
• Applying a Cathodic protection system. This is the most reliable system
to address corrosion issues related to steel piles, but is also a costly
option and that requires long term maintenance.

The standard of practice is summarized in the following reference:

• FHWA, 2016, Design and Construction of Driven Pile Foundations,
FHWA Publication No. FHWA-NHI-16-009, U.S. Department of
Transportation Federal Highway Administration.

3.1.3 Auger Cast Piles

Due mainly to their cost effectiveness, auger cast piles are gaining popularity
in recent decades. Typical diameter of auger cast piles range from 18 inches

2-10 Part 2: Geotechnical Engineering for Tall Buildings ATC-119-1

to 48 inches. Auger cast piles have successfully been installed to a depth of
200 feet.

The most suitable soil classification for installation of auger cast piles is
medium stiff to stiff clays and medium dense to dense sands. The most
challenging soil classification is loose sand (especially below ground water
table) and soft clays, as caving of soil could result in necking in the pile
section.

In addition, installation of long and small diameter auger cast piles in soft
clays requires checking for buckling mode of failure under static and seismic
loading conditions.

Until concrete has gained sufficient strength, newly installed auger cast piles
should be protected against construction related vibration (e.g., pile driving)
or installation of other auger cast piles within six shaft diameters (see
discussion on this topic presented in Section 3.1 under heading of “Drilled
Shafts”).

The most relevant reference for design of auger cast piles is given in the
following reference:
• FHWA, 2007, Design and Construction of Continuous Flight Auger Piles,
FHWA Publication No. FHWA-HIF-07-03, U.S. Department of
Transportation Federal Highway Administration.

3.1.4 Other Deep Foundation Types

Specialty constructed piles include torque-down piles, Tubex piles, Franki

piles, micropiles, and Fundex piles. Due to the limited axial capacity provided
by these piles (maximum allowable axial capacity is normally limited to about
600 kips or slightly higher with structural design calculations and load
testing), their use is not common for supporting new tall buildings, which have
a high demand on axial pile capacity.

These specialty piles are designed and constructed by specialty design-build

(DB) contractors using proprietary tools and techniques. In addition to
developing foundation performance criteria and reviewing plans and
specification for foundation construction and testing, the geotechnical
engineer of record (GEOR) should ask for and review the design submittal by
the DB contractor, as the GEOR is ultimately responsible for geotechnical
aspects of foundation design.

ATC-119-1 Part 2: Geotechnical Engineering for Tall Buildings 2-11

 3.2 Piles Going Through Soft or Liquefiable Soils and Firm Soil
Interface

Lateral reinforcing of concrete piles normally extends down below the point of
fixity (about 12 pile diameters below the pile top). However, in some cases,
lateral reinforcement should continue well below the interface between soft or
liquefiable soils and the underlying firm, competent soils. This is because
large shear strains are developed at this interface as seismic waves
propagate from competent soils at depth to the ground surface.

For example, if subsurface conditions at a site consist of 35 feet of liquefiable

sand and soft Young Bay mud underlain by alluvium consisting of dense
sands or stiff clays, lateral reinforcement of a 24-inch drilled shaft should
extend down 40 feet below the top of the shaft as opposed to stopping below
point of fixity at a distance of about 25 feet below the top of the shaft.

3.3 Downdrag Loads on Piles

Often, sites are raised by placing new fill over existing ground surface. The
added load would cause long term settlement within saturated clayey soils,
imposing downdrag loads on piles. Downdrag loads applied to a pile group
depends on pile geometry and layout, undrained shear strength of soil,
effective vertical stress, and soil plasticity. General information regarding
downdrag loads on piles are presented in the following publications:
• Kuwabara, F., Poulos, H.G., 1989, “Downdrag forces in group of piles,”
Journal of Geotechnical Engineering, Volume 115, No. 6.
• Lee, C.J., and Ng, C.W., 2004, “Development of downdrag on piles and
pile groups in consolidating soil,” Journal of Geotechnical and
Geoenvironmental Engineering, Volume 130, No. 9.

Useful information regarding case histories and methods for evaluation and
mitigation of downdrag loads are presented in the following reference:
• Fellenius, B.H., 1998, “Recent advances in the design of piles for axial
loads, dragloads, downdrag, and settlement,” ASCE and Port of NY&NJ
Seminar, April 22-23, 1998.

In case of San Francisco Young Bay mud, downdrag loads are computed by
multiplying vertical effective stresses by a factor between 0.25 to 0.35 for a
single pile and by a factor between 0.1 to 0.2 for a typical pile group (3 by 3, 4
by 4, or more). The corresponding range of values for sandy fill are 0.3 to
0.55 for a single pile and 0.20 to 0.35 for a typical pile group (3 by 3, 4 by 4,
or more). The smaller values are used for drilled shafts and auger cast piles
and the larger values are used for driven piles.

2-12 Part 2: Geotechnical Engineering for Tall Buildings ATC-119-1

To arrive at allowable axial pile capacity, downdrag load is subtracted from
ultimate axial pile capacity before applying a factor of safety (as opposed to
dividing the ultimate axial capacity by a factor of safety and then subtracting
the downdrag load).

3.4 Integrity Testing of Deep Foundations

Integrity testing of high capacity drilled shafts are performed through cross-
hole sonic, gamma-gamma, or thermal tests. Smaller diameter drilled shafts
could be tested through conventional static or dynamic pile load test
conducted from the ground surface.

To check the potential for necking of auger cast piles, field observation of
installation should be performed using a fully automated data acquisition
system. In addition, pile integrity tests should be performed for 5 to 10
percent of production piles and for piles with questionable recorded
installation data. Pile integrity tests could be performed through static or
dynamic pile load tests, cross-hole sonic, gamma-gamma, or thermal tests.
Cross-hole sonic and gamma-gamma tests cannot be performed for piles
with a diameter less than 24 inches (i.e., for 18-inch diameter auger cast
piles, only thermal test could be performed). It is noted that performing cross-
hole sonic, gamma-gamma, and thermal tests would require attachment of
the test instrument to the rebar cage before the cage is inserted into the fully
grouted drilled hole.

Other dynamic tests for confirmation of compression load capacity of

production or suspect deep foundations should be performed in accordance
with ASTM D49451 using proprietary software Statnamic 2 or GRL Apple 3.
The test could be performed in a stepped manner and up to twice the static
force acting on the deep foundation element. The CAPWAP 4 analysis is
often performed using the GRL Apple system that involves a falling weight
striking the head of piles or shafts. A concrete filled steel casing is placed
above the top of the pile to prevent damage caused by impact loading.

3.5 Allowable Tolerance for Installation of Deep Foundations

Deep foundations should meet plumbness criteria. Typical specifications

require a tolerance of less than 1 inch for every 10 feet of depth (1 in 75 to 1
in 100 ratios have also been specified in other practice areas). In addition,
1 ASTM, 2018, Standard Test Method for Shrinkage Factors of Cohesive Soils by the
Water Submersion Method, ASTM International, West Conshohocken,
Pennsylvania.
2 Berminghammer, Canada.
3 Load testing system, GRL Engineers, Inc., Cleveland, Ohio.
4 Case Pile Wave Analysis Program, Pile Dynamics, Inc., Cleveland, Ohio

ATC-119-1 Part 2: Geotechnical Engineering for Tall Buildings 2-13

 the center of deep foundation is normally specified to be within an inch from
center of a column and 3 inches from the design location.

3.6 Pile Load Test Procedure and Interpretation of Results

Axial pile compression load tests are performed in accordance with the
following ASTM standards:
• ASTM, 2013a, Standard Test Methods for Deep Foundations Under
Static Axial Compressive Load, ASTM International, West
Conshohocken, Pennsylvania.
• ASTM, 2017, Standard Test Method for High-Strain Dynamic Testing of
Deep Foundations, ASTM International, West Conshohocken,
Pennsylvania.

Lateral pile load tests are performed less frequently in practice than axial pile
load test for design of new foundations. Instead, design engineers rely on
interpretation of lateral pile load test results and development of p-y springs
by others as presented in the literature. The following ASTM standard
corresponds to lateral pile load tests of deep foundations:
• ASTM, 2013b, Standard Test Methods for Deep Foundations Under
Lateral Load, ASTM International, West Conshohocken, Pennsylvania.

In the San Francisco Bay Area, where liquefiable fill and Young Bay mud are
present, the contribution of the liquefiable sand layer and Young Bay mud to
axial pile capacity should be carefully evaluated and subtracted from ultimate
pile capacity obtained through conventional pile load tests conducted from
ground surface. In addition, down drag loads should be subtracted from
ultimate axial pile capacity (see discussion related to downdrag loads in
Section 3.3).

Comparison on the same piles show that the ultimate axial pile capacity
obtained from a dynamic load test in clayey soils is normally 1.3 times that
obtained from a static load test. This means that strain rate effects increase
the measured axial capacity of piles by 30 percent in clayey soils present in
the San Francisco Bay Area. However, the ultimate axial capacity of piles in
cohesionless soils are not affected by strain rate effects. Similar results are
obtained from static and dynamic test conducted in sandy soils.

3.7 Soil Liquefaction-Related Design Issues

Presence of liquefiable soils within saturated loose to medium dense sand

within surficial fill layer or sand lenses within Young Bay mud would impact

2-14 Part 2: Geotechnical Engineering for Tall Buildings ATC-119-1

design of deep foundations. The standard of practice for evaluation of soil
liquefaction potential is based on the following reference:
• Idriss, I.M, and Boulanger, R.W., 2008, Soil Liquefaction During
Earthquakes, EERI Monograph, Earthquake Engineering Research
Institute, Oakland, California.

The procedures and recommended methods for evaluation of soil liquefaction

potential presented in the above references are very similar for loose to
medium dense sands but are noticeably different for denser sands (with N1(60)
between 20 and 30, indicating SPT blow count normalized to 1 tsf (tonne/sqft)
confining pressure with 60% free fall energy of a 140 pound hammer
dropping 20 and 30 inches) at high cyclic stress ratios. Most practitioners
consider the difference between the methods as “epistemic uncertainty,” and
both methods are used to provide a range of solutions.

Most practitioners would agree that the effects of soil liquefaction (defined as
pore water pressure ratio of 100 percent) does not need to be considered for
sand with N1(60) greater than 25. This is because sands at this density are
dilative and therefore, soil has limited potential for developing large shear
strains or undergo appreciable loss of shear strength. Accordingly, the
effects of soil liquefaction could be ignored for sands with N1(60) of 25 or
greater.

Issues related to soil liquefaction are discussed below:

3.7.1 Temporary Loss of Axial Support

Soil liquefaction would result in temporary loss of axial support within the
liquefied soil layer. Therefore, no axial support in friction is assigned to this
layer.

3.7.2 Reduction of Lateral Support

The lateral support of piles is reduced as a result of soil liquefaction. The

standard of practice is to develop p-y springs (from L-PILE5 analysis) prior to
liquefaction (static condition) and reduce the static p-y curve by a factor
between 5 and 10, i.e., post liquefaction p-y being 10 % to 20% of static
values.

Other practitioners would assign an undrained shear strength to liquefied soil

based on N1(60) and use post liquefaction undrained shear when developing p-
y springs using soft clay criteria.

5 ENSOFT, Inc., Austin, Texas.

ATC-119-1 Part 2: Geotechnical Engineering for Tall Buildings 2-15

 3.7.3 Downdrag Loads

Post-earthquake settlement of liquefied soil can cause downdrag loads being

applied to deep foundation (piles and shafts) through negative skin friction
within soil layers above the liquefied soil layer. Downdrag loads are applied
slowly (compared to duration of an earthquake); therefore, they are combined
with static gravity and sustained live loads. Therefore, downdrag loads are
not combined with the loads caused by the inertial response of the building.

Evaluation of magnitude of downdrag loads are discussed in Section 3.3.

3.7.4 Effect on Ground Motion

Soil liquefaction has a significant effect on ground motion intensity and

characteristics, especially for cases where liquefiable soil layer is at depth
with relatively high pre-earthquake confinement pressure and corresponding
shear strength (as compared with liquefiable layer being at a shallow depth).

According to the California Building Code, sites with liquefiable sand layers or
deep soft clays (Young Bay mud in San Francisco Bay Area) fall into Site
Category F, requiring site response analysis, using such computer programs
as SHAKE 6, DEEPSOIL 7, or FLAC 8, except where soft clays and liquefiable
soil layers are completely removed as part of basement construction (i.e.,
basement mat is supported directly on competent soil / rock).

In the first step of performing site response analysis, ground motion is defined
at the surface of competent soil or rock at depth. Then, a one-dimensional
soil profile is constructed based on conventional geotechnical site
investigation and laboratory test results and site-specific shear wave velocity
measurement. Soil degradation curves that show the variation of shear
modulus and damping with shear strain amplitude are based on either:
(1) site-specific resonant column, cyclic triaxial, cyclic direct simple shear
tests, or a combination of these tests (seldom performed for design of
foundation of tall buildings); or (2) results of tests that have been performed
on similar soil units in the past (in San Francisco Bay Area, a number of high
quality tests were performed as part of design and construction of the new
east span of Bay Bridge and expansion of runways at SFO). Finally, site
response analysis is performed in frequency domain (SHAKE) or in time
domain (DEEPSOIL or FLAC).

6 Lysmer, J, Schnabel, P.B., and Seed, H.B.

7 University of Illinois at Urbana-Champaign and Yousef Hashash
8 Itasca Consulting Group, Inc., Minneapolis, Minnesota

2-16 Part 2: Geotechnical Engineering for Tall Buildings ATC-119-1

According to California Building Code, the ground motion computed at the
ground surface through site response analysis cannot fall below 80 percent of
Site Class E.

3.7.5 Liquefaction Related Ground Settlement

Liquefied soils undergo volumetric strain resulting in ground settlement. The

magnitude of settlement would depend on soil density before the earthquake,
earthquake magnitude, and induced cyclic stress ratio. Magnitude of
settlement could be obtained from the following reference:
• Tokimatsu, K., and Seed, H.B., 1987, “Evaluation of settlement in sands
due to earthquake shaking,” Journal of Geotechnical and
Geoenvironmental Engineering.

Although the above reference is three-decades old, it correctly and closely

predicted the magnitude of ground settlement observed in San Francisco as
a result of 1989 Loma Prieta earthquake.

Flexible connections should be used at the interface of utility lines and pile
supported structures to safely accommodate differential settlement caused by
liquefaction related ground settlement.

3.8 Floating Deep Foundations

In some cases, bedrock is very deep (>300 feet) and it would not be
economical or practical to extend deep foundations to gain support within
bedrock at depth. In these cases, if required, floating deep foundations are
used to reduce the total and differential settlement to acceptable limits
discussed in Chapter 2 of this Part.

Floating foundations normally consist of pile-supported mat or pile-supported

grade beam systems. Unlike mat or grade beams supported on end bearing
piles, in case of floating foundations, the mat or grade beams will always
remain in contact with the soil, and therefore, both the piles and the soil
below the mat or grade beams will contribute to the support of structural
loads.

Design of floating foundations requires careful and detailed sampling and

consolidation testing of clayey soils adjacent and below the pile tips and
three-dimensional settlement analyses using programs, such as 3D-PLAXIS 9
(preferred) or 3D-SETTLE 10. For preliminary estimates of foundation
settlement, a hand calculation of settlement is performed using one-

9 Plaxis, Delft, the Netherlands

10 Rocscience

ATC-119-1 Part 2: Geotechnical Engineering for Tall Buildings 2-17

 dimensional consolidation analysis and equivalent raft foundation model. In
this method, axial loads carried by piles are distributed on an imaginary raft
foundation located at a depth equal to 2/3 of pile length. The vertical
pressure is then distributed assuming a load spread based on a 2V (vertical):
1H (horizontal) distribution, spreading out in two orthogonal directions.

The procedure for design of pile-supported floating foundation is more

complicated than design of pile-supported structures where deep foundations
gain support within competent soil / rock at depth. This is because in this
case, pile head stiffness would depend on long term (consolidation related)
foundation settlement. In turn, foundation settlement would depend on pile
head axial load based on pile head stiffness. Therefore, foundation design of
pile-supported floating foundations is performed using an iterative procedure.

The steps taken in an iterative foundation design procedure consist of the

following:
• Long-term foundation settlement is estimated and axial pile head stiffness
is calculated by dividing pile loads by the magnitude of estimated
settlement.
• The structural engineer develops a model of the mat or grade beam
system, and represents axial stiffness of piles using soil spring constants.
The pile head axial loads are obtained from this model.
• The geotechnical engineer performs three-dimensional settlement
analysis using computer programs 3D-PLAXIS (preferred) or 3D-SETTLE
and recalculates the long-term foundation settlement.
• The iterative process continues until the assumed pile head axial stiffness
becomes close to (within 5 percent of) the calculated values.

It is noted that in case of large (3 to 4 inches) consolidation related foundation

settlement, dynamic axial pile head stiffness is much higher than, and is very
different from, those corresponding to the static loading conditions. This
means that axial pile and soil stiffness values during seismic loading
conditions could be higher than static values by an order of magnitude.

3.9 Pile Group Effects

To avoid reduction in axial capacity of a pile in a pile group, piles should have
a minimum center-to-center spacing of three pile diameters. Reduction in
lateral stiffness for pile with center-to-center spacing between 3 and 6 pile
diameters would depend on pile group geometry, pile layout, location of the
pile within pile group, and type / stiffness of soil between piles. Bending

2-18 Part 2: Geotechnical Engineering for Tall Buildings ATC-119-1

moment of a pile in a pile group is somewhat higher than that of a single pile
subjected to the average pile head lateral load within the pile group.

Lateral stiffness of a single pile could be estimated using the computer

program LPILE. Various methods for evaluation of pile group effects are
presented in the following reference:
• Poulos, H.G., 2017, Tall Building Foundation Design, CRC Press.

The hybrid method for evaluation of pile group effects has been developed by
the following reference:
• Ooi, P.S.K., and Duncan, J.M., “Lateral load analysis of groups of piles
and drilled shafts,” Journal of Geotechnical Engineering, Vol. 120, No. 6.

3.10 Modeling of Pile-Supported Structures and Soil-Structure

Interaction

The following report presents three methods for inclusion of seismic soil-
structure interaction effects:
• PEER, 2017, Guidelines for Performance-Based Seismic Design of Tall
Buildings, Version 2.03, Report No. 2017/06, Pacific Earthquake
Engineering Research Center, Berkeley, California.

The three methods are as follows:

• Ignoring presence of soils adjacent to the basement
• Bathtub modeling (soil springs and dashpots are attached to basement
walls) with a single set of ground motions applied to the ends of soil
springs and dashpots that are not connected to the basement walls
• Bathtub modeling, with location-specific (depth-varying) ground motions
applied to the back of soil springs and dashpots

However, according to the present state of practice, interaction between the

structure and foundation is typically ignored and seismic analysis is
performed in two steps:
• In the first step, a fixed base model of the structure is analyzed and the
design base shear force and maximum overturning moment are
calculated at the foundation level. If present, basements are included in
the model; however, interaction between basement walls and surrounding
soils is ignored (i.e., soil springs and dashpots are not included in the
model). The number, location, diameter, and length of pile foundations
are determined based on the magnitude of calculated base shear force

ATC-119-1 Part 2: Geotechnical Engineering for Tall Buildings 2-19

 and axial demand due to gravity, sustained live loads, and maximum
overturning moment.
• In the second step, a model is constructed that includes mat or grade
beams and soil springs representing axial and lateral pile stiffness. The
model is then subjected to base shear, gravity load, sustained live load,
and overturning moments to calculate axial force, lateral force, and
bending moment developed at the pile tops and shear and bending
stresses developed in the mat or grade beam system.

It should be noted that not including the surrounding soil in the numerical
model may be conservative when basement walls are in contact with
relatively soft soils (i.e., passive soil springs are not stiff enough to impact the
magnitude of axial or in-plane forces developed in the basement
diaphragms). However, in cases where basement walls are in contact with
very stiff soil or rock, eliminating passive soil springs may result in
underestimation of axial (in-plane) loads developed in the ground floor
diaphragm through redistribution of shear force in the core between ground
surface and the basement slab.

The current analysis procedure ignores the kinematic soil-structure

interaction (SSI) effects and elongation of structural period due to lateral
flexibility of the foundation. It is thought by the structural engineering
community that these assumptions (i.e., ignoring SSI effects) are
conservative.

The current analysis procedure described above is in contrast with practice in

the offshore and bridge industries. For more than half a century, foundation
of offshore platforms and bridges are included in the dynamic model of the
structure and both kinematic SSI effects and effects of foundation flexibility on
dynamic performance of the structure are accounted for in a more rigorous
manner. Perhaps, the contrast between state of practice in building
foundation design and design of offshore / bridge foundations stem from the
fact that for both offshore and bridge structure, kinematic motion (ground
motion transmitted to the structure) could substantially be different from
ground motion at the ground surface (mudline).

Kinematic motions could be obtained by either: (1) modeling of individual

piles, use of p-y springs to represent lateral soil stiffness, and use of depth-
varying free-field ground motion behind the soil springs (obtained through
one-dimensional site response analysis); or (2) modeling piles using beam
elements, modeling soil using two or three dimensional solid elements, and
attaching the soil mesh with the beam elements with p-y springs. The mass
of above-ground structure is not included when performing kinematic SSI

2-20 Part 2: Geotechnical Engineering for Tall Buildings ATC-119-1

analysis. There are a number of commercially available computer software
appropriate for kinematic SSI analysis including SASSI 11, FLAC, and LS-
DYNA 12, among others. SASSI performs analysis in frequency domain and is
only applicable to linear systems. Its practical application is limited to deeply
embedded structures with very rigid below-ground structural stiffness (BART
San Francisco Transition structure and Transbay Terminal Center structure
were analyzed using SASSI). FLAC is capable of accounting for soil and
structural nonlinearities; however, only the two-dimensional version of this
program is used in practice in dynamic mode. LS-DYNA has successfully
been used in two-, and three-dimensions to perform kinematic SSI analysis.

The simplest way to include pile foundation in analytical model of structure is

through the use of “dummy cantilever” beams. The process involves the
following numerical steps: (1) linearize the pile head lateral stiffness. This is
accomplished by assuming a certain pile head deflection and calculating pile
force based on L-PILE analysis results for either fixed pile head or free pile
head, depending on pile head fixity condition; and (2) calculate the length of a
dummy cantilever beam which has the same pile head stiffness. The cross-
section area of the dummy cantilever beams is adjusted to capture actual pile
axial stiffness values. Because the pile head lateral and axial stiffness in
each iteration are constant, the model is suitable for use in linear modal SSI
analysis.

The next step is to compare the pile head displacements obtained from the
SSI analysis to the assumed pile head displacement used for linearization of
the pile head stiffness. If the assumed and calculated values are different by
more than 5 percent, the process is repeated until the difference between
assumed and calculated pile head lateral displacements obtained from two
consecutive iterations becomes smaller than 5 percent.

This procedure is very well described in the following reference:

• PoLam, I., Kapuskar, M., Chaudhuri, D., 1998, Modeling of Pile Footings
and Drilled Shafts for Seismic Design, MCEER-98-0018, Multidisciplinary
Center for Earthquake Engineering Research, University at Buffalo, New
York.

According to the present standard of practice, if two towers share a common

mat foundation (whether the mat is pile supported or not), the interaction
between the two towers through structure-foundation-structure interaction is

11 System for Analysis of Soil-Structure Interaction, Lysmer, J., Tabatabaie-Raissi,

M., Tajirian, F., Vahdani, S., Ostadan, F.
12 Livermore Software Technology Corporation

ATC-119-1 Part 2: Geotechnical Engineering for Tall Buildings 2-21

 ignored, i.e., both towers are subjected to the same ground motion. More
studies are needed to validate this approach.

3.11 Expected Performance of Piles During the Design Seismic

Event

Because it would be difficult to inspect and repair damage to piles after a

major seismic event, although some yielding of the pile at the connection
points with the pile caps / mat foundation may be acceptable, piles are
generally designed to suffer no appreciable structural damage, with bending
moment and shear stress well below critical limits, i.e., plastic deformations
are allowed at the base of the tower, but not below the ground surface.
According to the current standard of practice, geotechnical engineers are
normally asked to use 0.3 to 0.7 times the EI of concrete pile section when
performing lateral pile analysis.

3.12 Use of Battered Piles

Use of battered piles for support of tall buildings should be avoided. If

needed to resist unbalanced lateral soil pressure in case of sloping ground,
shear keys are commonly used.

2-22 Part 2: Geotechnical Engineering for Tall Buildings ATC-119-1

Chapter 4

New Tall Buildings on Shallow

Foundations

A tall building is supported on a shallow foundation in case of one of the

following conditions: (1) the foundation is directly placed over bedrock; (2) the
foundation is underlain by dense to very dense Colma sand or alluvium
consisting of stiff to very stiff clays and dense to very dense sands; (3) the
building has a deep basement extending well below groundwater table where
highly compressible Young Bay mud or liquefiable soils are excavated as part
of basement construction; or (4) the foundation is placed over improved
ground extending to the surface of firm soil at depth.

In case of improved ground supporting the foundation of a tall building,

integrity of ground improvement elements (e.g., deep soil mixing - DSM)
during both static and seismic loading conditions should be carefully
evaluated. To satisfy integrity of DSM during a major seismic event, DSM is
typically placed in a grid pattern with center to center spacing of 15 feet or
less and wall thickness of 3 feet (replacement ratio of 30% or greater).

4.1 Bearing Capacity and Short or Long Term Settlement

If shallow foundation is not directly placed over bedrock, the bearing capacity
and short or long term settlement of soils below the foundation and above the
bedrock should be carefully evaluated. The density of sandy soil should be
evaluated by performing standard penetration tests or cone penetration test
(CPT) soundings. Sample of stiff to hard clayey soils (e.g., Old Bay clay)
below the foundation should be obtained using pitcher barrel sampler or
equivalent and high-quality tests should be performed on undisturbed
samples to measure shear strength and consolidation characteristics of
clayey soil layers. Use of Shelby tube for obtaining soil samples within stiff to
very stiff clayey soil is not desirable due to sample disturbance and uncertain
consolidation and strength characterization.

A very useful test often performed in the East Coast of United States and
other countries is the pressuremeter test. The results of this test provide
insitu soil stress-strain relations that can be used in computer modeling of

ATC-119-1 Part 2: Geotechnical Engineering for Tall Buildings 2-23

 soil-structure systems. However, pressuremeter tests are seldom performed
for projects in the San Francisco Bay Area.

Settlement calculations should be performed using computer programs such

as 3D SETTLE 1 or 3D PLAXIS 2 (preferred). Timeline of settlement should be
plotted using snapshots at various stages of construction including placement
of shoring, dewatering, completion of excavation, placement of foundation,
termination of dewatering, completion of construction, 5 years after
completion of construction, and 10 years after completion of construction.

When performing foundation settlement or bearing capacity analysis, effects

of hydrostatic uplift pressure in reducing foundation pressure may be
accounted for. However, the time required for recharge after termination of
dewatering should be accurately estimated or measured and considered. For
example, if it takes three years for water pressure under the foundation to
become equal to water pressure outside of the excavation and the
construction is completed in two years, accounting for full hydrostatic uplift
pressure as part of settlement or foundation bearing capacity calculation
could lead to excessive foundation settlement or bearing failure.

The factor of safety (FOS) against bearing failure should be greater than 3.0
for static plus sustained live loads and greater than 2.0 against total loads
including wind and seismic loads. Gain in shear strength within clayey soils
due to consolidation under dead plus sustained live loads may be accounted
for when calculating FOS against total loads.

When shallow foundations are placed directly over or within short distance
from stiff to very stiff clayey soils (e.g., Old Bay clay), in addition to
maintaining a FOS against global bearing failure mechanism, punching shear
failure mechanism should also be considered and FOS against this failure
mechanism should be evaluated. In this case, friction between basement
walls and shoring system may be accounted for provided that the basement
walls are designed to safely resist the resulting tensile forces and reduction in
frictional capacity between basement walls and shoring walls due to presence
of waterproofing membrane is accounted for.

Finally, if a shallow foundation is directly supported on ground improvement

elements such as deep soil mixing (DSM) or jet grouting (JG), the internal
stability, as well as global stability of the DSM or JG system, should be
evaluated during both static and seismic loading conditions. The
geotechnical engineer of record (GEOR) shall be responsible for design of

1 Rocscience
2 Plaxis, Delft, the Netherlands

2-24 Part 2: Geotechnical Engineering for Tall Buildings ATC-119-1

DSM or JG system and this responsibility shall not be passed on to the
specialty Design Build (DB) contractors. DSM or JG shall be placed in grid
pattern with limited size (normally center to center spacing of 15 feet or less)
to adequately resist tensile, shear, and compressional stresses developed
under application of total loads (dead plus sustained live loads plus seismic
loads). Using individual columns of improved ground shall be avoided. The
GEOR shall perform adequate studies to make certain that the ground
improvement system will perform satisfactorily during the design earthquake,
i.e., the ground improvement shall be considered to be part of the structural
system. This is because brittle failure of the ground improvement system and
its partial or total loss of support could result in excessive settlement of the
structure.

A properly designed and constructed ground improvement system could

change site classification from Site Class F (liquefiable soils or deep soft
clays such as Young Bay mud) to Site Class D.

4.2 Foundation-Soil Interaction and Foundation Design Against

Lateral Loads

A general discussion on state-of-the-art and state-of-practice related to

methods for accounting for seismic soil structure interaction (SSI) effects was
discussed previously in Chapter 2 of this Part.

To safeguard against migration of moisture through mat foundations directly

placed over competent soil below the permanent ground water table, a
waterproofing membrane is installed below the mat. Presence of a
waterproofing membrane will result in very low frictional resistance between
the base of the mat and the subgrade. The project team should consult with
waterproofing manufacturers to obtain a safe allowable friction coefficient and
check sliding of the structure under application of design base shear. If the
base shear force is larger than the frictional capacity of the interface between
the bottom of the mat and the waterproofing membrane or between the
waterproofing membrane and the subgrade, passive soil pressure against the
face of the mat and basement walls (if they exist) will be mobilized. In this
case, displacement compatibility between the base resistance and the
passive resistance should be evaluated and the basement walls should be
designed to safely resist the resulting passive forces.

4.3 Effects of Foundation Flexibility on Dynamic Response of a

Tall Building

Base rocking increases the structural period, and hence, reduces the seismic
demand (in the long period range where spectral acceleration decreases with

ATC-119-1 Part 2: Geotechnical Engineering for Tall Buildings 2-25

 an increase in period). Because of this, some practitioners conservatively
ignore foundation flexibility and do not include rotational soil springs in the
structural model, i.e., a fixed based analysis is performed. Although this is a
reasonable approach, it should be noted that rotation at the base will increase
later movement (drift) of the structure. For this reason, some practitioners
attach four vertical bi-linear springs to the base of the mat. The stiffness of
these springs and distance between them are selected such that the rocking
stiffness of the mat is reasonably modeled. If fixed-base analysis is
performed, it would be prudent for the geotechnical engineers to estimate
rocking stiffness of the foundation and the structural engineers to estimate
the magnitude of drift due to Inclusion of foundation flexibility and collectively
make a decision as to whether vertical soil springs should be included in the
structural model.

2-26 Part 2: Geotechnical Engineering for Tall Buildings ATC-119-1

Chapter 5

Evaluation and Retrofit of

Foundations of Existing Tall
Buildings

Evaluation and retrofit of foundation of existing tall buildings pose unique

challenges including:
• The soil report and detailed information about the foundation type and its
installation may or may not exist. In addition, the quality or quantity of
data in older reports may not meet current standard of practice.
• For tall buildings designed before 1970, soil reports (if they do exist) are
silent on design issues related to soil liquefaction, as the phrases “soil
liquefaction” and “cyclic mobility” were conceived in early 1970s (by Kenji
Ishihara of Tokyo University and Harry Seed of University of California at
Berkeley) following the 1964 Niigata and 1964 Alaska earthquakes.
• Based on damage observed during the 1971 San Fernando earthquake,
the building code changed substantially after 1975. For tall buildings
designed in accordance with building codes prior to 1975, design base
shear is lower than that of a tall building designed in accordance with the
current version of the building code.
• Access to the building and overhead clearance in the basement or ground
floor is usually very limited. As such, conventional soil investigation
equipment cannot be mobilized to conduct a thorough geotechnical field
investigation program according to modern standards.
• Retrofit of existing tall buildings normally requires installation of new
shear walls, energy absorbing devices, or new columns to carry gravity
loads. Therefore, either existing shallow spread footings are enlarged,
new shallow spread footings are installed, or micropiles are installed to
augment and strengthen the existing foundation system.

The most challenging aspect of retrofit of foundation of an existing tall

building is to maintain compatibility of stiffness or capacity between existing
foundation and new foundation elements. If compatibility between response
of existing and new foundation elements cannot be maintained, this could

ATC-119-1 Part 2: Geotechnical Engineering for Tall Buildings 2-27

 result in overstressing and thus possible failure of new foundation elements
and inadequate foundation support of the building.

5.1 Geotechnical Investigation Methods Operating from

Basement of an Existing Tall Building

Information regarding the elevation of the interface between soft and firm
soils in combination with knowledge of subsurface conditions and local
geology is very useful in evaluation of the capacity of an existing foundation
and design of new foundation elements.

Methods for geotechnical investigation from the basement of an existing tall

building include:
• Excavation of test pits for exposing footings or pile caps and observing
soil and pile conditions below the existing foundation. Depending on the
depth of footings and groundwater level, excavating test pits may require
shoring and dewatering.
• Excavation of shallow test borings and obtaining soil samples below the
base of the existing foundations using hand augers.
• Performing cone penetration test (CPT) soundings using mini CPT
equipment small enough to allow transport through doorways and
entrances and operable in confined spaces with limited vertical clearance.
Mini CPT equipment is anchored to the existing floor slab to develop
reaction needed for pushing the cone through subsurface soils. The
maximum axial compression load applied would be limited to tensile
capacity of anchors. Experience working in San Francisco indicates that
mini CPTs are able to go through fill and soft Bay mud, but not through
dense Colma sands.

5.2 Use of Micropiles for Strengthening the Foundation of an

Existing Tall Building

Micropiles are commonly installed to add axial capacity to the existing

foundation of buildings being seismically retrofitted.

A micropile is a small diameter (typically less than 12 inches), drilled and

grouted, replacement pile that is typically reinforced by placing a rebar at the
center of the pile. It is constructed by drilling a borehole, placing
reinforcement, and grouting the hole. Micropiles penetrating 40 to 60 feet
into competent soils have axial capacity of up to 600 kips and axial stiffness
of up to 400 kips/in. Post grouting is often used to increase the axial capacity
of micropiles.

2-28 Part 2: Geotechnical Engineering for Tall Buildings ATC-119-1

The most relevant reference for design, construction, and testing of
micropiles is presented below:
• FHWA, 2000, Micropile Design and Construction Guidelines, Publication
No. FHWA-SA-97-070, U.S. Department of Transportation Federal
Highway Administration.

Proper design of micropiles for strengthening of foundation of an existing tall

building would require detailed numerical modeling of the following: (1) axial
stiffness of existing foundation elements (spread footing or pile cap); and (2)
new micropiles. It is critically important to determine the percentage of axial
gravity and seismic loads carried by micropiles both in compression and
tension by properly modeling relative axial stiffness of the existing foundation
system and axial stiffness of new micropiles.

ATC-119-1 Part 2: Geotechnical Engineering for Tall Buildings 2-29

Chapter 6

Foundation Design of
Tall Buildings Near Shoreline

If a tall building is planned near the San Francisco Bay waterfront, foundation
design should consider potential for shoreline instability and its effects on
performance of foundation during a major seismic event.

Based on observations during the 1906 and 1989 earthquakes, detailed

recent studies performed at Treasure Island using nonlinear numerical
modeling (2D FLAC1 and 2D PLAXIS 2 analyses), and similar detailed studies
performed at waterfront sites in San Francisco, the magnitude of maximum
lateral spreading along the shoreline during a major seismic event is
estimated to be on the order of 10 to 15 feet. These studies indicate that, if
large (10 to 15 feet) lateral spreading occurred at the shoreline, the
magnitude of lateral ground movement will decay from its maximum value at
the shoreline to less than 6 inches at a distance of about 300 feet. It is also
judged that “pile pinning effects” would reduce the free field displacement of
less than 6 inches to much smaller values through kinematic soil structure
interaction. Therefore, well-designed and constructed pile-supported
structures are expected to perform satisfactorily at distances equal or greater
than 300 feet from the shoreline during a major seismic event. Thus, if a tall
building is planned at a distance greater than 300 feet from the shoreline,
foundation design could proceed without considering the potential for
shoreline instability. However, the effects of sea level rise should be
considered as discussed in Chapter 8 of this Part.

For sites located within 300 feet from the shoreline, seismic shoreline
instability has to be carefully evaluated through literature survey and review
of available data and reports with information related to seismic seawall and
shoreline stability. If only limited information is available regarding seismic
shoreline stability, performing simplified slope stability calculations alone
would not provide adequate or defensible estimates of variation of lateral
ground displacement with distance away from the shoreline. Thus, detailed,

1 Itasca Consulting Group, Inc., Minneapolis, Minnesota

2 Plaxis, Delft, the Netherlands

ATC-119-1 Part 2: Geotechnical Engineering for Tall Buildings 2-31

 two-dimensional, time-dependent numerical analysis should be performed
utilizing software such as 2D FLAC.

Ground motions should be developed at the bedrock or surface of competent

soil at depth and nonlinear behavior of soft and liquefiable soils should be
modeled considering generation of pore water pressure within sandy soils
and strain softening within clayey material. If the magnitude of computed
free-field lateral displacement is less than 6 inches at the pile location,
foundation design could proceed without considering potential for shoreline
instability. However, if the magnitude of free-field lateral displacement is
calculated to be larger than 6 inches at the pile location, the foundation
design should consider lateral loading on the foundation due to shoreline
instability. In this case, the most economical way to address foundation
design might be to mitigate the shoreline instability through the use of a
suitable ground improvement technique, such as deep soil mixing, jet
grouting, stone columns, compaction grouting, deep dynamic compaction,
vibro-compaction, vibro-replacement, rapid-impact compaction. If ground
improvement is used to mitigate seismic shoreline instability issues, the
geotechnical engineer of record (GEOR) should perform adequate numerical
modeling and analysis to make certain that the improved soil block will
maintain its global stability as well as internal stability. Individual columns of
improved ground shall not be used, deep soil mixing or jet grout system
should be placed in a grid pattern. The responsibility for design of ground
improvement system shall not be passed on to a design-build specialty
contractor.

It should be noted that evaluation of the time history of lateral ground

movement using a nonlinear, time-dependent analytical procedure follows a
performance-based design approach. As such, the risk-baced MCER ground
motion should be used to check against collapse of the structure as a result
of foundation failure.

2-32 Part 2: Geotechnical Engineering for Tall Buildings ATC-119-1

Chapter 7

Shoring and Dewatering

In cases where excavation cannot be completed by slopping back the ground

surrounding the excavation, shoring systems with vertical walls are often
used. Commonly used shoring systems in the San Francisco Bay Area
include sheet pile walls, soldier pile and lagging, secant piles, and deep soil
mixing (DSM) walls. Secant and DSM walls are reinforced by placing steel
H-beams in regular spacings of 4 to 8 feet on center while the mix is in a fluid
state. Information regarding design and construction of shoring systems
could be found in the following reference:
• FHWA, 2008, Earth Retaining Structures, Publication No. FHWA-NHI-07-
071, U.S. Department of Transportation Federal Highway Administration.

In case of DSM walls, the soil-cement mix acts as lagging between soldier
beams. The stresses developed in DSM due to apparent earth pressure and
deformation of soldier piles should be checked against design strength of
soil-cement mix (typically 300 psi in unconfined compressive strength).

Installation of soldier pile and lagging system in areas with high groundwater
table would result in lowering of groundwater table and settlement of ground
outside of excavation which would be problematic when used in well-
developed and congested areas. Overlapping secant piles and DSM shoring
system would minimize (if not eliminate) lowering of groundwater table
outside of the excavation. Therefore, overlapping secant piles and DSM
shoring systems are often used in case of a high groundwater table and when
excavation could have an adverse effect on the foundation of adjacent
structures. It is noted that lowering of the groundwater table outside of
excavation to historical low groundwater table would be acceptable as the soil
has already experienced the higher effective vertical stresses associated with
the lower historical groundwater table. Due to the cycles of drought in the
San Francisco Bay Area, a drop of 5 feet in groundwater table when
measured in rainy years is judged to be acceptable.

It is not practical to have a cantilever shoring system with a height of more

than 12 feet. Therefore, a shoring system with a height exceeding 12 feet
would require either internal bracing or tiebacks.

ATC-119-1 Part 2: Geotechnical Engineering for Tall Buildings 2-33

 Geotechnical investigation requirements for design and installation of
tiebacks; apparent earth pressure in sand, stiff clays, and soft to medium stiff
clays; gross and net water pressure diagrams; accounting for surcharge
loads; unbounded length requirements and design of anchor bond length;
tieback slope and spacing requirements; tieback testing procedure including
performance test, proof test, and creep test (creep test is specially required in
case of tiebacks installed in soft to medium stiff clays); lock-off load
requirements, anchor lift-off testing procedure; corrosion protection
requirements (in case of permanent shoring walls), depth of penetration
below excavation; design of soldier beams against axial loads due to inclined
tieback loads; expected lateral deformation of well-designed and well-
constructed shoring systems using tiebacks; inspection and short term / long
term monitoring requirements; and other pertinent information are presented
in the following reference:
• FHWA, 1999, Ground Anchors and Anchored Systems, Publication No.
FHWA IF-99-015, U.S. Department of Transportation Federal Highway
Administration.

A common failure mode during installation of soldier pile and lagging system
is flow of saturated loose sands into the excavation from areas below the
lowest installed lagging board, creating a cavity behind the wall that could
extend upwards to the ground surface. If this occurs, dewatering of soils
behind shoring wall could mitigate this failure mode.

A common failure mode during installation of tieback system is the flow of

loose saturated sands around the auger and creation of a cavity between the
auger and the ground surface. To investigate whether this failure mode is
occurring, the volume of soil being extracted should be carefully checked
against auger size and rate of advancement of the auger. In one case,
ground surface settlement was observed below a building due to a cavity
developed when the auger was spinning but not advancing with a rate
proportional to the volume of soil being extracted.

In cases of deep excavation in soft to medium stiff clays, special attention

should be given to basal heave and calculation of factor of safety (FOS)
against blow out conditions discussed in the above reference. In the San
Francisco Bay Area, there are sand lenses and layers within Young Bay mud
and Old Bay clay layers. Dewatering of these layers may be required to
prevent basal heave instability or blow out conditions. Additional information
for calculation of FOS against basal heave instability is provided in the
following reference:

2-34 Part 2: Geotechnical Engineering for Tall Buildings ATC-119-1

• Wu, S.-H., Ou, C.-Y., Ching, J., 2014, “Calibration of model uncertainties
in base heave stabilit for wide excavations in clay,” Soils and
Foundations, Vol. 54, No. 6.

It should be noted that for deep excavations, maximum ground settlement

may occur away from the shoring wall due to deep seated mode of ground
deformation. For example, for a 75 foot deep excavation in downtown San
Francisco, the maximum ground settlement was calculated to be about 50 to
70 feet away from the face of the excavation.

A conservative safe distance from the face of an excavation is twice the

depth of excavation. However, a more realistic safe distance is where the
ground deformation is reduced to less than 1/4 inch. In case of a properly
designed and constructed shoring system, ground deformation criteria of less
than 1/4 inch would generally be satisfied at a distance greater than or equal
to the height of the excavation.

The following reference provides additional information regarding design and

construction of DSM walls:
• FHWA, 2013, Deep Mixing Embankment and Foundation Support, FHWA
Design Manual, FHWA-HRT-13-046, U.S. Department of Transportation
Federal Highway Administration.

To prevent stress relaxation and excessive lateral movement of an internally

braced shoring system, struts are normally preloaded to 75% to 80% of loads
calculated based on the apparent earth pressure.

When DSM walls are installed within zone of influence of foundation of

adjacent structures, consideration should be given to potential lateral
movement of the DSM wall and resulting settlement of adjacent structures
while DSM remains in fluid state. This condition is especially critical when
DSM walls extend below the tip of pile foundations supporting adjacent tall
buildings.

Empirical ground settlement data for internally braced or tied-back shoring

walls is presented in the following reference:
• Peck, R.B., 1969, “Deep excavation and tunneling in soft ground,”
Proceeding of the 7th International Conference on Soil Mechanics and
Foundation Engineering, Mexico City, pp. 225-290.
In this reference, ground settlements are given at various distances from the
excavation. Both ground settlement and distance from excavation are
normalized to the depth of excavation.

ATC-119-1 Part 2: Geotechnical Engineering for Tall Buildings 2-35

 Maximum lateral movement of internally braced or tied-back walls as a
function of system stiffness and FOS against basal heave is presented in the
following reference:
• Clough, G.W., and O’Rourke, T.D., 1990, “Construction induced
movements of in-situ walls,” Design and Performance of Earth Retaining
Structures, No. 25, pp. 439-470.

Some practitioners use finite element or finite difference computer codes to

estimate the magnitude of ground deformation in case of complicated or
critical projects. The behavior of saturated clayey soils is generally modeled
considering short term loading (undrained) conditions using total stress
method of analysis. Although this method is appropriate for modeling of the
stiffness and capacity of clayey soils, if the construction duration is extended,
additional ground deformation may occur as the behavior of saturated clayey
soils change from undrained conditions to drained conditions. As such, the
analysis should employ the effective stress method as opposed to the total
stress method.

2-36 Part 2: Geotechnical Engineering for Tall Buildings ATC-119-1

Chapter 8

Sea Level Rise

The following reference presents a pilot study conducted by FEMA regarding

sea level rise for San Francisco County:
• FEMA, 2016, Sea Level Rise Pilot Study, Future Conditions Analysis and
Mapping San Francisco County, California, prepared by BakerAECOM for
the Federal Emergency Management Agency, Washington, D.C.

This study developed the following equations to estimate mid- and high-range
sea level rise (SLR):
• SLR (inches) = 0.0024 × t 2 + 0.12 × t, where t = years from 2000
SLR (inches) = 0.00366 × t + 0.2936 × t, where t = years from 2000
2
•

Accordingly, the mid-range estimate for sea level rise in San Francisco
County in the years 2050 and 2100 are 12 inches and 36 inches,
respectively. The high-range estimate for SLR in the years 2050 and 2100
are 24 inches and 66 inches, respectively. Groundwater table elevation at
sites close to the Bay will rise with sea level rise. However, the effect of sea
level rise on groundwater table elevation for sites away from the shoreline
(especially for site at higher elevations) diminishes as the distance from the
shoreline increases. This is relevant to geotechnical engineering because
sea level rise will increase the thickness of saturated sandy fills, and will
increase the potential for soil liquefaction related lateral spreading. Other
effects of sea level rise include an increase in hydrostatic pressure on
basement walls, an increase in hydrostatic uplift pressure, which would
impact design of ground anchors (tiedowns), and an increase in uplift
demand on deep foundations.

ATC-119-1 Part 2: Geotechnical Engineering for Tall Buildings 2-37

Chapter 9

Recommendations

Best practices are not fully codified and implementation of best practices
varies considerably between practitioners. AB083 set the bar for structural
design of tall buildings; need to develop a companion AB for foundation
design. It is essential to bring consistency and uniformity to geotechnical
practice for foundation design of tall Buildings by establishing guidelines for
best practices w/o adversely impacting the construction cost or project
schedule. It is recommended that SFDBI develop an Administrative Bulletin
(AB) or Guidelines to address the following:
1. Acceptable limits of total and differential settlement
2. Evaluation / mitigation of geologic hazards including soil liquefaction and
lateral spreading
3. Site characterization (field investigation and laboratory testing)
4. Foundation analysis and design
5. Numerical modeling, especially those dealing with bearing capacity
calculation and evaluation of foundation settlement
6. Design of shoring / dewatering systems
7. Development of design ground motions and incorporation of soil-structure
interaction effects
8. Design consideration related to sea level rise
The proposed AB / Guidelines will strengthen SFDBI in assessing the
appropriateness / completeness of foundation design recommendations
presented in geotechnical reports for tall buildings

The following 15 recommendations are extracted from material presented in

previous chapters of this Part. These recommendations should not be
followed without a thorough review of the relevant sections.
• Recommendation 1 - Total and Differential Settlement. There are no
universally accepted criteria for maximum allowable total and differential
settlement of foundation for support of commercial / residential tall
buildings. However, many practitioners and officials with building
departments of various cities consider the total settlement of less than 4

ATC-119-1 Part 2: Geotechnical Engineering for Tall Buildings 2-39

 inches and limiting differential settlement to an angular distortion of 1/500
for composite steel or concrete core buildings and 1/750 for reinforced
concrete towers (rigid structures) to be acceptable. The structural
system, cladding system, partitions, and other nonstructural components
may also control the acceptable differential settlement. Some buildings
(e.g., base-isolated structures) are more sensitive to differential
settlement than other conventional buildings and would require a tighter
limit on differential settlement. However, base-isolation systems are
seldom used in tall buildings. Detailed discussion is presented in Chapter
2 of this Part.
• Recommendation 2 - Monitoring of Building Settlement. Building
settlement should be monitored during construction and to at least 10
years after construction. Monitoring and reporting requirements should
follow SFDBI guideline presented on Sheet S-18.
• Recommendation 3 - Tall Buildings Supported on Large Diameter
(≥ 4 feet) Drilled Shafts Extending to or into Bedrock. The ultimate
skin friction in soil and skin friction / bearing capacity within rock socket
should be evaluated using Osterberg (O-Cell) or equivalent test method.
If end bearing is accounted for in the design, the bottom of drilled shaft
should be cleaned out using “one-eye” bucket or flat bottom clean out
bucket or airlifting method and verified by Mini-SID (Shaft Inspect Device)
or DID (Ding Inspection Device). For drilled shafts socketed in bedrock,
the minimum rock socket length should generally be 15 feet or twice the
shaft diameter, whichever is greater. The minimum rock socket length
requirement could be relaxed for sites where: (1) depth to bedrock is
greater than 25 times the shaft diameter; and (2) hard rock with
substantial lateral extension and depth below the bottom of drilled shafts
is encountered (i.e., the hard rock bolder is a size of a house and not a
compact car). Detailed discussion is presented in Section 3.1 of this Part.
• Recommendation 4 - Tall Buildings Supported on 24-inch or Larger
Auger Cast Piles. The use of Auger cast piles (ACP or CFA) should
either be avoided or carefully monitored if the subsurface conditions
include layers of soft clay (e.g., Young Bay mud) or saturated loose sand.
Installation of auger cast piles in all soil types should be installed by an
experienced foundation contractor, preceded by load testing, and
production pile installations monitored by skilled field inspectors or
engineers and automated monitoring equipment (AME) data acquisition
system. All Auger cast piles should include tubes for possible gamma-
gamma, cross hole sonic, or thermal testing. Ten percent of auger cast
piles should be tested for integrity as directed by the geotechnical

2-40 Part 2: Geotechnical Engineering for Tall Buildings ATC-119-1

 engineer of record (GEOR) using the above-mentioned methods. In
addition, all auger cast piles with questionable installation record should
be integrity tested. Static pile load test in combination with high strain
dynamic pile load test should be performed to verify the axial capacity of
the piles. Detailed discussion is presented in Section 3.1 of this Part.
• Recommendation 5 - Tall Buildings Supported on Specialty Deep
Foundations. If used, specialty piles (torque-down piles, Tubex piles,
enlarged base (Franki) piles, Micropiles, Fundex piles, etc.) are designed
and installed by specialty design build (DB) contractors. The GEOR
should review design submittals by DB contractors, including review of
both static and seismic loading conditions and load paths. The GEOR is
ultimately responsible for geotechnical aspects of foundation design.
Detailed discussion is presented in Section 3.1 of this Part.
• Recommendation 6 - Allowable Tolerance for Installation of Deep
Foundations. Deep foundations should meet plumbness criteria of less
than 1.25 inch for every 10 feet of depth. Detailed discussion is
presented in Section 3.5 of this Part.
• Recommendation 7 - Soil Liquefaction Design Issues. The potential
for soil liquefaction and its effects on ground motion and foundation
(temporary loss of axial support, reduction in lateral support, development
of downdrag loads, ground settlement) shall be carefully evaluated and
addressed. Detailed discussion is presented in Section 3.7 of this Part.
• Recommendation 8 - Tall Building on Floating Deep Foundations. In
case of tall buildings supported on deep foundations that don’t extend to
bedrock, the potential for short-term and long-term total and differential
settlement should be carefully evaluated. Undisturbed soil samples of
clayey soil should be obtained using Pitcher Barrel sampler and high-
quality consolidation tests should be performed on adequate number of
soil samples. Three-dimensional settlement analysis should be
performed using computer program 3-D PLAXIS (preferred) or 3-D
SETTLE. It is recognized that long-term, consolidation related foundation
stiffness is much lower than foundation stiffness during seismic loading
conditions. Therefore, both long-term and seismic foundation stiffness
should be evaluated and used in structural design. Pressure meter tests
are used in the eastern United States and other countries to provide very
useful information regarding the insitu stress-strain relationships, but
unfortunately is not being performed in the San Francisco bay Area.
Geotechnical engineers practicing is the Bay Area are encouraged to
perform pressure meter test and develop a data base correlating the

ATC-119-1 Part 2: Geotechnical Engineering for Tall Buildings 2-41

 pressure meter test results to laboratory test result. Detailed discussion is
presented in Section 3.8 of this Part.
• Recommendation 9 - Pile Group Effect. Pile group effects on axial pile
capacity and axial / lateral pile head stiffness depends on pile diameter,
pile spacing, and soil characteristics. These effects should be evaluated
as part of foundation design. Detailed discussion is presented in Section
3.9 of this Part.
• Recommendation 10 - Soil-Structure Interaction Effects. Evaluation
of soil-structure interaction (SSI) effects should follow TBI’s guidelines for
performance-based seismic design of tall buildings. Accounting for
kinematic SSI effects could potentially lead to economic design of the
buildings. Computer programs such as SASSI, FLAC, LS-DYNA, among
others could be used to perform two, or three-dimensional kinematic SSI
analysis. Detailed discussion is presented in Section 3.10 of this Part.
• Recommendation 11 - New Tall Buildings on Shallow Foundation.
The bearing capacity and short-term / long-term settlement of foundation
should be carefully evaluated. In-situ density of sandy soil should be
obtained using SPT or CPT, as appropriate. Samples of stiff to hard
clayey soils (Old Bay clay) below the foundation and above bedrock
should be obtained using Pitcher Barrel sampler and high-quality
consolidation tests should be performed on adequate number of samples.
Use of thin wall Shelby tube sampler in stiff clayey soils should be
avoided. Three-dimensional settlement analysis should be performed
using computer program 3-D PLAXIS (preferred) or 3-D SETTLE.
Foundation bearing capacity calculation should consider both punching
shear failure mechanism and global failure mechanism. Effects of
hydrostatic uplift pressure on reducing foundation bearing pressure may
be accounted for; however, time required for recharge and increase of
water pressure to a level outside of the excavation should be accurately
estimated or field measured during construction. Detailed discussion is
presented in Section 4.1 of this Part.
• Recommendation 12 - Evaluation and Retrofit of Foundation of
Existing Tall Buildings. Micropiles are commonly used to supplement
the axial compression / tensile capacity of the existing foundations. It is
critical to evaluate the axial capacity and stiffness of both the existing
foundations and new foundations and include axial stiffness of these
elements accurately in the computer model to evaluate a realistic
distribution of static and seismic loads between the existing and new
foundations. All micropiles (cased on uncased) shall have an upper pile
section (> 20 ft long) of segmental step-tapered joint, flush couples steel,

2-42 Part 2: Geotechnical Engineering for Tall Buildings ATC-119-1

 API rated pipe casing and be pile load tested regardless of pile load
capacity. Detailed discussion is presented in Chapter 4 of this Part.
• Recommendation 13 - Foundation Design of Tall Buildings Near
Shoreline. For sites within 300 feet from the shoreline, effects of
shoreline instability (lateral spreading) on performance of the foundation
should be evaluated by a geotechnical engineer experienced in design of
near-shore structures and mitigation measures be implemented to avoid
overstressing and failure of the foundation during the design seismic
event. The shoreline stability evaluation could be performed in phases
consisting of literature survey, simplified slope stability and deformation
analysis, detailed two-dimensional finite element / finite difference
modeling, with each phase of evaluation becoming more detailed, time
consuming, and costly than the previous phase. It is believed that if the
magnitude of free-field lateral spreading is six inches or less at the closest
pile location to the shoreline, the presence of pile foundation could
mitigate the lateral spreading through kinematic SSI effects (otherwise
referred to as pile pinning effects). Detailed discussion is presented in
Chapter 6 of this Part.
• Recommendation 14 - Shoring and Dewatering. Critical issues related
to shoring and dewatering include potential for settlement of adjacent
streets and structures, basal heave if shoring system is installed in soft to
medium stiff clays, loss of ground by flow of saturated loose sandy soil
behind the shoring into the excavation from areas below the lowest
installed lagging board, loss of ground in saturated sandy soils while
installing tie-backs, and increase in ground deformation adjacent to a
shored excavation with time as the behavior of saturated clayey soils
changes from undrained to drained conditions. The shoring should be
designed by qualified professional California Civil Engineer or California
Structural Engineer retained by the contractor. Detailed discussion is
presented in Chapter 7 of this Part.
• Recommendation 15 - Sea Level Rise. Potential for sea level rise
should be accounted for as part of design of foundation and basement
walls of a new tall building. Detailed discussion is presented in Chapter 8
of this Part.

ATC-119-1 Part 2: Geotechnical Engineering for Tall Buildings 2-43

Appendix A

Case Histories: Settlement

of Tall Buildings

The following reference presents case histories comparing predicated and

measured settlement of tall buildings:
• Poulos, H.G., 2017, Tall Building Foundation Design, CRC Press.

The case histories include La Azteca Building in Mexico, Emirates Twin

Towers in Dubai, Burj Khalifa in Dubai, Incheon 151 Tower in South Korea,
and a tower on karstic limestone in Saudi Arabia. The subsurface conditions
of the Incheon 151 Tower in South Korea are very similar to those in the San
Francisco Bay Area.

The following is a summary of the conclusions from the reference:

• Characterizing the ground conditions accurately is more critical than the
method of analysis employed to carry out the settlement calculations,
provided that the analysis method is reasonably sound and properly
reflects the mechanism of behavior.
• The following applies to immediate settlement of tall buildings on
unsaturated soils: When there are large number of piles in a pile-
supported mat foundation of a tall building, the calculation of interaction
among piles can have a large influence on group settlement, i.e.,
overestimating group effects could result in overestimation of foundation
settlement by a large factor. Also, assuming constant soil stiffness with
depth could result in overestimation of foundation settlement (because of
reduction in soil stress and strain, soil modulus at depth is much higher
than soil modulus at shallow depth).
• Foundation settlement calculation performed in case of tall buildings
supported on drilled shafts socketed into soft rock using pile load test
data and finite element analysis compare fairly well with measured data,
with predictions often being somewhat larger than recorded data.

Lessons learned from the Incheon 151 Tower with subsurface conditions
similar to the San Francisco Bay Area indicate that commercially available
computer programs could accurately predict foundation settlements

ATC-119-1 Part 2: Geotechnical Engineering for Tall Buildings 2-45

 consisting of drilled shafts socketed into soft rock by accounting for pile-soil-
pile interaction effects.

2-46 Part 2: Geotechnical Engineering for Tall Buildings ATC-119-1

 PART 3:
Performance Expectation
for New Buildings
Chapter 1

Introduction

1.1 Background

This Part examines the design requirements and expected seismic

performance of new buildings designed and constructed according to the
current San Francisco Building Code (SFBC) (CCSF, 2016). The Part
describes the seismic assessment of two archetype buildings—a reinforced
concrete shear wall (RCSW) residential building and a steel buckling
restrained braced frame (BRBF) office building—the results of which are used
to present the potential costs and benefits of higher performance goals for
new tall building construction. This Part corresponds to Recommendation 1B
presented in the Summary Recommendations.
This Part addresses aspects of the following tasks described in the CAPSS
Earthquake Safety Implementation Program’s Workplan 2012-2042 (CCSF,
2011):
• Task B.6.a: Update codes for new buildings to reflect desired
performance goals and acceptable confidence levels in meeting them

1.2 Seismic Performance Objectives

The primary objective of modern seismic codes is to ensure life safety under
extreme earthquake events. Building code requirements for seismic design
are primarily aimed at limiting the risk of structural collapse under severe
earthquake ground motions and damage to structural and nonstructural
components that can pose life-safety risks (e.g., falling hazards and blocked
egress). Even when the code is satisfied, damage to structural, architectural,
or mechanical components can still require extensive and costly repairs and
impact building functionality after a large earthquake.

In San Francisco and other West Coast cities with high seismic hazards,
most tall buildings constructed since about 2008 have been designed using a
performance-based seismic design approach, which helps ensure reliability of
building response under strong earthquakes. This is in contrast to
conventional seismic design methods, which rely on prescriptive procedures
and criteria that are less transparent in terms of the expected building
performance. Although the performance-based procedures provide a greater

ATC-119-1 Part 3: Performance Expectation for New Buildings 3-1

 degree of confidence in engineering design, their use in San Francisco (and
other West Coast cities) is rarely intended to provide an enhanced level of
performance. Rather, they are calibrated to provide performance similar to
that achieved with conventional prescriptive design methods.

About 10% of the San Francisco tall building inventory (see Part 1) has been
constructed or permitted since 2000, including a significant number of
residential buildings. The continuing high demand and prices for both
residential and commercial office space is likely to fuel continued
development of tall buildings in San Francisco for the foreseeable future. In
addition to the overall amount of new building construction, the increasing
density of residential and commercial occupancies raises concerns that
building damage and downtime due to earthquakes can have
disproportionate effects on building residents and urban communities (see
Part 5).

In addition to direct economic losses, building damage can lead to indirect

economic losses due to downtime, defined as the time required to achieve a
recovery state after an earthquake. Bonowitz (2011) defines three recovery
states: (1) reoccupancy; (2) functional recovery; and (3) full recovery.
Reoccupancy occurs when the building is deemed safe enough to be used
for shelter, although functionality may not be restored. Functional recovery
occurs when the building is repaired to restore most of its primary function,
(i.e., it is operational). Full recovery occurs when the building is restored to
its pre-earthquake condition. This Part focuses its evaluation on the
intermediate functional recovery criteria to align with the City’s tentative
recovery goals (ATC, 2018).

Two recent studies of a hypothetical reinforced concrete shear wall

residential building, representative of recently constructed buildings in San
Francisco, indicate that the building may incur damage requiring repairs
costing about 15% of building replacement cost under a design earthquake
(Tipler, 2014) and 5% under a magnitude-7 Hayward fault earthquake
(Almufti et al., 2018). The percent economic losses are not uncommon for
earthquake loss assessment studies of modern buildings. The same studies
indicate that the buildings may experience downtimes to functional recovery
on the order of 84 weeks under the design earthquake and 33 weeks under
the magnitude-7 Hayward earthquake. The studies highlight how damage
that results in moderate repair costs could result in excessive downtimes,
leading to displacement of residents with associated indirect costs.

3-2 Part 3: Performance Expectation for New Buildings ATC-119-1

Motivated by issues highlighted from these and other recent studies, this Part
further examines the expected seismic performance of modern tall buildings
and how the performance relates to current building code requirements.

1.3 Organization

Chapter 2 describes the designs of the archetypes buildings, the seismic

assessment methodology, and results.

Chapter 3 summarizes seismic performance results from other studies that

have investigated tall buildings.

Chapter 4 presents recommendations and potential cost implications for

achieving improvements in the expected seismic performance of new tall
buildings.

A list of references is provided at the end of this Part.

ATC-119-1 Part 3: Performance Expectation for New Buildings 3-3

Chapter 2

Archetype Tall Buildings

2.1 Design Basis and Objectives

The archetype tall buildings investigated in this study represent trends in

modern tall building construction: one is a reinforced concrete shear wall
(RCSW) residential building and the second is a steel buckling-restrained
braced frame (BRBF) office building. Two predominant forms of recent tall
office building construction in San Francisco are: (a) concrete shear wall core
with steel gravity framing (e.g., Salesforce Tower); and (b) steel BRBF
systems (e.g., 181 Fremont and Oceanwide Center Buildings). In addition,
the tall building inventory in Part 1 demonstrates that 17 buildings adopting a
RCSW system have been constructed or permitted for construction in San
Francisco since the year 2000. In this study, the BRBF system is selected as
representative of office buildings because the RCSW system is partially
represented by the residential concrete archetype. Both archetype buildings
were designed to comply with current San Francisco Building Code (SFBC)
seismic design requirements.

In San Francisco, new buildings are designed using either a code-prescriptive

or a performance-based seismic design approach. Prescriptive seismic
designs adhere to the design methods and materials prescribed by the
national standard ASCE/SEI 7-10, Minimum Design Loads for Buildings and
Other Structures (ASCE, 2010), which is adopted by the SFBC.
Performance-based seismic design employs advanced nonlinear structural
analysis to design structural building systems that are capable of the same
(or better) seismic performance as a code-prescriptive structural design. In
addition to providing more reliable assessment of seismic performance, the
performance-based method allows designers to employ structural systems
that are not allowed by traditional prescriptive requirements. The
performance-based methods are permitted by SFBC Section 104A.2.8, which
allows alternative materials and methods of construction as approved by the
building official. Despite these different analysis methods and design criteria,
buildings constructed using code-prescriptive or performance-based designs
are considered to be equivalent in their structural safety and adherence to
minimum building code standards. San Francisco’s Administrative Bulletin
(AB) 083, Requirements and Guidelines for the Seismic Design of New Tall

ATC 119-1 Part 3: Performance Expectation for New Buildings 3-5

 Buildings using Non-Prescriptive Seismic-Design Procedures (CCSF, 2008a)
provides basic requirements and guidelines for the seismic design of tall
buildings following performance-based seismic design procedures, which are
typically required to have an independent, third-party peer review to help
confirm that the proposed building’s structural system meets the minimum
code safety standards and requirements. AB 082, Guidelines and
Procedures for Structural Design Review (CCSF, 2008b), establishes
guidelines for the design peer review process. Most, if not all, projects that
are permitted by AB-083 will follow Guidelines for Performance-Based
Seismic Design of Tall Buildings (PEER, 2017).

As outlined in AB 083, the performance-based seismic design approach

entails evaluations under three levels of earthquake ground shaking: a
Service Level Earthquake (SLE), a code-level earthquake, referred to as a
Design Earthquake (DE), and a Maximum Considered Earthquake (MCE).
The implicit goals are to provide “collapse-prevention” under MCE motions,
life-safety in a DE event, and limited structural damage under SLE. These
checks and the associated acceptance criteria are intended to represent the
earthquake performance implied by national standards, such as ASCE/SEI
7-10, that are referenced by the SFBC.

The SLE evaluation is required to demonstrate acceptable seismic

performance for moderate earthquakes. It evaluates anticipated seismic
performance under a seismic hazard level that can be reasonably expected
during the service lifetime of a building (nominally, ground motions with a
50% probability of being exceeded in 30 years).

The DE evaluation is used to identify the exceptions being taken to the

prescriptive requirements of the SFBC and to define the minimum required
strength and stiffness for earthquake resistance. Minimum strength and
stiffness requirements are defined according to SFBC. Providing a non-
prescriptive seismic design with minimum strength and stiffness comparable
to code-prescriptive designs helps produce seismic performance at least
equivalent to the code. Minimizing the number of exceptions to prescriptive
requirements also helps achieve this aim. As specified in ASCE/SEI 7-10,
the DE-level ground motions are defined as two-thirds of the MCE-level
ground motions. For a Site Class D location in downtown San Francisco, the
DE-level motions have roughly a 10% probability of being exceeded in 50-
years.

The MCE evaluation is intended to verify that the structure has an acceptably
low probability of collapse under severe earthquake ground motions. The
evaluation uses nonlinear response-history analysis to demonstrate an

3-6 Part 3: Performance Expectation for New Buildings ATC 119-1

acceptable mechanism of nonlinear lateral deformation and to determine the
maximum forces to be considered for structural elements and actions
designed to remain elastic. As determined according to requirements of
ASCE/SEI 7-10, the MCE-level ground motions for a Site Class D location in
downtown San Francisco have roughly a 4% probability of being exceeded in
50-years (similar in shaking intensity to extreme ground motions experienced
in the 1906 magnitude-7.9 San Francisco earthquake).

Table 3-1 provides an overview of seismic design requirements at the three

levels of seismic evaluation per AB-083. The table provides an overview of
the objective of each evaluation, analysis methods, drift limits, acceptance
criteria, strength reduction factor and relevant code or guidance documents.
More complete details of the design requirements can be found in SFBC and
Guidelines for Performance-based Seismic Design of Tall Buildings (PEER,
2010).

Table 3-1 Performance-Based Design Criteria

Service Level Maximum Considered
Earthquake1 Design Earthquake2 Earthquake2
(SLE) (DE) (MCE)
Code Compliance
Overall Objective Limited Damage (Life-Safety) Collapse Prevention
Analysis Method Linear (Response Spectrum) Linear (Response Spectrum) Nonlinear Response History
Response Modification Factor --- Code-prescribed R-factor ---
Mean Transient < 3%
Story Drift Limits < 0.5% < 2%
Mean Residual < 1%
Demand < 1.5 Nominal Force and Deformation-
Component Acceptance Criteria Demand < Design Strength
Strength3 Controlled Component Checks
Code or Guidance Document PEER Guidelines (2017) ASCE/SEI 7-10 PEER Guidelines (2017)
1. In PEER Guidelines (2017), the SLE ground motion is specified based on a mean annual return period of 43 years, corresponding to
a 50% probability of exceedance in 30 years.
2. In ASCE/SEI 7-10, the MCE ground motion is determined by a combination of criteria. For a Site Class D soil site in downtown San
Francisco, the MCE ground motion intensity is generally governed by the 84th percentile hazard from a magnitude-8 earthquake on
the San Andreas fault, corresponding to a return period on the order of 1,200 years (~4% probability of exceedance in 50 years).
For a Site Class B rock site, the MCE intensity is generally governed by the risk targeted ground motion, corresponding to a return
period on the order of 2,500 years (~2% probability of exceedance in 50 years). The DE ground motion is specified as two-thirds of
the MCE intensity, which equates to about a 500-year return period (~10% probability of exceedance in 50 years).
3. In PEER Guidelines (2017): Demand < 1.5 Nominal Strength (i.e., no strength reduction factor is applied at the service-level check).
In PEER Guidelines (2010): Demand < 1.5 Design Strength (i.e., strength reduction factor is applied).

2.2 Reinforced Concrete Shear Wall Residential Building

The reinforced concrete shear wall (RCSW) structure evaluated in this study
is a core wall residential building designed to comply with the requirements

ATC 119-1 Part 3: Performance Expectation for New Buildings 3-7

 outlined in Section 2.1. The structure was originally designed by Magnusson
Klemencic Associates (MKA) for a site in Los Angeles (PEER, 2011) and
later re-designed for a site in San Francisco by Tipler (2014). The structure
consists of 42 stories above grade and four basement levels. Story heights
are approximately 10 ft, resulting in an overall height of 457 ft above the
ground (including the roof bulkhead). The gravity system consists of 8-inch-
thick slabs (measuring ~108 ft × 107 ft in the superstructure plan) supported
by reinforced concrete columns with sections typically ranging from 36 in × 36
in near the base to 18 in × 18 in at the roof. The seismic-force-resisting
system consists of coupled shear walls 32 inches thick up to floor 13 and 24
inches thick from floor 13 to the roof. The coupling beams are 30 inches in
depth in the superstructure and 34 inches in the basement. Nominal
compressive strength of concrete in the core is 8.0 ksi, and nominal yield
strength of steel in the core and coupling beams is 60 ksi and 75 ksi,
respectively. Figure 3-1 shows an isometric view of the building and its
superstructure plan view. The dynamic vibration periods of the RCSW
building are summarized in Table 3-2.

(a) (b)
Figure 3-1 Reinforced concrete shear wall residential building (a) isometric view
and (b) superstructure plan view. Adapted from Almufti et al. (2018).

Table 3-2 Dynamic Vibration Periods of the RCSW Building

Period (s)
Strong Direction (Y-axis) Weak Direction (X-axis)
T1 = 4.37 T1 = 5.27
T2 = 0.93 T2 = 1.10

3-8 Part 3: Performance Expectation for New Buildings ATC 119-1

The core-only RCSW system is designed to dissipate energy through plastic
hinging at the base of the wall piers and at the ends of the coupling beams.
A 3-dimensional model of this structure, developed by Almufti et al. (2018) for
the recently completed USGS HayWired study (USGS, 2018), was used to
evaluate the response of the structure at DE and MCE by means of nonlinear
response history analysis (NLRHA) in LS-DYNA (LSTC, 2009). In this model,
the wall piers are modeled with distributed plasticity fiber beam-column
elements, with fibers representing steel reinforcement and concrete with
different levels of confinement. The coupling beams are modeled with
concentrated plasticity elements, where the hinge response is validated
against testing at the University of California Los Angeles (Naish et al., 2009)
to reproduce their hysteretic behavior. Linear elastic beam-column elements
were used to model the columns and linear elastic shell elements to model
the slabs. More details of the archetype building design and modeling
approach can be found in Tipler (2014) and Almufti et al. (2018).

2.3 Steel Buckling-Restrained Braced Frame Office Building

The steel buckling-restrained braced frame (BRBF) structure was designed

following the requirements outlined in Section 2.1. The building has 40
stories above the grade and four basement levels. Story heights are typically
13.5 feet, resulting in an overall height of 545 feet above the ground. The
gravity system consists of a composite floor slab (3.25-inch slab on 3-inch
deck) supported by simply supported steel beams and wide flange steel
columns. The seismic-force-resisting system consists of a buckling-
restrained brace (BRB) core in the center of the building and mega-brace
BRB configuration on the perimeter in the Y-direction, as shown in Figure 3-2.

(a) (b)
Figure 3-2 Buckling-restrained braced frame office building (a) isometric view
and (b) superstructure plan view.

ATC 119-1 Part 3: Performance Expectation for New Buildings 3-9

 The core areas of the BRBs of the interior frames ranges from 6 in2 to 12 in2,
and the core area of the exterior mega-brace BRBs ranges from 50 in2 to
9 in2. Most of the columns are box shaped with concrete fill. Nominal
compressive strength of concrete fill is 10 ksi, and the steel for framing and
braces are A992 and A36, with yield strengths of 50 ksi and 36 ksi,
respectively. The dynamic properties of the steel BRBF building are
summarized in Table 3-3.

Table 3-3 Dynamic Vibration Periods of the Steel BRBF

Period (s)
Strong Direction (Y-axis) Weak Direction (X-axis)
T1 = 4.08 T1 =4.53
T2 = 1.35 T2 =1.54

The steel BRBF system dissipates energy primarily by yielding of braces,

while all other elements in the system are capacity designed to remain
essentially elastic. Planar (2-dimensional) nonlinear models of each principal
building direction were developed to evaluate the response of the structure at
DE and MCE by means of NLRHA in OpenSees (McKenna et al., 2000). In
these models, frame elements were simulated using concentrated plasticity
elements, with hinge properties calibrated per the recommendations of
Lignos and Krawinkler (2010, 2011) for W-section and tubular square
columns. The truss elements used to simulate the BRBs were calibrated per
the recommendations of Terashima (2018), using BRB test data. Finite-rigid
offsets are modeled at ends of BRBs to represent the non-yielding zone, and
rigid offsets are applied to connections between BRBs and frame elements to
consider the presence of gusset plates. The planar frame models
incorporated leaning columns to represent P-∆ effects, but otherwise, the
gravity systems were not modeled.

2.4 Assessment Methodology

The RCSW residential building (Section 2.2) and steel BRBF office building
(Section 2.3) comply with seismic design requirements for two locations in
San Francisco with Site Classes D and B. The Site Class is a classification
assigned to a site based on the soils present and their engineering properties
to a depth of 100 feet. Site Class D is representative of a site with stiff soil
(with shear wave velocities in the top 100 feet of soil in the range of 600 to
1,200 ft/s), and Site Class B is representative of a rock site (with shear wave
velocities in the top 100 feet of soil in the range of 2,500 to 5,000 ft/s). Figure
3-3 illustrates the archetype building site locations superimposed with the
City’s existing tall building inventory (See Part 1). While the majority of the

3-10 Part 3: Performance Expectation for New Buildings ATC 119-1

City’s tall buildings are located on Site Class D, the comparison to Site Class
B is useful to: (1) show the variability of performance that can occur in San
Francisco; and (2) provide insight on the performance of a building that is
essentially over-designed for its site, in other words a “better than code”
building.

Figure 3-3 Site locations of archetype buildings in

downtown San Francisco.

Since soft soils tend to amplify earthquake ground motion shaking, the
location with Site Class D has significantly higher shaking intensity than the
location with Site Class B. In spite of their distinct seismic hazard, the
designs of the archetype RCSW and BRBF buildings, introduced in Sections
2.2 and 2.3 respectively, are similar for both site classes due to minimum
code-prescribed base shear requirements and other design constraints (i.e.,
wind loads that control the design of the BRBF system). To the extent that
the designs could be further optimized, the archetype buildings are slightly
overdesigned for the ground motions at Site Class B. Nevertheless, the
performance assessment for the buildings at the two sites (B and D) are
considered to be representative of what could be expected from new code-
conforming buildings at each site. In addition, for the purpose of evaluating

ATC 119-1 Part 3: Performance Expectation for New Buildings 3-11

 the possible benefits of designing buildings with more stringent seismic
design criteria (e.g., smaller drift limits, more rugged elevator systems), the
relative performance of the buildings across the two sites provides insights
into how the building performance could be improved.

To demonstrate compliance with current performance-based seismic design

requirements, building drifts and member force and deformation demands
were evaluated at MCE shaking. Two suites of eleven ground motions were
selected and linearly scaled for each of the Site Classes, B and D. Figure 3-4
shows the target MCE spectrum, individual ground motion spectra (noted
RotD100 in the figures), as well as the average ground motion suite spectra,
which closely follows the target spectrum across the period range of interest
from 1 to 8 seconds. Following ASCE/SEI 7-10, the SFBC defines DE
shaking as two-thirds of MCE. To evaluate performance at DE shaking, the
MCE target spectrum and associated ground motion suite, as illustrated in
Figure 3-4, were scaled accordingly by a factor of two-thirds. In the figures,
RotD100 denotes the maximum direction spectral acceleration, which is the
greatest unidirectional acceleration value for any possible orientation of a pair
of horizontal ground motions and forms the basis of code design spectra. A
unique ground motion identifier (per the PEER 2013 ground motion
database), the earthquake event, its magnitude and distance (from the
seismic source to the site of the recording), as well as the corresponding
scale factor applied to match the target spectrum are summarized in Tables
3-4 and 3-5 for Site Classes B and D, respectively.

The expected damage and associated economic losses in the archetype

buildings are evaluated based on the FEMA P-58 (FEMA, 2012) methodology
as implemented in the Seismic Performance Prediction Program (SP3)
(HBrisk, 2018). This method is based on a building performance model,
which is a collection of structural and nonstructural building components that
are susceptible to seismic damage. Definition of the structural components is
based on the structural design of each building, as presented in Sections 2.2
and 2.3 for the RCSW and the BRBF, respectively. The nonstructural
components are defined using the Normative Quantity Estimation Tool of
FEMA P-58, which is based on a database of typical nonstructural
components in approximately 3,000 buildings of different occupancies. The
resulting nonstructural component definitions were reviewed for consistency
with previous studies on expected seismic performance of tall buildings
similar to those under evaluation (Tipler, 2014; PEER, 2011). Certain
nonstructural components required the calculation of parameters to
characterize their seismic resistance. These were calculated automatically

3-12 Part 3: Performance Expectation for New Buildings ATC 119-1

within SP3, which follows ASCE/SEI 7-10, Chapter 13: Seismic Design
Requirements for Nonstructural Components.

(a)

(b)
Figure 3-4 Maximum Considered Earthquake (MCE) target spectrum
superimposed with the average and individual ground motion spectra
for (a) Site Class B; and (b) Site Class D locations.

ATC 119-1 Part 3: Performance Expectation for New Buildings 3-13

 Table 3-4 Site Class B Selected Ground Motions
Sequence Scale Earthquake Distance
Number Earthquake Event Factor Magnitude (km)
286 Irpinia (Italy) 1.99 6.90 21.26
1511 Chi-Chi (Taiwan) 0.74 7.62 2.74
1521 Chi-Chi (Taiwan) 1.30 7.62 9.00
1541 Chi-Chi (Taiwan) 0.97 7.62 12.38
3750 Cape Mendocino (USA) 1.02 7.01 25.91
3947 Tottori (Japan) 1.83 6.61 5.86
3964 Tottori (Japan) 2.15 6.61 11.29
4852 Chuetsu-oki (Japan) 4.00 6.80 32.54
4882 Chuetsu-oki (Japan) 2.04 6.80 23.44
6928 Darfield (New Zealand) 2.32 7.00 25.67
8166 Duzce (Turkey) 2.10 7.14 3.58

Table 3-5 Site Class D Selected Ground Motions

Sequence Scale Earthquake Distance
Number Earthquake Event Factor Magnitude (km)
139 Tabas (Iran) 4.00 7.35 13.94
143 Tabas (Iran) 0.81 7.35 2.05
286 Irpinia (Italy) 4.00 6.90 21.26
1165 Kocaeli (Turkey) 3.12 7.51 7.21
1511 Chi-Chi (Taiwan) 1.64 7.62 2.74
1521 Chi-Chi (Taiwan) 2.75 7.62 9.00
1541 Chi-Chi (Taiwan) 3.18 7.62 12.38
3750 Cape Mendocino (USA) 2.49 7.01 25.91
4848 Chuetsu-oki (Japan) 3.51 6.80 17.93
4882 Chuetsu-oki (Japan) 4.00 6.80 23.44
8166 Duzce (Turkey) 4.00 7.14 3.58

In the FEMA P-58 methodology, performance estimates of the structural and

nonstructural building components are mathematically represented with
fragility functions. A fragility function is a statistical distribution that indicates
the conditional probability of incurring a damage state at a given value of
demand. From each damage state, the associated repair costs and times are
estimated by means of consequence functions. In addition to the family of
fragilities recommended by the Normative Quantity Estimation Tool, user
defined components were implemented to more accurately characterize
damage to elevators, mega braces, and façades.

3-14 Part 3: Performance Expectation for New Buildings ATC 119-1

The fragility function provided in FEMA P-58 to predict damage to the
elevator cabin is a function of peak ground acceleration developed based on
a review of elevator performance in low- and mid-rise buildings. To better
represent conditions in tall buildings, an additional fragility function was
introduced to estimate elevator shaft rail damage as a function of residual
story drift as outlined in Almufti et al. (2018). Residual story drift ratios were
derived from peak transient story drifts as outlined in Appendix C of FEMA
P-58.

User-defined fragility functions were developed to predict damage to the

interior BRBs (Almeter et al., 2018) and exterior mega BRBs (Hooper, 2018)
The latter functions are intended to recognize the differences between BRB
mega-braces and single story BRBs for which fragility functions in FEMA
P-58 were originally developed.

The parameters required to define fragility functions for the curtain wall
system are based on those outlined in Almufti et al. (2018), since the
standard curtain wall functions of FEMA P-58 where developed for low- and
mid-rise construction. While damage state definitions and consequence data
are consistent with that provided by FEMA, the drift values associated with
each damage state are defined in accordance with the performance-based
building design requirements.

The input demands to fragility functions are generally referred to as

engineering demand parameters (EDP), and are obtained from analysis.
Different EDPs are used for different components, depending on their ability
to predict damage (e.g., damage to acceleration-sensitive components can
be estimated by peak floor accelerations).

The following EDPs are used to evaluate performance in the RCSW

residential building: story drifts, residual drift, damageable wall drift, racking
drift, coupling beam rotation, and floor acceleration. Table 3-6 summarizes
the maximum value (throughout the building height) of the average set of
results (from the eleven ground motions) in each story for each of the EDPs
considered in the performance evaluation of the RCSW in Site Class B and D
under DE and MCE. There is a subtle difference in the way the residual drifts
are computed. The residual drifts noted in Table 3-6 are calculated by
obtaining the maximum residual drift (throughout the building height) in a
single ground motion simulation and then taking the average from the eleven
ground motions (a single residual drift parameter is input into the loss model
for each ground motion in each building direction). Figure 3-5 illustrates the
distribution of maximum transient story drifts, racking story drifts, and floor
accelerations up the building height.

ATC 119-1 Part 3: Performance Expectation for New Buildings 3-15

 Table 3-6 Engineering Demand Parameters for Evaluation of the Reinforced Concrete
Shear Wall Residential Building
Engineering Ground Motion Mean of Max Mean of Max
Demand Parameters Site Class Intensity x-Direction y-Direction
DE 0.82 % 0.77 %
D
MCE 1.18 % 1.14 %
Transient Story Drift
DE 0.42 % 0.34 %
B
MCE 0.58 % 0.49 %
DE 0.10 % 0.09 %
D
MCE 0.24 % 0.24 %
Residual Story Drift
DE 0.00% 0.00%
B
MCE 0.04 % 0.02 %
DE 0.21 % 0.24 %
D
MCE 0.64 % 0.66 %
Damageable Wall Drift
DE 0.09 % 0.11 %
B
MCE 0.14 % 0.17 %
DE 1.27 % 1.16 %
D
MCE 1.71 % 1.66 %
Racking Story Drift
DE 0.78 % 0.64 %
B
MCE 1.01 % 0.86 %
DE 0.012 rad 0.012 rad
D
MCE 0.020 rad 0.019 rad
Coupling Beam Rotation
DE 0.006 rad 0.005 rad
B
MCE 0.009 rad 0.008 rad
DE 1.60 g 1.68 g
D
MCE 2.31 g 2.24 g
Peak Floor Acceleration
DE 1.14 g 0.88 g
B
MCE 1.62 g 1.31 g

The following EDPs are used to evaluate performance in the BRBF office
building: transient story drift, residual story drift, and floor acceleration. Table
3-7 summarizes the maximum value (throughout the building height) of the
average set of results (from the eleven ground motions) for each of the EDPs
for the two sites (Site Class B and D) and earthquake intensities (DE and
MCE). As with the RCSW, there is a subtle difference in the way the residual
drifts are computed. The residual drifts noted in Table 3-7 are calculated by
obtaining the maximum residual drift (throughout the building height) in a
single ground motion simulation and then taking the average from the eleven
ground motions (a single residual drift parameter is input into the loss model
for each ground motion in each building direction). Figure 3-6 illustrates the
responses of the BRBF for the DE.

3-16 Part 3: Performance Expectation for New Buildings ATC 119-1

 (a)

(b)

(c)
Figure 3-5 Earthquake demands from non-linear response history
analysis of the reinforced concrete shear wall residential
building under the design earthquake: (a) transient story
drifts; (b) racking story drifts; and (c) floor accelerations.

ATC 119-1 Part 3: Performance Expectation for New Buildings 3-17

 Table 3-7 Maximum Engineering Demand Parameters for Evaluation of the Steel
Buckling-Restrained Braced Frame Office Building
Engineering Demand Ground Motion Mean of Max Mean of Max
Parameters Site Class Intensity x-Direction y-Direction
DE 0.90 % 1.01 %
D
MCE 1.82 % 1.28 %
Transient Story Drift
DE 0.45 % 0.62 %
B
MCE 0.58 % 0.75 %
DE 0.05 % 0.06 %
D
MCE 0.40 % 0.14 %
Residual Story Drift
DE 0.00 % 0.00 %
B
MCE 0.00 % 0.01 %
DE 0.71 g 1.12 g
D
MCE 1.06 g 1.67 g
Floor Acceleration
DE 0.56 g 0.93 g
B
MCE 0.78 g 1.33 g

The methodology assigns a repair class to each damage state of the

structural and nonstructural building components. The repair classes identify
damage levels that hinder achieving a certain recovery state, where the
building recovery is predicated on all the components achieving the
associated recovery state. The repair class associated with each damage
state for all components in the building are based on Almufti et al. (2018).

The methodology also considers a streamlined sequence of repairs on a

floor-per-floor basis. The necessary repair activities are grouped by trade,
including structural, façade (exterior partitions and cladding), egress (stairs
and elevators), MEP, and office fitouts. Once structural repairs are complete,
other repair activities can be carried out in parallel with limits to the number of
workers per trade and the total number of workers on site.

In addition to repairs, the methodology identifies a series of impeding factors

that may delay the initiation of repair work, such as post-earthquake
inspection, engineering mobilization, contractor mobilization, financing,
permitting and long-lead-time components. These impeding factors are
grouped into three delay sequences, as illustrated in Figure 3-7, the longest
of which controls the impeding factor delays. All delay sequences begin with
post-earthquake inspection. Following post-earthquake inspection, the first
sequence of delays considers time to mobilize an engineer, to carry out a
detailed structural evaluation and any necessary design work, as well as
delays associated with permitting such work. The second sequence

3-18 Part 3: Performance Expectation for New Buildings ATC 119-1

considers the time associated with the mobilization of a contractor to carry
out repair work as well as the procurement of specialty items (i.e., long-lead
components). The last sequence relates to the financing of repair work. The
median impeding factor delay estimates for each impeding factor assumed in
this study are summarized in Table 3-8. Note that in addition to the baseline
case, a recovery-planning option is considered to assess the benefits of
implementing measures to minimize different impeding factor delays.

(a)

(b)
Figure 3-6 Earthquake demands from non-linear response history analysis of
the steel buckling-restrained braced frame building under the design
earthquake: (a) transient story drifts and (b) floor accelerations.

ATC 119-1 Part 3: Performance Expectation for New Buildings 3-19

 Non-linear Response History Analysis

Engineering Demand Parameters

Long Lead
Permitting Components
Component Fragility Function

Engineering Contractor
Consequence Repair Repair Financing
Mobilization Mobilization
Function Time Class

Repair
Cost Impeding Factors Post-earthquake Inspection

LOSSES Repair Sequence Structure

Utility Disruption
Façade MEP Egress Office
DOWNTIME
DOWNTIME Fitouts
DOWNTIME
FUCTIONAL
FUNCTIONAL
RE-OCCUPANCY
RECOVERY
RECOVERY

Figure 3-7 Loss and downtime assessment methodology. Adapted from Molina
Hutt et al. (2016).

Table 3-8 Median Impeding Factor Delay Estimates for the Baseline Case and
with Recovery-Planning Measures in Place
Impeding Factor Delays Baseline Case Recovery Planning
Inspection 5 days 1 day
Engineering Mobilization
[6/12] weeks [2/4] weeks
weighted by [cosmetic/significant] damage
Contractor Mobilization
[28/40] weeks [3/7] weeks
weighted by [cosmetic/significant] damage
Permitting
[1/8] weeks [0.5/4] weeks
binary for [cosmetic/significant] damage
Financing
15 weeks 1 week
[only triggered for >10% costs]

The overall loss and downtime assessment methodology is graphically

illustrated in Figure 3-7. The total building downtime is calculated as the sum
of: (1) the time to repair damaged components that hinder achieving a
recovery state; and (2) the longest sequence of impeding factor delays that
control the overall downtime. Apart from repairs to the building itself,
important contributors to the functional recovery time of the building are
disruptions in water supply, electricity or natural gas distribution systems.
However, delays associated with utility disruption or long-lead components
are not evaluated in this study.

3-20 Part 3: Performance Expectation for New Buildings ATC 119-1

While the downtime assessment methodology outlined in REDi (Arup, 2013)
was used as the basis to evaluate the time to functional recovery in this
study, a number of modifications were introduced where limitations to the
method were identified in its application to tall buildings. Key changes are as
follows:
• Mobilization Estimates. Impeding factor delays associated with
engineer and contractor mobilization are calculated by weighting the
expected delays according to the percentage of components that
experience minor (cosmetic) damage versus more severe damage that
would prevent functionality. This change is introduced to scenarios where
severe damage to a single component would otherwise trigger unrealistic
impeding factors (e.g., one coupling beam in the RCSW building
triggering 12 weeks of impeding time for engineering mobilization and 40
weeks for contractor mobilization).
• Financing. The methodology requires conducting realizations, where
each realization represents one possible performance outcome for the
building considering a single combination of possible values of each
variable considered. For realizations where the loss ratio is less than
10% of building replacement cost, it is assumed that building owners
would have sufficient capital to cover repair costs, as opposed to
extensive delays associated with financing (on the order of 15 weeks).

Because there are many factors that can affect performance, such as
intensity of ground shaking, building construction quality, building response,
or vulnerability of contents, there is significant uncertainty in the predicted
performance of the building. This uncertainty can be accounted for by means
of a Monte Carlo simulation by conducting thousands of realizations.
Following this process, results can be expressed as a performance function
(i.e., probability of losses or downtime being less than or equal to a specified
value as a result of an earthquake). The results presented in Section 2.5
focus on mean, median, and 90th percentile estimates of response.

2.5 Loss and Downtime Results

The methods discussed in Section 2.4 to estimate building losses were used
to evaluate the expected performance of the RCSW residential and steel
BRBF office buildings across the two sites, Site Class B and D, and two
ground motion shaking intensities, DE and MCE. Table 3-9 provides a
summary of the median loss results associated with each archetype building,
Site Class and intensity level. The maximum story drift value (throughout the
building height) of the average set of results (from the eleven ground
motions) is also shown to gage the relationship between drift demands and

ATC 119-1 Part 3: Performance Expectation for New Buildings 3-21

 expected losses. Loss results are normalized over total building replacement
costs, which are estimated at $215 million ($312/square feet) for the RCSW
residential building and $280 million ($341/square feet) for the BRBF office
building. The results indicate that expected losses in the BRBF are
considerably lower than in the RCSW. For instance, at Site Class D under
DE shaking, the expected losses in the RCSW building are 8.1% versus 2.8%
in the BRBF building. While these differences are due in part to the
differences in construction costs, the expected absolute losses in the
residential building are still approximately twice those of the office building, as
illustrated in Figure 3-8.

Table 3-9 Summary of Median Loss Results and Associated Drifts

Median Loss
Drift* (% story height) (% replacement cost)
Archetype Site
Building Class DE MCE DE MCE
B 0.4 (0.8) 0.6 (1.0) 2.8 4.9
RCSW
D 0.8 (1.3) 1.2 (1.7) 8.1 15.5
B 0.6 0.8 0.9 1.8
BRBF
D 1.0 1.8 2.8 5.8
* Racking drifts shown in paranthesis.

Figure 3-8 Summary of median and 90th percentile loss results (loss ratio and
absolute losses) under DE shaking.

The loss methodology enables deaggregating losses into their constituent

parts, which can help inform loss mitigation strategies. Figure 3-9 illustrates
the average contribution of different building components to the overall
expected loss under DE shaking. Losses in the RCSW building are

3-22 Part 3: Performance Expectation for New Buildings ATC 119-1

dominated by damage to structural components and interior finishes,
particularly those components that are sensitive to racking drifts, which are
amplified by about 1.5 to 2 times over story drifts. Racking drift deformations
occur in the RCSW structure due to differences in axial deformation
(elongation) between the concrete walls and the gravity framing. The
amplified racking cause damage and losses associated with interior partition
wall finishes and slab-to-column connections. Losses in the steel BRBF,
which are about half of those for the RCSW, are dominated by damage to
cladding components (minor damage primarily associated with gasket seal
failure), plumbing, and HVAC. Note that while elevator damage is a relatively
small contributor to damage in both buildings, it is a more significant
contributor to downtime.

RCSW: Site D – DE BRBF: Site D – DE

Mean Losses: 10% Mean Losses: 3%
Figure 3-9 Contribution of different building components to overall loss at DE for
Site Class D.

Figure 3-9 shows that residual drift contributes to the overall losses in the
RCSW, but not in the steel BRBF. While the probability of residual drift
rendering the building irreparable for the RCSW building in Site Class D
under DE is less than 1%, because such a realization results in total building
replacement cost, the contribution to the mean loss statistics, as seen in
Figure 3-9, is significant.

Collapse realizations would also represent a total loss. However, based on

the nonlinear response history analysis results under MCE, the probability of
collapse under the ground motion shaking intensities considered is assumed
to be zero for both buildings and site classes considered.

The repair costs represented in Figures 3-8 and 3-9 are one type of direct
loss, downtime is another. The time needed to achieve reoccupancy and
make the repairs needed for functional recovery is good measure of the
impact of tall buildings on the recovery of San Francisco and its downtown
neighborhoods. The median repair times (for components that inhibit

ATC 119-1 Part 3: Performance Expectation for New Buildings 3-23

 functional recovery), and functional recovery times (which include delays
associated with different impeding factors) are summarized in Table 3-10.

Table 3-10 Summary of Median Repair Times and Functional Recovery

Time
Median Time to Reduced Median
Median Repair Functional Time to Functional
Time (days) Recovery (days) Recovery (days)
Archetype Site
Building Class DE MCE DE MCE DE MCE
B 24 34 113 154 52 80
RCSW
D 50 280 194 464 129 408
B 13 20 49 101 21 43
BRBF
D 29 118 134 245 68 171

Baseline Case Recovery Planning

Figure 3-10 Summary of median and 90th percentile functional recovery time
considering the baseline case and recovery planning under DE
shaking.

As evident from Table 3-10, in spite of moderate economic loss (repair cost)
levels, the associated repair and downtimes are extensive. For instance, at
Site Class D under DE shaking, the median losses in the RCSW residential
building of 8.1% result in repair times of 1.5 months, with functional recovery
extending to 6.5 months. Similarly, at Site Class D under DE shaking, the
expected losses in the BRBF office building of 2.8% corresponds to a repair
time of 1 month and a time to functional recovery of 4.5 months.

3-24 Part 3: Performance Expectation for New Buildings ATC 119-1

Table 3-10 shows significantly shorter repair and recovery times at Site Class
B than at Site Class D. As discussed above, the Site Class B results can be
understood as estimates of Site Class D performance of a “better than code”
building, specifically one designed with tighter drift limits. Another way to
reduce downtime is to incorporate recovery considerations into the initial
design. The benefits of designing with recovery in mind are shown in the two
right-hand columns of Table 3-10., labeled, “Reduced median time to
functional recovery.” Recovery-focused measures can include: (1) expediting
post-earthquake inspection; (2) pre-earthquake arrangements to have an
engineer on contract to minimize delays associated with engineering
mobilization; (3) pre-earthquake arrangements to have a general contractor
on retainer to minimize delays associated with contractor mobilization; and
(4) expediting permits for building repairs. Schemes (1) and (2) already exist,
through the Building Occupancy Resumption Program (BORP), and through
arrangements that building owners frequently hold with engineers to carry out
certain activities as needed. See Part 6 for further discussion on BORP.
Schemes (3) and (4), while not routinely available, are evaluated to explore
their impact on expediting recovery. The impact of these measures on the
median recovery times are summarized in the last two columns of Table 3-10,
and the median and 90th percentile results are illustrated in Figure 3-10 for
both buildings under DE shaking. To the extent that effective pre-earthquake
plans can be implemented and carried out, they can significantly reduce the
building downtimes—potentially cutting downtime in half for many cases.

Although deaggregating the contributors to downtime is not as straightforward

to deaggregating losses, it is possible to identify the major contributions to
repair and functional recovery times. As described previously, depending on
the component, the influence of component damage and repairs on downtime
is amplified when impeding factors are triggered, so the relationship between
repair and functional recovery times is not one-to-one.

For the RCSW building subjected to DE ground motions at Site Class D, the
major controlling contributors to functional repairs and downtime are:
elevators, structural repairs, and mechanical equipment. For the BRBF
building subjected to DE ground motions at Site Class D, the major
controlling contributors to the repair and functional recovery times are:
elevators, curtain wall, and mechanical equipment.

Figures 3-11 and 3-12 summarize median and 90th percentile estimates of
functional recovery time, which includes the impeding factor estimates
previously discussed in Section 2.4, such as post-earthquake inspection,
engineering mobilization, contractor mobilization, permitting and financing for
the RCSW and BRBF buildings. The figures also illustrate the different

ATC 119-1 Part 3: Performance Expectation for New Buildings 3-25

 sequences of delays, which are quantified in the recovery “paths,” termed
engineering, contractor, and financing.

with Recovery Planning

Figure 3-11 RCSW functional recovery time deaggregation at Site Class D under
DE: baseline (top) and with recovery planning (bottom).

with Recovery Planning

Figure 3-12 Steel BRB functional recovery time deaggregation at Site Class D
under DE: baseline (top) and with recovery planning (bottom).

All three recovery paths begin with the time for inspection. Beyond this, the
engineering path includes time for mobilizing engineering services and
permitting of the proposed repairs; the contractor path includes contractor
mobilization and procurement of replacement items with long lead times; and
the financing path includes time to arrange financing of repairs. The total
downtime estimate is essentially the combination of the longest of these three

3-26 Part 3: Performance Expectation for New Buildings ATC 119-1

recovery paths plus the repair time itself. Note that the median values of the
recovery paths for the total downtime are close to but not equal to simple
additions of the constituent components, since the combined totals are
statistical realizations of the medians that incorporate variability of the
component parts. For the RCSW building (Figure 3-11), the longest
sequence of delays comes on the engineering path, due to engineering
mobilization and permitting. Comparing the upper and lower plots,
implementation of recovery-planning strategies can reduce functional
recovery time from approximately 6.5 months (194 days) to just over 4
months (128 days). Similar trends are seen for the BRBF building (Figure
3-12), where the engineering path is the longest sequence and
implementation of resilience strategies can reduce functional recovery time
from 4.5 months (134 days) to just over 2 months (68 days). The functional
recovery time estimates in Figures 3-11 and 3-12 are for realizations that do
not require total building replacement in the event of demolition due to
excessive residual drifts. The probability of requiring demolition due to
excessive residual drifts is zero in all cases except for the RCSW at Site
Class D, with a 0.9% probability under DE and a 2.4% under MCE. In these
cases, the downtime is assumed to be the building replacement time of 3
years.

Note that the functional recovery time estimates in Figure 3-11 do not include
realizations where residual drifts render the building irreparable, which have a
0.9% probability of occurrence at Site Class D under DE. Note that the
functional recovery time estimates in Figure 3-12 do not include realizations
where residual drifts render the building irreparable, which have a 0.0%
probability of occurrence at Site Class D under DE.

2.6 Variations in Archetype Tall Building Design Criteria

To evaluate how story drift demands can affect the damage, repair cost loss,
and downtimes, sensitivity analyses were conducted for varying story drift
demands. This was done by scaling the story drifts from NLRHA, as
illustrated in Section 2.4, based on maximum of mean story drift ratios
ranging from 1% to 3% in 0.5% increments, and re-running the SP3 loss
analyses. In the re-evaluation of loss and downtime, the distributions of peak
ground accelerations and floor accelerations are assumed to remain
consistent with the baseline case, since the accelerations are assumed to be
less sensitive to changes in the building stiffness. Thus, only repair costs and
repair time associated with drift sensitive components varied, whereas those
associated with acceleration sensitive components remained unchanged.

ATC 119-1 Part 3: Performance Expectation for New Buildings 3-27

 Figures 3-13, 3-14, and 3-15 illustrate the results of the sensitivity analyses
for cost, repair time, and functional recovery time. The 90th percentile values
are included in addition to the median values to indicate how the risk of large
losses increases with increasing mean drift demands. The results suggest
that to reduce the risk of large expected losses, the mean drift demands in
the steel BRBF should be limited to about 2%. This is based on the dramatic
increase in the trend in 90th percentile losses at the 2% story drift demand.
Similarly, results suggest that the mean drift demands in the RCSW would
need to be limited to about 1%.

The smaller limit for the RCSW system is due to two factors. First, due to
wall elongation associated with flexural cracking, shear wall structures
experience racking drifts that can double the damaging effects of story drifts,
and thus, the RCSW story drifts of 1% can impose damaging racking drifts on
the order of 2%. Second, structural damage tends to be experienced sooner
in RCSW structures due to the larger localized deformation in coupling
beams and wall-floor slab connections.

Figure 3-13 Mean story drift demand versus repair cost losses.

Figure 3-14 Mean story drift demand versus repair times.

3-28 Part 3: Performance Expectation for New Buildings ATC 119-1

Figure 3-15 Mean story drift demand versus functional recovery times.

ATC 119-1 Part 3: Performance Expectation for New Buildings 3-29

Chapter 3

Other Tall Building Studies

3.1 Recent Tall Building Projects

In addition to evaluating the archetype residential and office buildings

presented in Chapter 2, the project team collected information from over 30
recently completed performance-based seismic design tall building projects in
San Francisco, Los Angeles, Seattle, San Diego, and other locations. The
purpose of collecting this information was twofold: (1) to benchmark the
design of the archetype buildings against real project data; and (2) to
understand how modifications to drift limits would affect the state of practice.
Figure 3-16 shows the mean story drift demands for MCE evaluations in the
principle directions for the 30 buildings.

Figure 3-16 Mean MCE-level story drift demands from recently completed tall buildings
designed using performance-based methods.

These results suggest that despite the 3% drift limit at MCE, currently
prescribed by PEER TBI guidelines, about two-thirds of the buildings have
maximum drifts less than 2% and half have drifts less than 1.5%. These data
indicate that the two archetype buildings are fairly representative of the
average stock of buildings, and further, that limiting MCE drifts to values less
than the current 3% limit is not inconsistent with current design practice.

3.2 Research Publications

A literature review was carried out to compare the performance results from
this study against those of similar tall building research projects. Data were
collected primarily from the following studies: PEER (2011), Tipler (2014),
and Jayaram et al. (2012), including loss results for buildings consistent with
the steel buckling-restrained braced frame (BRBF) and reinforced concrete

ATC 119-1 Part 3: Performance Expectation of New Buildings 3-31

 shear wall (RCSW) archetype buildings presented in Chapter 2, as well as for
dual systems including a reinforced concrete core wall and perimeter
reinforced concrete frame, and steel moment resisting frames. Expected
losses at Design Earthquake (DE) and Maximum Considered Earthquake
(MCE) levels found in the literature, as well as those previously reported for
the archetype buildings, are shown in Figure 3-17.

(a)

(b)
Figure 3-17 Loss versus drift data for different structural systems
at (a) DE and (b) MCE ground motion intensities.

The data shown in Figure 3-17 indicate that the loss estimates for the
archetype buildings in the current study, while at the lower range of estimates
from the other studies, are generally consistent with those from past studies.

3-32 Part 3: Performance Expectation of New Buildings ATC 119-1

The data also confirm a clear trend between drift demand and expected
losses, suggesting that losses are about 10% of building replacement value
for story drift demands of 1%.

3.3 Mid-Rise Building Loss Studies

To further investigate the trends identified in Section 2.6, FEMA P-58

Performance Estimate Tool (PET) (FEMA, 2018) was utilized to evaluate the
losses and repair times for a 12-story steel BRBF office building and 12-story
RCSW building. Note that FEMA P-58 PET does not include data for
buildings taller than 12-stories, and the building systems considered in FEMA
P-58 PET do not include the same features as the two archetype buildings
(e.g., the RCSW in the study consists of a coupled shear wall building,
whereas the RCSW in FEMA P-58 PET consists of a system with single
shear walls). Nevertheless, the FEMA P-58 PET results can help
substantiate the trends identified in Section 2.6.

Loss and downtime results of the FEMA P-58 PET analyses are summarized
for the 12-story BRBF and RCSW systems in Figures 3-18 and 3-19,
respectively. In these analyses, losses and downtime are calculated for
alternate building designs that were designed to have DE drift demands of
0.5% to 2%. Therefore, in contrast to the plots from the archetype study
(Section 2.6), in these plots, the horizontal axes correspond to the DE drift
limits used in the design, rather than the calculated drift demands. To the
extent that the drift demands are consistent with the drift limits, then the loss
results for the DE shaking (blue lines in the figures) should be roughly
consistent with the drift limits. On the other hand, the larger losses for MCE
drifts would presumably be related to larger drift demands, as compared to
the DE design drift limit.

In the FEMA P-58 PET analyses, the repair cost losses and downtimes seem
remarkably low for the 12-story RCSW building, as compared to both the
BRBF building and the 42-story RCSW archetype building presented in
Chapter 2. On the other hand, the results for the 12-story BRBF building are
comparable to those shown previously for the 40-story BRBF archetype of
Chapter 2.

Similar to the trends observed previously with the 40-story BRBF, the
12-story FEMA P-58 PET BRBF design shows a distinct increase in losses
for buildings with design drift limits above about 1.3%. This is about two-
thirds of the 2% DE limit specified in current building codes. Assuming a
linear relationship between DE and MCE drifts, the two-thirds reduction would
correspond to an MCE drift limit of 2%, instead of the 3% permitted by current

ATC 119-1 Part 3: Performance Expectation of New Buildings 3-33

 codes. While the overall losses for the 12-story FEMA P-58 PET RCSW are
relatively small, they similarly show an increase at design drifts of about
1.3%.

Figure 3-18 FEMA P-58 PET drift versus loss sensitivity results.

Figure 3-19 FEMA P-58 PET drift versus repair time sensitivity results.

3-34 Part 3: Performance Expectation of New Buildings ATC 119-1

Chapter 4

Summary, Recommendations, and

Cost Implications

4.1 Summary

Current design requirements for tall buildings in San Francisco are primarily
intended to protect life safety by minimizing the risk of collapse or significant
structural or façade damage under extreme (i.e., Maximum Considered
Earthquake) ground motions. While the requirements also include
serviceability requirements to limit damage to the building structural system,
this check is made under frequent earthquake ground shaking (with a 50%
chance of exceedance in 30 years), whose intensity is much lower than that
of the Design Earthquake (DE) ground motions (with roughly a 10% chance
of exceedance in 50 years).

One of the key design criteria that controls the level of building damage under
earthquake ground motions is the story drift ratio. Current building codes limit
the maximum story drift ratios to 3% under MCE ground motions, which is
roughly equivalent to story drift ratios of 2% under DE ground motions. While
these are the maximum story drifts permitted by the building code, the drift
ratios of many new tall buildings are significantly less than this limit due to
other design constraints and considerations. For example, a survey of
recently constructed buildings in San Francisco, Los Angeles, Seattle and
San Diego reveals their MCE-level drift ratios to range from 1% to 3%, with
typical values between 1.5% to 2% (see Figure 3-16). For comparison, the
drift levels under DE ground motions would typically be about two-thirds of
the MCE level drifts; hence the surveyed MCE drift ratios of 1% to 3% would
relate to DE drift ratios of 0.7% to 2%.

Studies of two archetype tall buildings designed to meet current San

Francisco Building Code requirements indicate that both would experience
functional recovery times significantly longer than the tentative recommended
recovery goals for occupancies common in San Francisco tall buildings (ATC,
2018). Those goals call for a functional recovery time of “weeks” for hotels,
major employers, and most multi-family residential buildings, while allowing
“months” for other business uses typical in office towers.

ATC 119-1 Part 3: Performance Expectation for New Buildings 3-35

 A performance evaluation for a 42-story concrete shear wall residential
building indicates that damage resulting from DE ground motions will take on
the order of 1.5 months to repair, with a functional recovery time extending to
6.5 months, including time for inspections, mobilization of engineering and
contractor services, and permitting. By employing Building Occupancy
Resumption Program (BORP) and other measures to reduce the impeding
factors, estimates suggest that the recovery times can be reduced to about 4
months, which is still longer than the City’s tentative recovery goal of “weeks”
for multi-family residential buildings (ATC, 2018). These long repair and
recovery times are in spite of the fact that this building is estimated to
experience moderate story drift ratios of 0.8%, which is less than half the
building code limit, although wall elongation creates amplified racking drifts of
about 1.3%. The calculated repair costs of about 8% of the building
replacement value are not out of line with expectations for standard repair
cost loss analyses.

Damage and downtime analyses of the 40-story steel braced-frame office

building yield slightly better performance, with an estimated repair time of 1
month and functional recovery times of 2.5 to 4.5 months, depending on
measures to reduce impeding factors. This office building has estimated
story drifts of 1%, again about half of the maximum permitted by the building
code, and repair costs equal to about 3% of the building replacement value.

Studies of these same two buildings for sites with lower ground motions and
lower drifts indicate that the repair and recovery times would significantly
reduce with tighter story drift limits. For example, for the residential building,
the repair and functional recovery times would reduce to about 3/4 month and
2.5 months, respectively, by limiting the building story drifts to 0.4% (about
20% of the code limit) and employing measures to reduce impeding factors.
For the office building, the repair and functional recovery times would reduce
to about 1/2 month and 3/4 month, respectively, by limiting the building story
drifts to 0.6% (about 30% of the code limit). These are approaching San
Francisco’s tentative recommended recovery goals, although the economics
and practicality of achieving these tight drift limits is unclear for sites with high
ground motions (e.g., sites with Site Class D).

Parametric studies to examine the performance of buildings with drifts larger

than those observed in the archetype studies indicate that story drift ratios up
to the building code limit of 2% under DE ground motions would lead to
extended downtimes on the order of a year or more.

3-36 Part 3: Performance Expectation for New Buildings ATC 119-1

4.2 Recommendations

Available evidence indicates that tall buildings designed per current minimum
building code requirements may experience earthquake damage that will
result in functional recovery times significantly in excess of San Francisco’s
tentative recommended recovery goals under DE ground motions.

Fully achieving the recommended functional recovery goal for functional

recovery of “weeks” for multi-unit residential tall buildings is not economical
with standard construction technologies, even when measures are included
to reduce impeding factors. (The goal could potentially be achieved with
seismic isolation or supplemental damping, but this is beyond the scope of
the current investigation).

The target of 1 month to functional recovery may be achievable by limiting

story drift ratios of tall buildings under the DE ground motions to less than
1%. Assuming that DE drift demands are roughly two-thirds of MCE drift
demands shown in Figure 3-16, many recently completed tall buildings come
close to meeting this reduced drift demand. In addition to the reduction in
seismic drift limits, achieving the 1 month to functional recovery goal would
require: (1) enhanced design and specification of critical MEP/elevator
systems; and (2) implementation of BORP and other measures to mitigate
impeding factors for recovery. With regard to the latter, Part 6 of this report
and Lang et al. (2018) outline suggestions for improving the BORP program
to better achieve functional recovery.

4.3 Cost Implications

Although there are cost premiums associated with the different strategies
proposed in this study to achieve enhanced seismic performance, in the long
term, these initial costs result in long term savings associated with the
reduction in economic losses and downtime. Studies such as Tipler (2014)
suggest benefit-to-cost ratios on the order of two to three when considering
direct losses only, and five to twenty when considering both direct losses and
downtime.

The cost of pursuing an enhanced design can vary significantly whether

these are achieved through enhanced structural design (e.g., thicker walls or
supplemental damping devices) or enhanced nonstructural components (e.g.,
seismic detailing to accommodate larger deformations and/or accelerations
prior to damage). PEER (2011) evaluated the costs associated with
performance-based versus code-based design of tall buildings (consistent
with the RCSW and BRBF in this study) and estimated construction cost
premiums for the performance-based designs of less than 2% of the building

ATC 119-1 Part 3: Performance Expectation for New Buildings 3-37

 construction cost. A study by NIST (2013) showed that the benefits
associated with improved seismic design can be significant and cost-
beneficial (though that study was about areas of moderate study and not
specifically about tall buildings). Tipler (2014) also estimated the construction
costs of an archetype building consistent with the RCSW in this study, as well
as variations in design that included enhanced structural solutions (i.e.,
damped outriggers and base isolation) along with enhanced nonstructural
components. The study estimated premiums within 3% of the baseline
construction costs. This is consistent with a report by Almufti et al. (2016)
which noted that premiums associated with enhanced structural designs
generally range from 0 to 5%.

The costs associated with participation in programs such as BORP are

estimated at $30,000 to $50,000 as a one-time fee to prepare the required
inspection plan (BORP Submittal Phase). This fee is likely to be less if the
report is prepared in conjunction with the design of a new building. In the
event that emergency inspection services are needed following an
earthquake, these would generally be billed separately under a time and
material pricing model. The cost of having an engineering design firm on
retainer to minimize engineering mobilization details is estimated at $5,000 to
$10,000 per year. (These estimates were derived in consultation with design
firms that provide such services.)

The costs associated with other strategies such as having a contractor on

retainer or earthquake insurance are not easily estimated. Having a
contractor on hold to carry out specific repair activities, as needed, is not
generally a service offered by general contractors. Furthermore, the retainer
costs for contractors could vary significantly depending on the nature of the
repair work (e.g., light partition wall damage versus severe structural
damage). Similarly, while earthquake insurance is available to building
owners, developing transparent estimates of earthquake insurance premiums
is difficult because: (1) earthquake insurance is often lumped with other
perils; (2) owners do not necessarily insure their entire asset value; and
(3) premiums are derived based on portfolio level assessments (not individual
buildings). As a result, it is difficult to disaggregate the corresponding costs
associated with earthquake insurance alone.

While not many, there are examples of modern tall buildings that have
adopted some of the strategies presented in this study. For instance, the 181
Fremont Building, located in downtown San Francisco next to the Transbay
Transit Center, was reportedly designed to exceed the minimum seismic
performance objectives of the San Francisco Building Code at “little-to-no
cost premium” (Almufti et al., 2016). The building was designed to achieve

3-38 Part 3: Performance Expectation for New Buildings ATC 119-1

immediate occupancy and limited disruption to functionality after a 475-year
return period ground motion (approximately equal to the DE earthquake in
this study). This was achieved by means of different strategies, many of
which align with those presented in this study:
• Structural Design: Elements to remain essentially elastic under a DE
earthquake.
• Nonstructural Design: Enhanced component design including elevators
(consistent with requirements for California hospitals), staircase (to
accommodate more movement and sustain less damage), façade system
(air-tight and water-tight up to drift limits of 2%), and enhanced anchorage
of nonstructural components.
• Recovery Planning: Participation in the BORP program, inclusion of
back-up system for the building to remain functional in the event of utility
disruption, training of personnel to re-start elevators after an earthquake.
and the development of an owner’s guideline to earthquake resilience.

ATC 119-1 Part 3: Performance Expectation for New Buildings 3-39

Appendix A

References

Almeter, E., Wade, K.F., Haselton, C.B., Saxey, B., 2018, “Rapid risk
assessment of buckling restrained braced frames, with focus on
residual drifts,” Proceedings, 11th National Conference on Earthquake
Engineering, Earthquake Engineering Research Institute, Los
Angeles, California.
Almufti, I., Krolicki, J., and Crowther, A., 2016, “The resilience-based design
of 181 Fremont Tower,” Structure Magazine, June, pp. 42-46.
Almufti, I., Molina Hutt, C., Mieler, M. W., Paul, N.A., and Fusco, C.R., 2018,
The HayWired Earthquake Scenario—Engineering Implications: Case
Studies of Tall-Building Structural Analyses, Downtime and Loss
Assessment for the HayWired Scenario Mainshock, US Geological
Survey, Menlo Park, California.
Arup, 2013, Resilience-based Earthquake Design Initiative (REDi) for the
Next Generation of Buildings, Arup, San Francisco, California.
ASCE, 2010, Minimum Design Loads for Buildings and Other Structures,
ASCE/SEI 7-10, American Society of Civil Engineers Structural
Engineering Institute, Reston, Virgina.
ATC, 2018, Recommended Earthquake Performance Goals for San
Francisco's Buildings (ESIP Task A.6.h), final report submitted to the
City and County of San Francisco prepared by the Applied
Technology Council, Redwood City, California.
Bonowitz D., 2011, Resilience Criteria for Seismic Evaluation of Existing
Buildings, a 2008 Special Projects Initiative Report to the Structural
Engineers Association of Northern California, San Francisco,
California.
CCSF, 2008a, Requirements and Guidelines for the Seismic Design of New
Tall Buildings using Non-Prescriptive Seismic-Design Procedures,
Administrative Bulletin 083, Department of Building Inspections, San
Francisco, California.

ATC 119 -1 Part 3: Performance Expectation for New Buildings 3-41

 CCSF, 2008b, Guidelines and Procedures for Structural Design Review,
Administrative Bulletin 082, Department of Building Inspections, San
Francisco, California.
CCSF, 2011, CAPSS Earthquake Safety Implementation Program Workplan
2012-2042, the City and County of San Francisco, California.
CCSF, 2016, The San Francisco Building Code, the City and County of San
Francisco, California.
FEMA, 2012, Seismic Performance Assessment of Buildings, FEMA P-58,
prepared by the Applied Technology Council for the Federal
Emergency Management Agency, Washington, D.C.
FEMA, 2018, Seismic Performance Assessment of Buildings, Volume 3—
Supporting Electronic Materials and Background Documentation,
Second Edition, FEMA P-58-3, prepared by the Applied Technology
Council for the Federal Emergency Management Agency,
Washington, D.C.
HBrisk, 2018, Seismic Performance Prediction Program, SP3, Haselton
Baker Risk Group, Chico, California.
Hooper, 2018, Personal communication with Peter Morris who previously
provided cost estimates for earthquake repairs as part of the FEMA
P-58 development process.
Jayaram, N., Shome, N., and Rahnama, M., 2012, “Development of
earthquake vulnerability functions for tall buildings,” Journal of
Earthquake Engineering and Structural Dynamics, Vol. 41, No. 11, pp.
1495-1514.
Lang, A., Buckalew, J., Mayes, R., Wade, K., 2018, “FORWARD: Second
generation B2B/BORP and resilient planning for existing buildings,”
SEAOC 2018 Convention, Palm Springs, California.
LSTC, 2009, LS-DYNA, Livermore Software Technology Corporation,
Livermore, California.
McKenna, F., Fenves, G.L., Scott, M.H., and Jeremic, B., 2000, Open System
for Earthquake Engineering Simulation, OpenSees, University of
California, Berkeley, California.
Molina Hutt, C., Almufti, I., Willford, M., and Deierlein G.G., 2016, “Seismic
loss and downtime assessment of existing tall steel-framed buildings
and strategies for increased resilience,” Journal of Structural
Engineering, Vol. 142, No. 8.

3-42 Part 3: Performance Expectation for New Buildings ATC 119-1

Naish, D., Wallace, J.W., Fry, J.A., and Klemencic, R., 2009, Reinforced
Concrete Link Beams: Alternative Details for Improved Construction,
UCLA-SGEL Report 2009-06, Structural & Geotechnical Engineering
Laboratory, University of California, Los Angeles, 103 pp.
NIST, 2013, Cost Analysis and Benefit Studies for Earthquake-Resistant
Construction in Memphis, Tennessee, NIST GCR 14-017-26,
prepared by the NEHRP Consultants Joint Venture, a partnership of
the Applied Technology Council and the Consortium for Universities
for Research in Earthquake Engineering, for the National Institute of
Standards and Technology, Gaithersburg, Maryland.
PEER, 2010, Guidelines for Performance-based Seismic Design of Tall
Buildings, Version 1.0, Report No. 2010/05, Pacific Earthquake
Engineering Research Center, University of California, Berkeley,
California.
PEER, 2011, Case Studies of the Seismic Performance of Tall Buildings
Designed by Alternative Means, PEER Report 2011/05, Task 12
Report for the Tall Buildings Initiative, Pacific Earthquake Engineering
Research Center, University of California, Berkeley, California.
PEER, 2013, NGA-West2 Database, PEER Report 2013/03, Pacific
Earthquake Engineering Research Center, University of California,
Berkeley, California.
PEER, 2017, Guidelines for Performance-based Seismic Design of Tall
Buildings, Version 2.03, Report No. 2017/06, Pacific Earthquake
Engineering Research Center, University of California, Berkeley,
California.
Terashima, M., 2018, in review, Ductile Fracture Simulation and Risk
Quantification of Buckling Restained Braces under Earthquakes,
Dissertation (PhD), Stanford University, Stanford, California.
Tipler, J., 2014, Seismic Resilience of Tall Buildings—Benchmarking
Performance and Quantifying Improvements, Dissertation (MSc),
Stanford University, Stanford, California.
USGS, 2018, The HayWired Earthquake Scenario—Engineering Implications,
USGS Scientific Investigations Report 2017-5013-I-Q, US Geological
Survey, Menlo Park, California.

ATC 119 -1 Part 3: Performance Expectation for New Buildings 3-43

 PART 4:
Post-earthquake
Structural Evaluation
Chapter 1

Introduction

1.1 Background

This Part reviews and explains current San Francisco policy for post-
earthquake structural evaluation, especially as it relates to the City’s existing
tall buildings. It recommends modifications to the current policy and
proposes additional study or development of new regulations to address
identified issues. This Part corresponds to Recommendations 2A, 3D, 3E,
and 3F presented in the Summary Recommendations.

Current San Francisco policy requires the repair of earthquake damage in

nearly all cases and checks three conditions, called triggers, under which
required repair must be supplemented by a full-building seismic evaluation
and, if warranted, a seismic retrofit.

The three triggers represent policy judgments about when retrofit, which adds
cost and recovery time, is nevertheless viewed as beneficial to the
community and reasonable for owners and tenants. This judgment, which
applies to all San Francisco buildings except one- and two-family dwellings,
should be reviewed as it relates to tall buildings, for which recovery time can
be critical and retrofit costs impractical.

This Part addresses aspects of the following tasks described in the CAPSS
Earthquake Safety Implementation Program Workplan 2012-2042 (CCSF,
2011):
• Task A.4.d: Adopt disproportionate damage trigger
• Task B.4.b: Develop post-earthquake repair and retrofit standards

1.2 Organization

Chapter 2 reviews current San Francisco policy for post-earthquake structural

evaluation.

Chapter 3 presents recommendations for improving San Francisco policy for

post-earthquake structural evaluation.

Appendix A provides an annotated set of current San Francisco Existing

Building Code provisions related to this topic.

A list of references is provided at the end of this Part.

ATC-119-1 Part 4: Post-earthquake Structural Evaluation 4-1

Chapter 2

Current Policy

The San Francisco Existing Building Code (SFEBC) (CCSF, 2016a) regulates
most aspects of post-earthquake structural evaluation of buildings. SFEBC
provisions combine the California Existing Building Code (CEBC) (CBSC,
2016), itself an amended version of the International Existing Building Code
(IEBC) (ICC, 2018), and San Francisco amendments.

The SFEBC is organized by project type, recognizing five categories:

additions, alterations, repairs, changes of occupancy, and relocations.
Because post-earthquake evaluation involves the assessment of damage, it
is addressed by the code’s provisions for repairs. With exceptions for one-
and two-family dwellings, the provisions are the same for buildings of any
height.

2.1 Evaluation, Repair, and Triggered Retrofit

2.1.1 2016 SFEBC Provisions

The SFEBC regulates the task of restoring a damaged building’s pre-

earthquake condition and, in some cases, improving it.

SFEBC Section 404.1 requires earthquake damage to be repaired. The

remaining question is whether the code, as public policy, should also require
the structure to be seismically improved at the same time. The answer has
evolved over decades and is represented in the current SFEBC by four ideas:
• When earthquake damage to the seismic-force-resisting system (the wall
and frame elements expected to resist earthquake effects) is substantial,
the repair work will already be extensive, so the repair presents an
opportunity to evaluate and improve an obsolete structure in anticipation
of several more decades of service. The triggering level of damage is
defined in SFEBC Chapter 2 as substantial structural damage to the
lateral system (lateral SSD), and the trigger provision is in SFEBC
Section 404.2.
• When earthquake damage to the gravity system is substantial, it indicates
that the building’s seismic-force-resisting system is grossly deficient,
since the purpose of the seismic-force-resisting system is to protect the
gravity system from collapse. Returning such a deficient building to

ATC-119-1 Part 4: Post-earthquake Structural Evaluation 4-3

 decades of service would be inconsistent with what is expected from new
buildings. This triggering level of damage is a special case of substantial
structural damage to the gravity system (gravity SSD caused by lateral
load), and the trigger is in SFEBC Section 404.3.1.
• When earthquake damage to the seismic-force-resisting system is
surprisingly high, given a low level of shaking, it indicates a grossly
deficient seismic-force-resisting system prone to collapse in a larger
earthquake. Again, returning such a deficient building to decades of
service would be inconsistent with policy for new buildings. This
triggering level of damage, together with the low hazard, is defined as
disproportionate damage (DD), and the trigger is in SFEBC Section
404.4.1. The concept of DD was pioneered in San Francisco, but it has
been adopted by the 2018 IEBC model code and will appear, slightly
modified, in the 2019 edition of the statewide CEBC.
• Whenever triggering SSD or DD exists, the pre-earthquake structure
should be evaluated, but not with the full earthquake loads (or forces) that
would be used for the design of a similar new building. Rather, the
evaluation should be allowed to use reduced loads; this ensures that any
building failing the evaluation is unquestionably deficient, not a marginal
case. If the building fails the evaluation, it must be retrofitted as well as
repaired. The retrofit, like the evaluation, is allowed to use reduced loads.
San Francisco has amended this CEBC allowance to require the larger of
the reduced seismic loads and the loads from “the code under which the
building or structure was designed” (SFEBC Section 404.1.1).

These four codified ideas comprise three repair-based retrofit triggers and
one set of evaluation and design criteria. The triggers represent policy
judgments about when retrofit, which adds cost and recovery time, is
nevertheless viewed as beneficial to the community and reasonable for
owners and tenants. These policy judgments should be reviewed as they
relate to building groups for which recovery time is critical or retrofit costs are
impractical.

Retrofit triggered by repair is more controversial as policy than retrofit

triggered by other project types. Other projects—additions, alterations, or
changes of occupancy—are essentially voluntary. If the building owner does
not want the additional cost of retrofit, she has the option to change the scope
of the project. Repairs, by contrast, are mandatory. Requiring a retrofit at the
same time as the owner is forced to make unplanned repairs (and deal with
tenants and insurers) adds a burden that is likely to delay the building’s
recovery. Further, if many buildings face similar delays, especially if those

4-4 Part 4: Post-earthquake Structural Evaluation ATC-119-1

buildings are clustered geographically or serve the same function, the
aggregate effect might impact recovery on a neighborhood or city scale.

Of the three repair-based triggers, lateral SSD is the most controversial.

Unlike DD and gravity SSD caused by earthquake, lateral SSD is not
necessarily an indicator of a collapse-prone structure; in a large earthquake,
even some new buildings are expected to see significant structural damage.
Thus, the lateral SSD trigger could catch some buildings that would pass a
structural evaluation.

The question, then, is whether the additional time and expense needed to
make these evaluations—which, for complex or tall structures, can be
significant—are worth the benefit. The question is amplified when applied to
dozens of tall buildings (and possibly hundreds more non-tall buildings) at the
same time. On one hand, there is clear value in bringing dozens of tall office
buildings back into service more quickly by eschewing triggered evaluations
and retrofits. On the other hand, if dozens of tall buildings have already
sustained substantial structural damage, just making repairs will not be quick
either; downtime is likely to be on the order of months even in the more
permissive case.

2.1.2 Administrative Bulletins

Two of the code’s three triggers rely on a measurement of damage in terms

of lost “lateral load-carrying capacity.” A 33% capacity loss defines lateral
SSD, while a 10% loss, together with a low shaking level, defines DD.

Building code provisions should have quantitative, enforceable terms, and

this sometimes results in bright-line definitions, such as those of lateral SSD
and DD. Even so, the triggering values in these two definitions are entirely
subjective.

First, there is nothing technically special about a 10% or 33% capacity loss;
that is, a building with a 30% loss is not clearly in a different category than a
building with a 35% loss. Rather, SSD and DD should be understood to have
conceptual meanings despite their quantitative definitions.
• The 10% figure in the DD definition is meant to convey a small but not
insignificant level of damage. The intent of the DD provisions is to identify
structural damage in cases where a properly designed and constructed
building would have none, so the 10% level is meant as “something
noticeable,” as opposed to “practically nothing.”
• The 33% figure in the lateral SSD definition is meant to convey a level of
damage that is serious but still well short of collapse. When this figure

ATC-119-1 Part 4: Post-earthquake Structural Evaluation 4-5

 was debated in building code hearings, proponents suggested that a loss
of half the lateral system on one side of a building—that is, a 25% loss—
should not trigger an upgrade. By consensus judgment, 33% was agreed
to as a more appropriate number representing a substantial and
somewhat unexpected loss even in cases where some structural damage
is anticipated.

Second, there is no agreed definition of capacity loss and no standard

procedure for measuring it, given the observed damage. As with many code
provisions, in-house rules and procedures will be used until a consensus
emerges, perhaps through a formal standards process, but more likely
through informal sharing of practices between implementing agencies. Until
then, it is expected that San Francisco, like every other jurisdiction, will
interpret and apply these definitions with ample discretion.

As the first jurisdiction to codify the concept of disproportionate damage, San

Francisco has made these interpretations through a series of Administrative
Bulletins (ABs) that provide procedures for calculating capacity loss or
replace the somewhat-arbitrary values with more practical and definable
damage descriptions. AB-099, Post-Earthquake Repair and Retrofit
Requirements for Concrete Buildings (CCSF, 2012a), defines lateral SSD
and DD for concrete shear wall and moment frame structures, including those
in tall buildings. AB-098, Post-Earthquake Repair and Retrofit Requirements
for Wood-Frame Residential Buildings with Three or More Dwelling Units
(CCSF, 2012b), and AB-100, Post-Earthquake Repair and Retrofit
Requirements for One- and Two-Family Units (CCSF, 2012c), define lateral
SSD and DD for wood-frame multi-unit buildings and wood-frame dwellings;
they do not apply to tall buildings.

AB-099 uses two methods to categorize damage as meeting the lateral SSD
or DD thresholds:
• A quantitative method, involving structural analysis, based on the
guideline document FEMA 306, Evaluation of Earthquake Damaged
Concrete and Masonry Wall Buildings (FEMA, 1998), and
• A mostly qualitative method based on observed damage patterns, crack
orientation, and crack width.

AB-099 also allows for nonstructural or cosmetic damage that need not be
repaired at all.

AB-099 is mostly applicable to concrete shear wall structures. It should apply

also to dual systems with concrete shear walls and concrete moment frames;
for the moment frames, only the qualitative method is required, and the

4-6 Part 4: Post-earthquake Structural Evaluation ATC-119-1

cosmetic damage level is omitted. Presumably, AB-099 would apply to dual
systems with concrete shear walls and steel moment frames, but the AB is
not completely clear on this point.

Currently, there are no ABs for other structural systems common to San
Francisco’s tall buildings. The San Francisco Department of Building
Inspection’s guidelines for its Building Occupancy Resumption Program refer
to FEMA 352, Recommended Postearthquake Evaluation and Repair Criteria
for Welded Steel Moment-Frame Buildings (FEMA, 2000), which addresses
some of the same questions for steel structures that FEMA 306 does for
concrete, but FEMA 352 is not required and is not written into enforceable
provisions through an AB.

2.2 Relation to Post-Earthquake Safety Inspection

The structural evaluations that guide repairs are different from post-
earthquake safety inspections. Repairs are planned, even if the damage
being repaired was not. The applicable code provisions are concerned with
regulating the building stock over the long term. Repairs therefore are
regulated by the code as defined projects, along with administrative
provisions for documentation, permitting, fees, and quality assurance.

By contrast, post-earthquake safety inspection is performed with a more

immediate interest in short-term safety. As such, the code’s repair provisions
offer no rules or procedures for accessing or occupying a potentially
damaged building, finding or ruling out damage, or posting or removal of
placards (Red, Yellow, or Green). Post-earthquake safety inspection is
addressed in Part 6.

The intersection of the SFEBC with safety inspection is limited to the code’s
general provisions regarding dangerous conditions. Whether a building is
identified as dangerous will often be related to the findings made during post-
earthquake safety inspection. Immediate measures to stabilize the structure
might be made to eliminate dangerous conditions and to allow further
assessment or reoccupancy. This is all SFEBC Section 401.3 contemplates;
the code still requires permits for these “emergency repairs” but allows the
paperwork to be completed after the fact. Those immediate measures are
likely to be different from the code-required repairs, however. How structural
damage will be repaired, and specifically whether a seismic upgrade will be
required, is unrelated to the building’s status as dangerous.

Indeed, by the time planned repairs are able to begin, the structure will have
been stabilized and any Red or Yellow placards will have been removed or
modified. In cases of light damage, it is even likely that general reoccupancy

ATC-119-1 Part 4: Post-earthquake Structural Evaluation 4-7

 will be allowed before the repair-guiding evaluations are complete and before
the repair work is begun.

2.3 Relation to Other Code-Regulated Work

Repairs (and the structural evaluations that guide them) are distinct from
maintenance, alteration, and retrofit projects, as noted in SFEBC Section
404.1.

In general, all buildings must be maintained in serviceable and habitable

conditions. In recent code cycles, the International family of codes, which
includes the IEBC, has begun to clarify the intended distinction between
maintenance and repair, both of which are required. While the resolution is
not yet complete, a fair statement of the intent is that maintenance preserves
an acceptable condition, while repair corrects an unacceptable condition. In
any case, earthquake damage is a condition likely to require repair, not
maintenance.

Related to general maintenance requirements is San Francisco’s façade

inspection and maintenance program for pre-1998 buildings taller than four
stories. Program requirements are given in SFEBC Chapter 4E and AB-110,
Building Façade Inspection and Maintenance (CCSF, 2017). In addition to
deadlines for regular inspection (the first of which, for buildings constructed
before 1910, is not until the end of 2021), the program requires inspection
when façade elements “exhibit significant damage or failure” due to an
earthquake (Section 403E.3). By responding to apparent damage only, the
requirement is not proactive and is thus redundant to the code’s general
requirements to repair damage and to eliminate dangerous conditions. Also,
the earthquake provision does not trigger any evaluation or retrofit of the
structural system.

Alteration, addressed by SFEBC Section 403, is distinct from repair and

unrelated to post-earthquake evaluation. Model code revisions since 2009
have made this distinction clearer, and SFEBC Section 404.1 further clarifies
that any work necessary to complete a repair is to be treated as part of the
repair, not as an alteration. Thus, earthquake damage should not invoke, or
trigger, any requirements that normally attach to alteration projects.
Nevertheless, this distinction is worth noting because past codes covered
alterations and repairs with the same set of provisions, and local practices in
some jurisdictions continue to conflate these project types. Current policy
and recommendations regarding alteration projects are addressed in Part 7.

Retrofit, whether voluntary, mandated by city policy, or triggered by a

substantial project, is a type of alteration. Like all alterations, retrofit is

4-8 Part 4: Post-earthquake Structural Evaluation ATC-119-1

distinct from repair and unrelated to post-earthquake evaluation. This
distinction is worth noting, however, because many owners and tenants refer
to retrofit as “fixing” a deficient building, leading to possible confusion.

ATC-119-1 Part 4: Post-earthquake Structural Evaluation 4-9

Chapter 3

Issues and Recommendations

3.1 Revisions and Amendments to the SFEBC

3.1.1 Policy on Repair-Triggered Retrofit

San Francisco policy regarding seismic retrofit triggered by damage and

repairs is embodied in three SFEBC provisions. The three triggering
provisions represent judgments about when the community’s benefit from
retrofit outweighs the expected cost and possible delays in recovery. Of the
three triggers, two specifically target highly deficient structures. The third,
known as lateral SSD, is more likely to affect buildings that time-consuming
and costly structural evaluations will often find acceptable. In concept, if tall
buildings represent a community function for which recovery time is critical or
retrofit cost is impractical, then the lateral SSD trigger might be waived or
adjusted for buildings in this group. (Current provisions gravity SSD caused
by earthquake, however, should not be waived or relaxed.) Since the issue
here is loss of community function, not building height, any reconsideration of
current policy should apply to all buildings that support that function, or at
least to large and complex buildings for which costs and delays will have
similar effects regardless of height.

To address these questions, San Francisco could study the recovery impacts
of the lateral SSD trigger through neighborhood-level simulation. A more
limited (less costly) study would estimate the recovery impacts on individual
buildings of different types. Informed by the findings, the City could decide to
waive the lateral SSD trigger for certain buildings when the damage is due to
an event, like an earthquake, that affects thousands of buildings
simultaneously.

If the policy adjustment is made after the event, the process would be able to
adjust to facts on the ground and grant waivers only as needed. Although
flexibility is valuable, however, making a general policy under emergency
conditions is administratively and politically fraught. Doing so would take
attention away from more pressing issues, would make preparations more
difficult, could lead to confusion or inconsistency, and is likely to be subject to
pressure from building owners, jeopardizing the trust of other stakeholders.
For these reasons, advance policy making is recommended.

ATC-119-1 Part 4: Post-earthquake Structural Evaluation 4-11

 The policy-making process should also consider that the Federal Emergency
Management Agency (FEMA) endorses all three current triggers, and FEMA
policy requires compliance with the model code in order to be eligible for
post-disaster assistance. This means that any relaxation of the lateral SSD
trigger, which has been accepted in the CEBC since 2009, could jeopardize
federal recovery funds. That said, FEMA assistance does not generally go to
office buildings, so the waiver or code change might avoid this issue by
applying only to certain occupancies, building sizes, or neighborhoods.

Reduced loads for triggered evaluation and retrofit are meant to mitigate
some of the concerns about repair-triggered retrofit. However, by amending
the criteria to include the building’s original design loads, San Francisco
negates that mitigating measure, especially for younger buildings. To
preserve the allowance intended by the CEBC, San Francisco should amend
the SFEBC to allow reduced loads in the event of repair-triggered retrofit.
(The question of which loads to use is related to the currently unclear
application of the SFEBC table of benchmark dates, discussed in Appendix A
of this Part under SFEBC Section 301.2.1.)

Relaxing the current policy for repair-triggered retrofit of tall buildings might
raise concerns about favoring recovery over safety. In fact, relaxing a repair
trigger does not make any building less safe, but any perceived loss of safety
can be made up through other programs. Indeed, there are more effective
ways to address both safety and recovery than tinkering with the code’s
repair triggers. Pre-earthquake mandates and incentives, together with
recovery planning, are likely to be more effective at risk reduction or risk
transfer than changes to the evaluation rules that only get applied after losses
have occurred. These alternatives are discussed in Part 6 and Part 7.

Recommendation. Study the benefits of relaxing SFEBC provisions

regarding substantial structural damage to the lateral-force-resisting system.
• Consider the effects of lateral SSD-triggered evaluation and retrofit on
post-earthquake recovery time, for individual tall and non-tall buildings.
• Consider the aggregate policy effects on neighborhoods with
concentrations of tall buildings and buildings supporting recovery-critical
functions.
• Consider waiving the current SFEBC requirement to consider “original
loads” for lateral SSD-triggered seismic evaluation and retrofit.
• Consider allowing benchmarking per SFEBC Section 301.2.1 for lateral
SSD-triggered evaluation and retrofit.

4-12 Part 4: Post-earthquake Structural Evaluation ATC-119-1

3.1.2 Routine Code Development

While San Francisco pioneered the concept of disproportionate damage and

has for years enforced its own provisions for triggered retrofits, many of its
local amendments are now out of coordination with the CEBC and IEBC. In
many cases, as shown in Appendix A of this Part, because the CEBC has
adopted some of San Francisco’s ideas, the SFEBC is now duplicative or out
of step. To improve the SFEBC, at least with respect to repair-triggered
retrofit provisions, the City should review Appendix A of this Part and address
the issues raised in its commentary. The easiest way to do this would be to
make any necessary changes or corrections as the City updates its
amendments for the forthcoming 2019 CEBC, since that process will already
require a careful review and reorganization of current SFEBC Chapter 3 and
Chapter 4.

Recommendation. Update the SFEBC.

• Correct section numbering errors for SFEBC amendments.
• To avoid duplication, omit the SFEBC definition and triggers for
disproportionate damage. Use the CEBC content instead.
• Coordinate Sections 301.2 and 404.1.1 with Section 301.1.4.2.
o Revise the text that incorrectly says Section 301.2.1 applies only
when invoked by Section 403.
o Clarify whether SFEBC benchmarking in Section 301.2.1 is allowed
for repair projects.
o Clarify whether SFEBC benchmarking is allowed when original design
loads are also required per Section 404.1.1.
o Since Section 301.2.1 applies only to cases where reduced forces are
used (alterations and repairs), consider combining with Section
301.1.4.2 (to be Section 303.3.2 in the 2019 SFEBC).
o Delete the unnecessary pointer from Section 301.2.3 to Section
301.1.4.2. Pointers to Section 301.2.3 should instead point directly to
Section 301.1.4.2.
o Move Section 404.1.1 into Section 301.2 (or Section 301.1.4.2) and
clarify whether the requirement to consider original loads applies to
gravity SSD cause by lateral load (Section 404.3.1) and to DD
(Section 404.4.1).
o Clarify whether Section 404.1.1 original loads requirement applies
when AB-099 (and others) is applied, as those ABs are premised on
reduced loads.

ATC-119-1 Part 4: Post-earthquake Structural Evaluation 4-13

 • Make revisions 2016 CEBC to correct errors that occurred while adopting
the 2015 IEBC.
• Clarify or revise limits on upgrade triggers imposed by the Department of
Housing and Community Development.
• Omit Section 404.4.2.

3.2 Administrative Bulletins and Reference Documents

3.2.1 AB-099

AB-099 will help implement SFEBC repair provisions after the next damaging
earthquake. Even so, AB-099 has not been used since its adoption by DBI in
2012. Most important, AB-099 references and relies on FEMA 306,
published in 1998 yet still unfamiliar to most engineers, plan checkers, and
potential peer reviewers. Simply put, there have not been enough
opportunities to apply and vet FEMA 306 with actual earthquake damage. In
addition, parts of AB-099 and FEMA 306 have fallen out of coordination with
the CEBC and SFEBC and with their reference standards. Therefore, an
update of AB-099 and a review of FEMA 306 with respect to San Francisco
tall building types is recommended.

Recommendation. Update AB-099 and clarify its application to tall concrete

structural systems.
• Reference SFEBC Table 301.2.1 or remove the obsolete reference to the
May 21, 1973, benchmark date.
• Reference current codes and standards, including the forthcoming 2019
CEBC and SFEBC and ASCE/SEI 41-17 (ASCE, 2017), replacing
references to outdated documents, such as ASCE/SEI 31-03, ASCE/SEI
41-06, ASCE/SEI 41-13, FEMA 273, and FEMA 356. This should include
revising the performance objective in AB-099 Step 3 to reference the
BSE-1E hazard level.
• Make the alternative performance objective consistent with SFEBC Table
301.1.4.2, and with SFEBC Section 404.1.1 as needed.
• Update the definition of lateral SSD, now based on 33% (not 20%) lateral
capacity loss. This will include revisions to the text, to Figure 1, to the
FEMA 308 procedure, and possibly to the Table 1 thresholds.
• Clarify application to concrete elements in dual systems with steel frames.
• Review FEMA 306 to ensure that its references to FEMA 273 (or AB-
099’s references to ASCE/SEI 41-13) are consistent with ASCE/SEI
41-17.

4-14 Part 4: Post-earthquake Structural Evaluation ATC-119-1

• Review FEMA 306 applicability to tall buildings in light of the note in
FEMA 306 Section 4.4.1 regarding applicability of the method’s basic
procedures to structures with significant response outside the
fundamental mode.
• Coordinate post-earthquake safety inspection procedures for concrete
buildings (see Part 6) with FEMA 306 guidelines for inspection and
damage-finding.

3.2.2 New Administrative Bulletins

AB-099 offers procedures and interpretations that relate technical damage

observations to SFEBC’s policy prescriptions. Having procedures and
interpretations defined in advance will greatly improve implementation of
these rarely-used code provisions. The idea should be extended to other
structure types. In particular, as discussed in Part 1, San Francisco’s tall
buildings are more likely to have steel than concrete seismic-force-resisting
systems. A similar AB for steel moment frames and dual systems with steel
moment frames, referencing FEMA 352, should be developed. After that,
ABs for less common structural systems, including concentric and eccentric
braced frames should be considered.

It is possible that some welded steel moment frames in San Francisco have
damage from the 1989 Loma Prieta earthquake that has still not been
repaired. An AB implementing FEMA 352 will be needed as the City
develops a policy or program to find and repair, or rule out, that damage, as
discussed in Part 6.

Recommendation. Develop a new AB for post-earthquake inspection and

evaluation of welded steel moment frames.
• The AB should address both safety inspection and structural evaluation,
as FEMA 352 covers both topics. The AB should reference FEMA 352
and set specific requirements where the document allows options. See
also Part 6.
• Review FEMA 352 for application to representative San Francisco
structures. In particular, review the findings of the SEAONC EBC Steel
Frame Subcommittee (2004) showing that results are sensitive to the joint
sampling method.
• The AB should apply FEMA 352 to interpret SFEBC repair-based retrofit
triggers for pre-Northridge welded steel moment frames.

ATC-119-1 Part 4: Post-earthquake Structural Evaluation 4-15

Appendix A

Annotated Building Code

Provisions

This appendix presents the current building code provisions applicable to

post-earthquake structural evaluation and repair-triggered seismic retrofit.
The current, 2016, San Francisco Existing Building Code (SFEBC) amends
the 2016 California Existing Building Code (CEBC), which itself adopts and
amends the 2015 International Existing Building Code (IEBC). This appendix
shows the relevant provisions from all three documents organized by
chapters as they appear in the 2016 CEBC. In addition, it provides a
commentary on the combined provisions and their application to existing tall
buildings, and it offers notes on the forthcoming 2019 CEBC, which will adopt
the 2018 IEBC.

This appendix focuses on earthquake evaluation and retrofit provisions that

apply to post-earthquake assessment of tall buildings associated with repair
projects. As such, it does not include all existing building provisions, all tall
building provisions, or even all repair or retrofit provisions. In particular, the
following specialty provisions might apply to certain tall buildings but are
omitted from this appendix for brevity:
• Historical Building Code provisions
• Provisions for state-owned buildings in CEBC Sections 317-322; Section
317.1.1 allows these sections to be adopted by local jurisdictions for use
on private buildings
• Regulations used by California’s Office of Statewide Health Planning and
Development (OSHPD) and Division of State Architect (DSA)
• San Francisco amendments regarding buildings “on a military base
selected for closure” (SFEBC Section 326) and homeless shelters
(SFEBC Section 401.7), which are not likely to affect tall buildings.

California amendments from the Department of Housing and Community

Development (HCD) are included, since they are meant to cover apartment
buildings, condominiums, and hotels that might be in tall buildings.

ATC-119-1 Part 4: Post-earthquake Structural Evaluation 4-17

 The text is formatted to distinguish between the various source documents
using the following legend:
2015 IEBC. Model code. Plain text (except for section headings in bold).
2016 CEBC. California amendments, applicable in San Francisco. Italic plain
text (except for section headings in bold).
2016 SFEBC. Additional San Francisco amendments. Bold italic text.
Commentary. Plain italic text, indented.
2018 IEBC. Model code language already published and expected to
be adopted into the 2019 CEBC and SFEBC. Plain text, shaded and
indented.

CHAPTER 1
SCOPE AND ADMINISTRATION

1.8 Department of Housing and Community Development

1.8.10 Other Building Regulations

1.8.10.1 Existing structures. Notwithstanding other provisions of law, the

replacement, retention, and extension of original materials and the use of
original methods of construction for any existing building or accessory
structure, or portions thereof, shall be permitted in accordance with the
provisions of this code as adopted by the Department of Housing and
Community Development. For additional information, see California Health
and Safety Code, Sections 17912, 17920.3, 17922 and 17958.8.

Commentary: This and similar California amendments for HCD are

generally interpreted to prohibit upgrade triggers for residential
buildings. See the commentary at Section 401.2.1.

4-18 Part 4: Post-earthquake Structural Evaluation ATC-119-1

 CHAPTER 2
DEFINITIONS

DANGEROUS. Any building, structure or portion thereof that meets any of

the conditions described below shall be deemed dangerous:
1. The building or structure has collapsed, has partially collapsed, has moved
off its foundation, or lacks the necessary support of the ground.
2. There exists a significant risk of collapse, detachment or dislodgement of
any portion, member, appurtenance or ornamentation of the building or
structure under service loads.

Commentary: The definition of dangerous is qualitative in order to give

discretion to the code official. In the context of post-earthquake
evaluation, a damaged building might be deemed dangerous under
either part of the definition. The identification of dangerous buildings,
however, has nothing to do with the evaluation of seismic deficiencies
or potential seismic damage under normal pre-earthquake conditions;
“service loads” is understood to mean normal day-to-day conditions.
In rare cases, the code official might designate an undamaged
building dangerous in anticipation of aftershock damage.

DISPROPORTIONATE DAMAGE. A condition of earthquake-related

damage where:
1. The 0.3-second spectral acceleration at the building site as estimated
by the United States Geological Survey for the earthquake in question is
not more than 0.40g; and
2. In any story, the vertical elements of the lateral force-resisting system
have suffered damage such that the lateral load-carrying capacity of the
structure in any horizontal direction has been reduced by more than
10% from its pre-damage condition.

Commentary: The SFEBC originated the concept of disproportionate

earthquake damage with the provision shown above. Neither the 2016
IEBC nor the 2016 CEBC uses this term, but the 2018 IEBC (and
2019 CEBC) will include a similar term with a similar triggering
provision (see SFEBC Section 404.4.1, below).
San Francisco ABs 098, 099, and 100 provide interpretations
regarding calculation of capacity loss as needed to implement this
definition. AB-099, for concrete buildings, will apply to some tall
buildings.

ATC-119-1 Part 4: Post-earthquake Structural Evaluation 4-19

 The 2018 IEBC definition follows here. Note the differences relative to
the SFEBC:
• Both the IEBC and SFEBC address only earthquake damage, but
the IEBC term is more explicit about that.
• In the SFEBC, the bounding ground motion is 0.40g. In the 2018
IEBC, to cover the entire country, it is the more generic 0.4SS.
• Relative to the SFEBC, the 2018 IEBC makes an editorial
clarification to the second condition, with no change in intent.
DISPROPORTIONATE EARTHQUAKE DAMAGE. A condition of
earthquake-related damage where both of the following occur:
1. The 0.3-second spectral acceleration at the building site as
estimated by the United States Geological Survey for the earthquake
in question is less than 40% of the mapped acceleration parameter
SS.
2. The vertical elements of the lateral force-resisting system have
suffered damage such that the lateral load-carrying capacity of any
story in any horizontal direction has been reduced by more than 10%
from its predamage condition.

SUBSTANTIAL STRUCTURAL DAMAGE. A condition where one or both of

the following apply:
1. In any story, the vertical elements of the lateral force-resisting system have
suffered damage such that the lateral load-carrying capacity of the structure
in any horizontal direction has been reduced by more than 33% from its
predamage condition.
2. The capacity of any vertical gravity load-carrying component, or any group
of such components, that supports more than 30% of the total area of the
structure’s floor(s) and roof(s) has been reduced more than 20% from its
predamage condition, and the remaining capacity of such affected elements,
with respect to all dead and live loads, is less than 75% of that required by
this code for new buildings of similar structure, purpose and location.

Commentary: San Francisco ABs 098, 099, and 100 provide

interpretations regarding calculation of capacity loss as needed to
implement this definition. AB-099, for concrete buildings, will apply to
some tall buildings.
Item 1 of the definition will usually be of greatest interest to post-
earthquake evaluation; the corresponding trigger is found in Section
404.2. The definition was changed with the 2012 IEBC to increase

4-20 Part 4: Post-earthquake Structural Evaluation ATC-119-1

 the critical damage level from 20% to 33%. AB-099 still references
the old definition.
Item 2 of the definition is also applicable to post-earthquake
evaluation, especially in older buildings in which the lateral force-
resisting system is inadequate to protect the gravity system from
damage; the corresponding trigger is found in Section 404.3.1.
The 2018 IEBC definition for substantial structural damage (SSD) has
been edited for clarity, with no change in intent. It will also include a
third category for damage caused by snow that will likely not apply to
San Francisco tall buildings.
SUBSTANTIAL STRUCTURAL DAMAGE. A condition where any of
the following apply:
1. The vertical elements of the lateral force-resisting system have
suffered damage such that the lateral load-carrying capacity of any
story in any horizontal direction has been reduced by more than 33%
from its predamage condition.
2. The capacity of any vertical component carrying gravity load, or any
group of such components, that has a tributary area more than 30% of
the total area of the structure’s floor(s) and roof(s) has been reduced
more than 20% from its predamage condition, and the remaining
capacity of such affected elements, with respect to all dead and live
loads, is less than 75% of that required by the International Building
Code for new buildings of similar structure, purpose and location.
3. The capacity of any structural component carrying snow load ....

CHAPTER 3
PROVISIONS FOR ALL COMPLIANCE METHODS

301.1.4 Seismic evaluation and design procedures. The seismic

evaluation and design shall be based on the procedures specified in the
California Building Code or ASCE 41. The procedures contained in Appendix
A of this code shall be permitted to be used as specified in Section 301.1.4.2.

Commentary: In the 2018 IEBC, Section 301.1.4 has been edited for
clarity and to coordinate with ASCE/SEI 41-17 instead of ASCE/SEI
41-13. The entire section has also been relocated and renumbered
as Section 303.3.

The 2016 CEBC adopts only three of the 2015 IEBC’s five Appendix A
chapters. None of the three adopted chapters applies to tall buildings
(or to steel or concrete structural systems of any size).

ATC-119-1 Part 4: Post-earthquake Structural Evaluation 4-21

 301.1.4.1 Compliance with International Building Code-level seismic
forces. ...

Commentary: Section 301.1.4.1 gives the “full” earthquake design

criteria applicable when seismic work is triggered by an addition or
change of occupancy project; it does not apply to repairs.

301.1.4.2 Compliance with reduced International Building Code-level

seismic forces. Where seismic evaluation and design is permitted to meet
reduced California Building Code seismic force levels, the criteria used shall
be in accordance with one of the following:

1. The California Building Code using 75% of the prescribed forces. Values of
R, Ω0 and Cd used for analysis shall be as specified in Section 301.1.4.1 of
this code.

2. Except where these requirements are triggered by Section 403.12,

structures or portions of structures that comply with the requirements
of the applicable chapter in Appendix A as specified in Items 2.1
through 2.5 and subject to the limitations of the respective Appendix A
chapters shall be deemed to comply with this section.
...
2.5. Seismic evaluation and design of concrete buildings assigned to Risk
Category I, II or III are permitted to be based on the procedures specified in
Chapter A5.

3. ASCE 41, using the performance objective in Table 301.1.4.2 for the
applicable risk category.

Commentary: Section 301.1.4.2 gives the “reduced” earthquake

design criteria for use when seismic work is triggered by certain
alteration or repair projects. See Section 404.
The 2016 CEBC reference to Chapter A5 is inappropriate, as
California does not adopt that chapter of the model code. Also, the
2018 IEBC (and presumably the 2019 CEBC) will no longer have
Chapter A5; it was omitted because it is essentially identical to
ASCE/SEI 41-17. As noted above, none of the other Appendix A
chapters adopted by the CEBC is applicable to tall buildings or to
concrete or steel structural systems of any size.

4-22 Part 4: Post-earthquake Structural Evaluation ATC-119-1

 Section 301.1.4.2 is referenced by SFEBC Section 301.2, which
contains San Francisco’s traditional reduced load criteria with
benchmark dates. The CEBC allows three options where reduced
loads are allowed. Presumably, all three options are acceptable
within the intent of SFEBC Section 301.2.
• Option 1 is the traditional “75%” approach long embraced by San
Francisco.
• Option 2 allows five prescriptive approaches for specific building
types. As noted above, however, California adopts only three of
the five Appendix A chapters, none of which is applicable to tall
buildings, and Chapter A5 (not adopted by the CEBC) has already
been removed from the 2018 IEBC.
• Option 3 allows the use of ASCE/SEI 41-13 with performance
objectives that vary by risk category: For RC I or II: Structural Life
Safety with a BSE-1E hazard; for RC III: Structural Damage
Control with a BSE-1E hazard. Tall buildings would not be
expected to be assigned to Risk Category IV.

301.2 Minimum Lateral Force for Existing Buildings

301.2.1 General. This section is applicable to existing buildings when

invoked by SFEBC Section 403. This section may be used as a standard
for voluntary upgrades.

Commentary: Section 301.2 gives additional provisions for seismic

work triggered by alteration projects in Section 403. Despite the
wording of Section 301.2.1, these provisions are also invoked for
repair projects by Sections 404.1.1 and 404.2.1 (and by Section
404.4.1, which points back to 404.2). However, Section 404.1.1
requires consideration of two force levels—the reduced forces
generally allowed by Section 301.2 and the original design forces,
which for relatively recent buildings are likely to be larger than the
reduced forces.

An existing building or structure which has been brought into

compliance with the lateral force resistance requirements of the San
Francisco Building Code in effect on or after the dates shown in Table
301.2.1 [Table 4-1], shall be deemed to comply with this section except
when a vertical extension or other alterations are to be made which
would increase the mass or reduce the seismic resistance capacity of
the building or structure. Where multiple building types apply, the later

ATC-119-1 Part 4: Post-earthquake Structural Evaluation 4-23

 applicable date shall be used. Where none of the building types apply,
compliance shall be at the discretion of the Director. Building type
definitions are given in ASCE 41-13, Table 3-1.

Commentary: Starting with the 2016 edition, this section includes

benchmark dates for each structural system (as opposed to the prior
provision, which gave a single date for all buildings). Where the
benchmark table, as repeated in Table 4-1 below applies, it effectively
eliminates the requirement for triggered seismic evaluation (and
possible retrofit) in more recent buildings, some as old as 1995.
It is possible (but unclear in the SFEBC) that the benchmark table is
not meant to apply when Section 301.2 is invoked as criteria for
upgrades triggered by either substantial structural damage or
disproportionate damage. As discussed below, the San Francisco
amendment in Section 404.1.1 says that seismic evaluations and
retrofits triggered by substantial structural damage must consider not
only the reduced loads prescribed by Section 301.2 but also “the code
under which the building ... was designed.” By a plain reading, this
changes nothing about whether the benchmark table applies to
exempt relatively new buildings. Another interpretation, however,
would recognize that the benchmark table (like the similar table in
ASCE/SEI 41-17) is meant as a rough approximation of when
traditional reduced criteria would be expected to be satisfied. Any
requirement to use higher forces represents an attempt to be more
conservative than the traditional reduced criteria, so the “original
code” requirement in Section 404.1.1 means the benchmark table
should not apply. By this interpretation, no building with triggering
damage levels would be exempt from upgrade just because it is
relatively new.
For example, consider a steel moment-resisting frame structure
designed with the 1997 UBC in 1999. According to Table 301.2.1, the
benchmark date for this system is 12/28/1995, so the building would
be exempt from triggered seismic evaluation—unless the requirement
to consider the original 1997 UBC design loads is understood to
override the benchmark. Even with the additional requirement, the
structure will not necessarily need to be retrofitted. Since it was
designed with the 1997 UBC, any evaluation using its original design
forces should find the structure adequate, so that it need only be
repaired to its pre-damage condition. If the evaluation reveals an
original design flaw, however—and especially if the triggering damage
is thought to be related to it—then prohibiting the use of the

4-24 Part 4: Post-earthquake Structural Evaluation ATC-119-1

 benchmark table will have been worthwhile. Thus, the point of
prohibiting the benchmark table in cases of structural damage is not to
force an upgrade but to ensure a thorough review of the original
design. On the other hand, if the original design was well
documented, it would seem that the engineer of record could simply
show compliance by documentation, without the need for a costly and
time-consuming full evaluation, so original errors, if there were any,
might still not be found.

Table 4-1 Dates Required to Demonstrate Building Compliance (reproduced

from Table 301.2.1 in Section 301.2 of SFEBC)
Date of Model Code
Building Type Compliance (for reference)
Wood Frame, wood shear panels (Types W1 & W2) 1/1/1984 UBC 1976
Wood Frame, wood shear panels (Type W1A) 7/1/1999 UBC 1997
Floor areas greater than 3,000 ft2 per level
Steel moment-resisting frame (Types S1 & S1a) 12/28/1995 UBC 1994
Steel concentrically braced frame (Types S2 & S2a) 7/1/1999 UBC 1997
Steel eccentrically braced frame (Types S2 & S2a) 1/1/1990 UBC 1988
Buckling-restrained braced frame (Types S2 &S2a) 1/1/2008 IBC 2006
Light metal frame (Type S3) 1/1/2008 IBC 2006
Steel frame w/ concrete shear walls (Type S4) 12/28/1995 UBC 1994
Steel plate shear wall (Type S6) 1/1/2008 IBC 2006
Reinforced concrete moment-resisting frame (Type C1) 12/28/1995 UBC 1994
Reinforced concrete shear walls (Types C2 & C2a) 12/28/1995 UBC 1994
Tilt-up concrete (Types PC1 & PC1a) 7/1/1999 UBC 1997
Precast concrete frame (Types PC2 & PC2a) 1/1/2008 IBC 2006
Reinforced masonry (Type RM1) Flexible diaphragms 7/1/1999 UBC 1997
Reinforced masonry (Type RM2) Stiff diaphragms 12/28/1995 UBC 1994
Seismic isolation or passive dissipation 7/1/1992 UBC 1991

301.2.2 Wind forces. Buildings and structures shall be capable of

resisting wind forces as prescribed in San Francisco Building Code
Section 1609.

301.2.3 Seismic forces. Buildings and structures shall comply with the
reduced International Building Code-level seismic forces, as defined in
Section 301.1.4.2. The building separation limitations of Section ASCE
7-10 Section 12.12.3 need not be considered. ...

ATC-119-1 Part 4: Post-earthquake Structural Evaluation 4-25

 Commentary: Section 301.2.3 gives the “reduced” earthquake design
criteria for use when a seismic upgrade is triggered by certain
alteration or repair projects. The first part of this provision is not
actually needed; as shown, it merely points to criteria already
provided in CEBC Section 301.1.4.2. This sentence remains,
however, as a vestige of prior SFEBC provisions that specified
reduced forces directly, in coordination with San Francisco upgrade
triggers for alteration projects (SFEBC Sections 403.12.1 and
403.12.2)—triggers the IEBC/CEBC does not have.
In addition to setting the force level, Section 301.2 provides
benchmarking (Section 301.2.1) and waives building separation
requirements (Section 301.2.3); these provisions are unique and
traditional to San Francisco. As such, they are perhaps meant to
apply only to work triggered by San Francisco amendments, but the
way the code is written, they would apply to any project subject to a
code section that invokes Section 301.2. As discussed above,
however, in cases of substantial structural damage, Section 404.1.1
invokes Section 301.2 but also requires consideration of “the code
under which the building or structure was designed.” So, for a
building with substantial structural damage, if the original design code
was more conservative than the reduced forces allowed by Section
301.2, then the triggered evaluation, as well as any subsequent
upgrade, would have to use the larger forces. As discussed at
Section 301.2.1, it is unclear whether the requirement to consider the
original design code is meant to override or negate the benchmarking
allowance.
CHAPTER 4
PRESCRIPTIVE COMPLIANCE METHOD

401.1.2 Existing state-owned structures. [BSC] ... The provisions of

Sections 317 through 322 may be adopted by a local jurisdiction for
earthquake evaluation and design for retrofit of existing buildings.
...
401.2.1 Existing materials. ... [HCD 1] Local ordinances or regulations shall
permit the replacement, retention and extension of original materials, and the
use of original methods of construction [if the] structure complied with the
building code provisions in effect at the time of original construction and ...
does not become or continue to be a substandard building. ...

Commentary: This and similar California amendments for HCD are

generally interpreted to prohibit upgrade triggers for residential

4-26 Part 4: Post-earthquake Structural Evaluation ATC-119-1

 buildings. However, the final phrase of the provision, regarding
“substandard building,” alters that general interpretation with respect
to earthquake design. “Substandard building” is defined in Health and
Safety Code Section 17920.3 to include any residential building with
“inadequate structural resistance to horizontal forces.”
With an explicit statement of policy in AB-099, San Francisco defines
any building with either disproportionate damage or substantial
structural damage to the seismic-force-resisting system to be
“substandard” for purposes of applying the HCD amendments to the
CEBC. Thus, both of the repair-based seismic upgrade triggers in the
SFEBC will apply even to HCD-regulated buildings.
This is a helpful policy clarification as well as a reasonable
interpretation. The disproportionate damage trigger is explicitly about
damage-prone, and thus inadequate, buildings. Substantial structural
damage can occur even in otherwise adequate buildings, but it would
not trigger upgrade unless the building would also fail an evaluation
with reduced forces, which would indicate inadequacy.

401.3 Dangerous conditions. The building official shall have the authority to
require the elimination of conditions deemed dangerous.

401.4 Dangerous conditions. [BSC] Regardless of the extent of structural

or nonstructural damage, the building official shall have the authority to
require the elimination of conditions deemed dangerous.

Commentary: CEBC Section 401.4 is an unnecessary restatement,

slightly edited, of Section 401.3. The additional phrase regarding the
extent of damage made sense when this provision was located in a
different section. Now that the provision is at the top of Chapter 4, the
leading phrase is no longer needed. In the 2018 IEBC, this provision
has been moved to Chapter 3 where its broad application is even
clearer.

SECTION 404
REPAIRS

404.1 General. Buildings and structures, and parts thereof, shall be repaired
in compliance with Sections 401.2 and 404. Work on nondamaged
components that is necessary for the required repair of damaged
components shall be considered part of the repair and shall not be subject to
the requirements for alterations in this chapter. Routine maintenance required

ATC-119-1 Part 4: Post-earthquake Structural Evaluation 4-27

 by Section 401.2, ordinary repairs exempt from permit in accordance with
Section 105.2, and abatement of wear due to normal service conditions shall
not be subject to the requirements for repairs in this section.

404.1.1 Replacement, retention and extension of original materials.

[HCD 1] Local ordinances or regulations shall permit the replacement,
retention and extension of original materials, and the use of original methods
of construction [if the] structure complied with the building code provisions in
effect at the time of original construction and ... does not become or continue
to be a substandard building. ...

Commentary: Section 401.2 allows existing materials to remain unless

dangerous. The HCD 1 provision added as Section 404.1.1 is
consistent with the amendment to 401.2.1. See the commentary
there.

404.1.1 Repairs. Unless otherwise approved by the Building Official, all

structural damage shall be repaired.

Repairs to buildings or structures which have sustained substantial

structural damage to lateral force resisting elements shall comply with
the minimum lateral force design requirements of Section 301.2 or with
the code under which the building or structure was designed,
whichever is more restrictive.

Damage may be caused by events or a combination of events,

including, but not limited to, fire, explosion, structural pest or wood-
destroying organism attack, earthquake, wind storm, vehicular impact,
ground subsidence or failure, or the collapse or dislodgement of any
portion of any adjacent building or structure. The removal or alteration
of structural elements as part of the work described in an approved
building permit application shall not be considered to be “damage.”

Commentary: This additional SFEBC section is misnumbered; it is not

the intent of the SFEBC to replace CEBC Section 404.1.1. Further,
the second sentence would be more logically located with Section
404.2, since it is specifically about the topic of that section: substantial
structural damage to the lateral system.
The second sentence invokes the reduced forces of Section 301.2
(which points to CEBC Section 301.1.4.2) but also requires
consideration of “the code under which the building ... was designed.”

4-28 Part 4: Post-earthquake Structural Evaluation ATC-119-1

 This means buildings with substantial structural damage to the lateral
system (from any cause) must be evaluated with the full original
design loads, not with reduced current loads. In addition, it might also
mean that the benchmarking exemption in Section 301.2.1 does not
apply to cases of substantial structural damage. See the commentary
for Section 301.2.1.

404.2 Substantial structural damage to vertical elements of the lateral

force-resisting system. A building that has sustained substantial structural
damage to the vertical elements of its lateral force-resisting system shall be
evaluated in accordance with the applicable provisions of Sections 404.2.1
through 404.2.3.

Exceptions:
1. Buildings assigned to Seismic Design Category A, B or C whose
substantial structural damage was not caused by earthquake need not be
evaluated or rehabilitated for load combinations that include earthquake
effects.
2. One- and two-family dwellings ....

Commentary: Exception 1 does not apply in San Francisco, where

high seismicity prevents any building from being assigned to SDC A,
B, or C. In the 2018 IEBC, this provision is edited for clarity and
renumbered:
405.2.3 Substantial structural damage to vertical elements of the
lateral force-resisting system. A building that has sustained
substantial structural damage to the vertical elements of its lateral
force-resisting system shall be evaluated in accordance with Section
405.2.3.1, and either repaired in accordance with Section 405.2.3.2 or
repaired and retrofitted in accordance with Section 405.2.3.3,
depending on the results of the evaluation.
Exceptions:
1. Buildings assigned to Seismic Design Category A, B or C whose
substantial structural damage was not caused by earthquake need not
be evaluated or retrofitted for load combinations that include
earthquake effects.
2. One- and two-family dwellings need not be evaluated or retrofitted
for load combinations that include earthquake effects.

404.2.1 Evaluation. The building shall be evaluated by a registered

design professional, and the evaluation findings shall be submitted to

ATC-119-1 Part 4: Post-earthquake Structural Evaluation 4-29

 the code official within 60 days of completion of the evaluation. The
evaluation shall establish whether the damaged building, if repaired to
its predamage state, would comply with the provisions of this code for
wind and earthquake loads. Evaluation for earthquake loads shall be
required if the substantial structural damage was caused by or related
to earthquake effects or if the building is in Seismic Design Category C,
D, E or F.

Wind loads for this evaluation shall be those prescribed in San

Francisco Building Code Section 1609. Earthquake loads for this
evaluation, if required, shall be according to section 301.2.

Commentary: SFEBC Section 404.2.1 replaces the similar provision in

the CEBC. For reference, the corresponding wording from the CEBC
and from the 2018 IEBC are given at the bottom of this commentary.
Notes on the SFEBC wording in Section 404.2.1:
• SFEBC includes a 60-day deadline for submittal of the evaluation.
• SFEBC makes an unnecessary (and potentially confusing)
statement about when earthquake loads must be considered. It’s
unnecessary because this SFEBC wording merely duplicates
Exception 1, already provided in Section 404.2. As indicated in the
commentary there, this SFEBC wording is also moot, since every
building in San Francisco would be assigned to SDC D, E, or F.
• SFEBC specifies reduced loads by referring to Section 301.2.
Although not stated here, the additional requirement in Section
404.1.1 regarding the original design code also clearly applies.

Following is the CEBC/IEBC wording that SFEBC replaces:

404.2.1 Evaluation. The building shall be evaluated by a registered
design professional, and the evaluation findings shall be submitted to
the building official. The evaluation shall establish whether the
damaged building, if repaired to its predamage state, would comply
with the provisions of the California Building Code for wind and
earthquake loads. [Commentary: The balance of the section
prescribes full wind loads per CBC Section 1609 but allows reduced
seismic loads, including the ASCE 41 option in CEBC Section
301.1.4.2.]

Following is the 2018 IEBC wording, further clarified and renumbered:

405.2.3.1 Evaluation. The building shall be evaluated by a registered
design professional, and the evaluation findings shall be submitted to

4-30 Part 4: Post-earthquake Structural Evaluation ATC-119-1

 the code official. The evaluation shall establish whether the damaged
building, if repaired to its predamage state, would comply with the
provisions of the International Building Code for load combinations
that include wind or earthquake effects, except that the seismic forces
shall be the reduced seismic forces.

404.2.2 Extent of repair for compliant buildings. If the evaluation

establishes compliance of the predamage building in accordance with Section
404.2.1, then repairs shall be permitted that restore the building to its
predamage state.

Commentary: In the 2018 IEBC, this provision is edited for clarity and
renumbered:
405.2.3.2 Extent of repair for compliant buildings. If the evaluation
establishes that the building in its predamage condition complies with
the provisions of Section 405.2.3.1, then the damaged elements shall
be permitted to be restored to their predamage condition.

404.2.3 Extent of repair for noncompliant buildings. If the evaluation does

not establish compliance of the predamage building in accordance with
Section 404.2.1, then the building shall be rehabilitated to comply with the
applicable provisions of the California Building Code for load combinations
that include wind or seismic loads. [Commentary: The balance of this section
requires full wind loads per CBC Section 1609 if the damage was caused by
wind, but allows original wind loads otherwise. The section also allows
reduced seismic loads.]

Commentary: In the 2018 IEBC, this provision is edited for clarity and
renumbered:
405.2.3.3 Extent of repair for noncompliant buildings. If the
evaluation does not establish that the building in its predamage
condition complies with the provisions of Section 405.2.3.1, then the
building shall be retrofitted to comply with the provisions of this
section. The wind loads for the repair and retrofit shall be those
required by the building code in effect at the time of original
construction, unless the damage was caused by wind, in which case
the wind loads shall be in accordance with the International Building
Code. The seismic loads for this retrofit design shall be those required
by the building code in effect at the time of original construction, but
not less than the reduced seismic forces.

ATC-119-1 Part 4: Post-earthquake Structural Evaluation 4-31

 404.3 Substantial structural damage to gravity load-carrying
components. Gravity load-carrying components that have sustained
substantial structural damage shall be rehabilitated to comply with the
applicable provisions of the California Building Code for dead and live loads.
[Commentary: The balance of this section accounts for snow loads and live
loads and requires consideration of load path from any affected members.]

Commentary: This section is generally unrelated to earthquake,

except for the subsection 404.3.1. Also, in the 2018 IEBC, edited for
clarity and renumbered:
405.2.4 Substantial structural damage to gravity load-carrying
components. Gravity load-carrying components that have sustained
substantial structural damage shall be rehabilitated to comply with the
applicable provisions for dead and live loads in the International
Building Code. Snow loads shall be considered if the substantial
structural damage was caused by or related to snow load effects.
Undamaged gravity load-carrying components that receive dead, live
or snow loads from rehabilitated components shall also be
rehabilitated if required to comply with the design loads of the
rehabilitation design.

404.3.1 Lateral force-resisting elements. Regardless of the level of

damage to vertical elements of the lateral force-resisting system, if
substantial structural damage to gravity load-carrying components
was caused primarily by wind or seismic effects, then the building shall be
evaluated in accordance with Section 404.2.1 and, if noncompliant,
rehabilitated in accordance with Section 404.2.3.

Exceptions:
1. One- and two-family dwellings ....
2. Buildings assigned to Seismic Design Category A, B, or C whose
substantial structural damage was not caused by earthquake need not be
evaluated or rehabilitated for load combinations that include earthquake
effects.

Commentary: The point of this section is to address archaic structures

in which the gravity system elements are acting as a de facto lateral
system, as evidenced by their damage in a wind or seismic event.
The provision points back to Section 404.2.1, indicating that this case
should be treated like a case of substantial structural damage (SSD)
to the lateral system.

4-32 Part 4: Post-earthquake Structural Evaluation ATC-119-1

 There is one unclear aspect to this provision. For criteria, Section
404.2.1 points to Section 301.2, but Section 404.1.1 requires
additional consideration of original design loads. Since Section
404.3.1 points back only to Section 404.2.1, but not to 404.1.1, so
some might argue that the additional requirement there does not
apply. This is almost certainly an oversight in the SFEBC, however,
as the point of this section is to treat this case of SSD the same way
one would treat SSD to the lateral system. If additional criteria apply
to cases subject to Section 404.2.1, the same criteria should apply
here.
In the 2018 IEBC, this provision is edited for clarity and renumbered:
405.2.4.1 Lateral force-resisting elements. Regardless of the level
of damage to vertical elements of the lateral force-resisting system, if
substantial structural damage to gravity load-carrying components
was caused primarily by wind or seismic effects, then the building
shall be evaluated in accordance with Section 405.2.3.1 and, if
noncompliant, retrofitted in accordance with Section 405.2.3.3.
Exceptions:
1. Buildings assigned to Seismic Design Category A, B, or C whose
substantial structural damage was not caused by earthquake need not
be evaluated or retrofitted for load combinations that include
earthquake effects.
2. One- and two-family dwellings need not be evaluated or retrofitted
for load combinations that include earthquake effects.

404.4 Less than substantial structural damage. For damage less than
substantial structural damage, repairs shall be allowed that restore the
building to its predamage state. New structural members and connections
used for this repair shall comply with the detailing provisions of the California
Building Code for new buildings of similar structure, purpose and location.

Commentary: In the 2018 IEBC, this provision is edited for clarity and
renumbered:
405.2.1 Repairs for less than substantial structural damage.
Unless otherwise required by this section, for damage less than
substantial structural damage, the damaged elements shall be
permitted to be restored to their predamage condition.

404.4.1 Disproportionate Damage. Buildings with Disproportionate

Damage shall be subject to the requirements of Section 404.2 for
earthquake evaluation and rehabilitation as if they had substantial

ATC-119-1 Part 4: Post-earthquake Structural Evaluation 4-33

 structural damage to vertical elements of the lateral-force-resisting
system.

Permit application for required rehabilitation work shall be submitted to

the Department within 1 year after the earthquake, and the work shall be
completed as specified in Table B of San Francisco Building Code
Section 106A.4.4.

Commentary: SFEBC Section 404.4.1 is the San Francisco

amendment that implements the definition of disproportionate damage
given in Chapter 2. Since it is not necessarily true that
disproportionate damage will always be less than substantial
structural damage, and for general clarity, it would be better if future
SFEBC editions would make this provision separate from Section
404.4. The 2018 IEBC organization and numbering are
recommended.
By pointing to SFEBC Section 404.2 for the triggered scope and
criteria, this additional upgrade trigger invokes the reduced seismic
loads and other criteria given in SFEBC Section 301.2. Presumably,
by also referencing the requirements for substantial structural damage
to the lateral system, this trigger also invokes the additional
requirement in Section 404.1.1 to consider the original design loads
as well. The SFEBC could be clearer, but it stands to reason that if
additional criteria apply to cases subject to Section 404.2.1, the same
criteria should apply here.
Following is the similar triggering provision in the 2018 IEBC. The
IEBC trigger applies only to SDC D, E, and F, but this limit is moot in
San Francisco, where every building would be assigned to SDC D, E,
or F.
405.2.2 Disproportionate earthquake damage. A building assigned
to Seismic Design Category D, E or F that has sustained
disproportionate earthquake damage shall be subject to the
requirements for buildings with substantial structural damage to
vertical elements of the lateral force-resisting system.

404.4.2. Other damage. For damage less than substantial structural

damage that is not Disproportionate Damage, repairs shall be allowed
that restore the building to its predamage state, based on material
properties and design strengths applicable at the time of original
construction. New structural members and connections used for this

4-34 Part 4: Post-earthquake Structural Evaluation ATC-119-1

repair shall comply with the detailing provisions of this code for new
buildings of similar structure, purpose and location.

Commentary: This San Francisco amendment was perhaps

necessary when the SFEBC introduced disproportionate damage, but
it is no longer needed, as it merely duplicates the intent of
CEBC/IEBC Section 404.4. For future SFEBC editions, the 2018
IEBC organization and numbering are recommended.

ATC-119-1 Part 4: Post-earthquake Structural Evaluation 4-35

Appendix B

References

ASCE, 2017, Seismic Evaluation and Retrofit of Existing Buildings, ASCE/SEI

41-17, American Society of Civil Engineers Structural Engineering
Institute, Reston, Virginia.
CEBC, 2016, 2016 California Existing Building Code, California Code of
Regulations, Title 24, Part 10, California Building Standards
Commission, Sacramento, California.
CCSF, 2011, CAPSS Earthquake Safety Implementation Program Workplan
2012-2042, the City and County of San Francisco, California.
CCSF, 2012a, Post-Earthquake Repair and Retrofit Requirements for
Concrete Buildings, Administrative Bulletin 099, Department of
Building Inspections, San Francisco, California.
CCSF, 2012b, Post-Earthquake Repair and Retrofit Requirements for Wood-
Frame Residential Buildings with Three or More Dwelling Units,
Administrative Bulletin 098, Department of Building Inspections, San
Francisco, California.
CCSF, 2012c, Post-Earthquake Repair and Retrofit Requirements for One-
and Two-Family Units, Administrative Bulletin 100, Department of
Building Inspections, San Francisco, California.
CCSF, 2016, The San Francisco Existing Building Code, the City and County
of San Francisco, California.
CCSF, 2017, Building Façade Inspection and Maintenance, Administrative
Bulletin 110, Department of Building Inspections, San Francisco,
California.
FEMA, 1998, Evaluation of Earthquake Damaged Concrete and Masonry
Wall Buildings, FEMA 306, prepared by the Applied Technology
Council for the Federal Emergency Management Agency,
Washington, D.C.
FEMA, 2000, Recommended Postearthquake Evaluation and Repair Criteria
for Welded Steel Moment-Frame Buildings, FEMA 352, prepared by
the SAC Joint Venture, a partnership of the Structural Engineers
Association of California, the Applied Technology Council, and

ATC 119-1 Part 4: Post-earthquake Structural Evaluation 4-37

 California Universities for Research in Earthquake Engineering, for
the Federal Emergency Management Agency, Washington, D.C.
ICC, 2018, International Building Code, International Code Council, Country
Club Hills, Illinois.
SEAONC EBC Steel Frame Subcommittee, 2004, “Post-Earthquake
Evaluation of Welded Steel Moment Frames: Case Studies Using
FEMA-352,” Proceedings, 2004 SEAOC Convention, Monterey,
California.

4-38 Part 4: Post-earthquake Structural Evaluation ATC 119-1

 PART 5:
Tall Building Effects on
Post-earthquake Recovery
Chapter 1

Introduction

1.1 Background

In an effort to understand the impact that tall building performance may have
on the recovery of the City of San Francisco after a major earthquake, this
Part provides an overview of seismic risks that are disproportionately
associated with tall buildings (i.e., tall building effects), with recommendations
for new policy considerations and further research. This Part corresponds to
Recommendations 2C, 2D, 3A, 3C, and 3G presented in the Summary
Recommendations.

A characterization of building tallness in relevant building codes suggests that

while no unique height threshold is valid to distinguish between “tall” versus
“non-tall” buildings, the fact that different thresholds do exist for fire safety,
seismic or planning amongst other considerations, is an indication that tall
buildings present unique challenges.
This Part addresses aspects of the following tasks described in the CAPSS
Earthquake Safety Implementation Program Workplan 2012-2042 (CCSF,
2011):
• Task A.1.b: Provide information and assistance about insurance
• Task A.6.i: Study fire-related earthquake resilience topics
• Task A.4.f: Update post-earthquake inspection policies and procedures

1.2 Acknowledgements

The findings were developed by means of a literature review and consultation

with a wide range of professionals including design engineers (structural and
building services), building owners and managers, first responders, and the
insurance sector among other stakeholders. ATC gratefully acknowledges
the following parties for the insightful comments and discussion in the
development of this report and its recommendations: Ibbi Almufti, Mary
Comerio, Wayne Gaw, Kevin Geraghty, Prasad Gunturi, Akshay Gupta, Matt
Hansen, Dennis Mulqueeney, John Osteraas, Robert Postel, Marcelo
Ramirez, Nilesh Shome, members of the San Francisco Lifelines Council,
and members of the San Francisco Building Owners and Managers
Association (BOMA).

ATC 119-1 Part 5: Tall Building Effects on Post-earthquake Recovery 5-1

 1.3 Organization

Chapter 2 reviews tall building effects and characterization of building tallness

in current San Francisco policy for post-earthquake structural evaluation.

Chapter 3 discusses safety and emergency response.

Chapter 4 reviews consequences from damage.

Chapter 5 reviews programs and procedures to help accelerate recovery,

including inspection and tagging, insurance, and concepts of re-occupancy
and functional recovery.

Chapter 6 provides a summary of tall building effects identified, along with

recommendations for new policy considerations and further research to
mitigate these effects, and a brief discussion on their relevance to tall versus
all buildings, and future versus existing buildings.

A list of references is provided at the end of this Part.

5-2 Part 5: Tall Building Effects on Post-earthquake Recovery ATC-119-1

Chapter 2

Tall Building Effects

2.1 Overview

There are presently 156 buildings 240 ft or taller either constructed or

permitted for construction in San Francisco (see Part 1). About 60% of these
buildings house commercial (office) occupancies, and 40% are residential or
hotel occupancies. With a large concentration of the tall buildings in the
downtown districts of San Francisco, the tall buildings play a key role in the
socio-economic activity of the City. Tall buildings house large concentrations
of businesses and residents, which raises concerns that seismic damage to
one or more tall buildings can have a significant impact on the building
occupants and the surrounding neighborhood. Events such as the 2010-
2011 Canterbury earthquake sequence in Christchurch, New Zealand
highlight the impact of damaged tall buildings on the business continuity of
districts where buildings are clustered close together. For example, following
the 2011 Christchurch earthquake, the 26-story Hotel Grand Chancellor
sustained severe damage, where the risk of collapse from aftershocks
prompted authorities to set up a cordon around the building with a roughly
300-foot radius, roughly equal to the height of the damaged building (NZPA,
2011). Thus, there were significant indirect losses attributed to business
disruption in surrounding buildings in addition to the direct economic loss
associated with the damaged hotel building.

Roughly 75% of the existing tall buildings in San Francisco were constructed
before 2000, following what are now considered outdated code prescriptive
design requirements that do not necessarily provide consistent levels of
safety from major earthquakes. Tall buildings designed since about 2008,
when San Francisco introduced Administrative Bulletin (AB) 083 (CCSF,
2008), typically follow a performance-based seismic design approach where
building’s structural response under strong earthquakes is explicitly evaluated
using advanced structural analysis. While the design approach is called
“performance-based,” the provisions of AB-083 are generally intended only to
demonstrate compliance with minimum code requirements to ensure life
safety under extreme earthquakes. Thus, while more recently designed
buildings are more reliable in terms of life safety, both older existing and
newly designed tall buildings are susceptible to damage from strong

ATC 119-1 Part 5: Tall Building Effects on Post-earthquake Recovery 5-3

 earthquakes, which could result in extensive downtime and potential need for
demolition and replacement.

In an effort to understand the impact that tall building performance may have
on the recovery of the City of San Francisco after a major earthquake, this
Part provides an overview of seismic risk factors that are disproportionately
present with tall buildings (in contrast with the entire building stock), with
recommendations for new policy considerations and further research to better
understand and control the risks.

2.2 Characterization of Building Tallness

From a structural and seismic perspective, tall buildings are generally

associated with larger drifts, higher overturning forces, and more complex
dynamic response. None of those complexities, however, necessarily makes
a taller building riskier, in part because building codes have implemented
provisions to account for such effects since at least the 1960s. In taller
buildings, the structural design requirements are often governed as much or
more by wind loading as compared to earthquake ground shaking. In
addition, there are countless examples of short buildings behaving poorly in
earthquakes, due to structural irregularities, falling hazards, and collapse-
prone deficiencies.

To the extent that certain seismic deficiencies or potential risks might be

associated with tallness, it is useful to review the precedents in existing
building codes, standards, and policies that categorize buildings by height.
This review includes the 2016 edition of the San Francisco Building Code
(SFBC) and San Francisco Existing Building Code (SFEBC), hereinafter
referred to as the “Current Code,” which reference the 2016 California
Building Code (CBC) and 2016 California Existing Building Code (CEBC).
Two previous editions of the Uniform Building Code (UBC) adopted in San
Francisco are also reviewed. Provisions in or similar to the 1964 UBC would
have applied at the start of the tall building era in San Francisco. Provisions
in or similar to the 1985 UBC would have applied at the peak of the
downtown tall building boom of the 1970s and 1980s. Part 1 presents a
detailed inventory of tall buildings in the City with regard to age.

The CBC defines a “high-rise” building, independent of its use, as one

“having floors used for human occupancy located more than 75 feet above
the lowest floor level having building access.” The roof elevation of a high-
rise building would therefore be about 85 feet above grade, but in this report,
for simplicity, the “high-rise” critical height is defined as 75 feet. The CBC
uses this “high-rise” definition only in provisions for fire safety and egress, so

5-4 Part 5: Tall Building Effects on Post-earthquake Recovery ATC-119-1

although a distinction is made based on height, this has little effect on
earthquake design and makes no implication regarding the seismic
performance of taller buildings.

Guidelines for Performance-Based Seismic Design of Tall Buildings (PEER,

2017), which is widely used for the seismic design of tall buildings in San
Francisco, specifies its scope of application as applying to buildings having
the unique seismic response characteristics of tall buildings including: (1) a
fundamental translational period of vibration significantly in excess of 1
second; (2) high mass participation and lateral response in higher modes of
vibration; and (3) a seismic-force-resisting system with a slender aspect ratio
such that substantial portions of the lateral drift result from axial deformation
of the walls and/or columns as compared to shearing deformation of the
frames or walls. The following sections introduce and discuss structural
response parameters that are considered meaningful measures of “tallness,”
and how past and current building codes have dealt with these parameters:

2.2.1 Building Period

The building period is a characteristic vibrational property of a building,

corresponding to the time for the building to undergo one cycle of lateral sway
motion. Assuming that a building period of 2 seconds qualifies as tall by the
terms of the PEER Guidelines, i.e., “significantly in excess of 1 second,” and
applying the current building code’s default formula for estimating the
fundamental period, then the building heights that the PEER Guidelines might
consider “tall” range from about 200 ft for moment frame systems to over 400
ft for stiffer wall and braced frame systems.

2.2.2 High Mode Effects

Higher mode effects cover building lateral drifts and accelerations and
response properties associated with building vibrational modes with higher
frequencies (shorter periods) than the fundamental vibration mode. Through
the equivalent lateral force method, Current Code implicitly considers higher
mode effects for every building. In past codes, higher mode effects were
captured by modifying the vertical distribution of lateral design forces for
certain buildings.

At the time of the downtown development boom of the 1970s and 80s, the
UBC required consideration of high mode effects only where the building’s
fundamental period exceeded 0.7 seconds. Using1985 UBC’s formulas for
period, this corresponds to a structure height of about 7 stories (80 ft) for a
ductile moment frame structure and up to about 12 stories (130 ft) for stiffer
systems.

ATC 119-1 Part 5: Tall Building Effects on Post-earthquake Recovery 5-5

 2.2.3 Aspect Ratio

This is the ratio of the building height to the plan view dimension at ground
level. Because of its formula for estimating period, the critical heights listed in
1985 UBC for considering higher mode effects were related to the building’s
aspect ratio. In earlier codes from the 1970s, the higher mode effect was
ignored for any building with an aspect ratio less than 3:1. In the 1964 UBC,
higher mode effect was ignored for any building with an aspect ratio up to 5:1.

Some of San Francisco’s oldest buildings still show a typical lot width of
about 25 ft, suggesting a critical height of 125 ft in the 1960s or 75 ft in the
1970s. But the taller buildings from the 1960s and later do not have a typical
width, thus the critical height by contemporary codes would have been
significantly greater, but highly varied.

Aspect ratio was also considered as an indicator of higher floor accelerations

in the design requirements for nonstructural components such as rooftop
tanks, chimneys, and penthouses. The 1964 UBC increased the design force
for these components by 50% when the building’s aspect ratio exceeded 5:1.
In the 1985 UBC, this requirement changed to a more general requirement to
consider actual dynamic properties of the component and the building
structure.

2.2.4 Story Drift Ratio

The story drift ratio is a measure of the lateral displacement of a building

under earthquakes, calculated as the peak lateral drift occurring over one
story, divided by the story height. At the peak of tall building construction, the
1985 UBC set a peak elastic story drift ratio limit of 0.005 (roughly equivalent
to an inelastic drift ratio limit of 0.025), but it was independent of the structural
system, dynamic characteristics, or height. At the start of the tall building era,
the 1964 UBC required only that “lateral deflections or drift ... shall be
considered in accordance with accepted engineering practice.” The 2017
PEER Guidelines limit the inelastic story drift ratio under Maximum
Considered Earthquake (MCE) ground motions to 0.03, which is roughly
equivalent to the limit of 0.02 on design story drift ratio of the current CBC
and ASCE/SEI 7-16, Minimum Design Loads and Associated Criteria for
Buildings and Other Structures (ASCE, 2017).

2.2.5 System Ductility Requirements

These are building code requirements that limit the applicability of structural
systems and materials, based on their ability to deform and absorb energy
during a large earthquake. Aside from the criteria suggested by the PEER

5-6 Part 5: Tall Building Effects on Post-earthquake Recovery ATC-119-1

Guidelines, the building code has long used height limits to guide the
selection of seismic-force-resisting systems, as follows:
• 1964 UBC Section 2312(j): “Buildings more than 13 stories or one
hundred and sixty feet (160’) in height shall have a complete moment
resisting space frame ... made of a ductile material or a ductile
combination of materials.”
• 1985 UBC Section 2312(j)1.B: “Buildings more than 160 feet in height
shall have ductile moment-resisting space frames ....”

Instead of assigning systems to buildings of certain heights, current code

provisions specify height limits to buildings with different systems (ASCE/SEI
7-16 Table 12.2-1):
• For typical office, hotel, or residential buildings in high seismic areas
(Seismic Design Category D or E), ductile moment frames have no height
limit, whereas ductile shear wall and braced frame systems are limited to
240 ft, or 160 ft if the building has an extreme torsional irregularity or an
unbalanced configuration.
• For essential facilities (Risk Category IV and Seismic Design Category F),
those same structural systems are limited further, to 160 or 100 ft, if the
building does not have extreme torsional irregularity or an unbalance
configuration.

2.2.6 Collapse-Prone Existing Buildings

San Francisco has three programs that mandate retrofit of existing buildings.
SFEBC Chapters 4A and 4B cover the mandatory retrofit of unreinforced
masonry bearing wall buildings, but their requirements are independent of
building height. Similarly, SFEBC Section 329 mandates seismic evaluation
of private schools, but it makes no distinctions based on building height.

SFEBC Chapter 4D covers the mandatory retrofit of woodframe residential

“soft-story” buildings. Only buildings three stories tall (including two above
grade stories and a basement) or taller are required to comply.

2.2.7 Importance or Risk Category

The building code acknowledges differences between building uses by

assigning every new building to a risk category. Different Risk Categories are
then subject to different design criteria, including the use of Importance
Factors to amplify the Design Earthquake loads.
• 1964 UBC had no Importance Factors and no special earthquake design
provisions even for what are now considered essential facilities.

ATC 119-1 Part 5: Tall Building Effects on Post-earthquake Recovery 5-7

 • 1985 UBC increased design forces by 25% for assembly occupancies,
and by 50% for essential facilities, but neither category was linked to
building height.
• Current Code (2016 CBC Table 1604.5) requires a 25% increase in
design forces for buildings assigned to Risk Category III: “Buildings ...
that represent a substantial hazard to human life in the event of a failure,”
regardless of height.
• Current Code also requires a 50% increase, as well as consideration of
post-earthquake functionality, for hospitals, fire stations, emergency
shelters, and other “essential facilities” assigned to Risk Category IV,
regardless of height.

In summary, the building code’s conception of importance and risk is only

loosely connected with building height. In concept, a collapse-prone building
of almost any size could be said to represent “a substantial hazard to human
life,” but in practice, buildings are only assigned to Risk Category III if they
match one of the examples given in the code. None of those examples is
specifically related to building height, but the category does include “any
other occupancy with an occupant load greater than 5,000.” To establish a
building’s occupant load for design, the code presumes 100 net square feet
per occupant in an office or similar area. Using that value, an occupant load
of 5,000 would only occur in a building with at least 500,000 net square feet,
or a gross square footage of at least 750,000 square feet.

Per San Francisco Property Information Map (CCSF, 2017a), only eight
buildings within the inventory presented in Part 1 are office buildings with
building areas of at least 750,000 square feet, and all of them are over 400
feet tall. Thus, it can be assumed that in San Francisco, the building code
envisions office buildings taller than about 400 feet to be candidates for
assignment to Risk Category III. Per Current Code, such a building would
then be subject to the same additional requirements of other Risk Category III
facilities, including restaurants or assembly halls with seating for 300. Due to
lower occupant density, tall residential buildings would rarely, if ever, trigger
the Risk Category III criteria.

2.2.8 Structural Integrity

Current Code (2016 CBC Section 1615) has special provisions for structural
integrity meant to provide additional resistance to collapse, usually
associated with blast or other unanticipated loads. They apply only to “high-
rise buildings that are assigned to Risk Category III or IV.”

5-8 Part 5: Tall Building Effects on Post-earthquake Recovery ATC-119-1

The current CBC code defines “high-rise building” to include most buildings
taller than 75 feet. As noted above, only a building approaching 400 ft tall
would be assigned to Risk Category III on the basis of its height alone. Even
so, the code’s association of special safety provisions with height does imply
a systematic categorization of “tall” buildings.

2.2.9 Falling Hazards

The earthquake design requirements of building codes have provisions that

apply to the design and installation of architectural components (façade and
ceiling systems, partition walls), mechanical components (mechanical
equipment) and other nonstructural components that can represent potential
falling hazards. As with its drift limits, the building code’s nonstructural
provisions are independent of the building’s height.

In 1969, San Francisco mandated bracing for unreinforced masonry parapets

(2016 SFEBC Chapter 4C), but the requirements were independent of
building height.

San Francisco has also implemented a mandatory façade inspection program

(2016 SFEBC Chapter 4E) to identify damaged, deteriorated, or otherwise
vulnerable cladding conditions that represent earthquake falling hazards.
The program applies to buildings five stories and taller.

2.2.10 Fire Safety

As noted above, Current Code (2016 CBC Section 403) imposes special fire
safety and egress requirements for high-rise buildings, defined as those taller
than about 75 feet. Within Section 403, the code sets another line at 420 ft;
buildings taller than that are limited in the types of fire-resistive construction
they may use. The 1985 UBC (Sec 1807(a)) made a similar requirement for
office and general use buildings (Group B, Division 2) and for hotels and
apartment houses (Group R, Division 1).

Also related to fire resistance, but in the building code chapter on allowable
building heights and areas, 2016 CBC Table 504.3 makes height distinctions
for different types of construction with and without sprinkler systems. For
most occupancies:
• 160 to 180 ft, for Type I.B construction, non-sprinklered and sprinklered,
respectively
• 65 to 85 ft, for Type II, III, or IV construction, non-sprinklered and
sprinklered, respectively
• 50 to 70 ft, for Type V construction (woodframe, generally), non-
sprinklered and sprinklered, respectively

ATC 119-1 Part 5: Tall Building Effects on Post-earthquake Recovery 5-9

 where construction Types I and II are those in which the building elements
are of noncombustible materials; Type III is that in which the exterior walls
are of noncombustible materials and the interior building elements are of any
material permitted by the code; Type IV is that in which the exterior walls are
of noncombustible materials and the interior building elements are of solid or
laminated wood without concealed spaces; Type V is that in which the
structural elements, exterior walls and interior walls are of any materials
permitted by this code; Type A refers to fire-protected systems; and Type B
refers to unprotected ones. See International Building Code (ICC, 2016)
Section 602 for more detailed definitions.

1985 UBC Table 5-D sets similar limits for office buildings (Occupancy Group
B) and hotels (Occupancy Group R-1):
• 160 ft or 12 stories, for Type II fire resistive construction
• 65 ft or 4 stories, for Type II, III, or IV one-hour construction
• 50 ft or 3 stories, for Type V one-hour construction

1964 UBC Table 5-D height limits for office buildings (Occupancy Group F-2)
and hotels or apartment houses (Occupancy Group H) were slightly more
restrictive:
• 95 ft or 6 stories (5 stories for H occupancy), for Type II construction
• 65 ft or 4 stories, for Type III or IV one-hour construction
• 50 ft or 3 stories, for Type V one-hour construction

2.2.11 Planning and Zoning

In addition to building code requirements, San Francisco sets height and bulk
limits for purposes of planning and zoning. In the northeast portion of the
city, which includes the downtown and adjacent residential neighborhoods,
these are shown on San Francisco Zoning Map Sheet HT01 (CCSF, 2017b).
The map includes special block-size zones for specific projects, and in most
downtown neighborhoods the limits vary from block to block. Some
neighborhood have consistent limits:
• 40 ft: Russian Hill, North Beach
• 50 ft: Chinatown
• 65 ft: Nob Hill
• 80 ft: Union Square
• 150 ft: Second Street

5-10 Part 5: Tall Building Effects on Post-earthquake Recovery ATC-119-1

2.2.12 Summary Table

Table 5-1 summarizes the precedents discussed. Although each of the

height values in the table has some meaning, the main point of the table is to
illustrate there is no single height value that distinguishes “tall” buildings from
“non-tall” buildings for all circumstances.

Table 5-1 Height Precedents in Past and Current Codes and Policies
Critical
Height Issue Source and Description
420 ft Fire Safety Current Code (2016 CBC Section 403): Taller buildings must meet all fire-resistance
requirements for Type I.A construction, as well as additional egress requirements
~400 ft Risk Category Current Code: Approximate height at which an office building would be assigned to Risk
Category III based on occupant load of 5,000
240 ft Risk Category, Current Code: Height limit for shear wall and braced frame systems in normal occupancy
seismic, structural without torsion-prone configuration
160 – 180 ft Fire Safety Current Code (2016 CBC Table 504.3): Maximum height for Type I.B construction*, non-
sprinklered or sprinklered, respectively
160 ft Seismic, structural 1985 UBC: Taller buildings must have a ductile moment-resisting frame
160 ft Risk Category, Current Code: Height limit for shear wall and braced frame systems in normal occupancy
seismic, structural with torsion-prone configuration
Current Code: Height limit for shear wall and braced frame systems in essential facilities*
without torsion-prone configuration
13 stories Seismic, structural 1964 UBC: Taller buildings must have a ductile moment-resisting frame
12 stories or Fire Safety 1985 UBC (Table 5-D): Maximum height for Type II construction
160 ft
150 ft Planning and SF Zoning Map Sheet HT01 (CCSF, 2017b): Height limit in Second Street neighborhood
Zoning
80 – 130 ft Seismic, structural 1985 UBC: Structure subject to provision for higher mode effects
100 ft Risk Category, Current Code: Height limit for shear wall and braced frame systems in essential facilities*
seismic, structural with torsion-prone configuration
95 ft or Fire Safety 1964 UBC (Table 5-D): Maximum height for Type II construction (6-story limit for office
5 – 6 stories buildings; 5-story limit for hotels and apartment houses)

65 – 85 ft Fire Safety Current Code (2016 CBC Table 504.3): Maximum height for Type II, III, or IV construction,
non-sprinklered or sprinklered, respectively
80 ft Planning and SF Zoning Map Sheet HT01 (CCSF, 2017b): Height limit in Union Square neighborhood
Zoning
75 ft Fire Safety Current Code (2016 CBC Section 403): “High-rise buildings” must comply with special fire
safety and egress provisions.
1985 UBC (Section 1807a): Taller office buildings, hotels, and apartment houses must
comply with special fire safety and egress provisions
75 ft Risk Category; Current Code: “High-rise buildings” assigned to Risk Category III or IV must comply with
Structural Integrity structural integrity provisions
7 stories Seismic, structural 1985 UBC: Moment frame structure subject to provision for higher mode effects

ATC 119-1 Part 5: Tall Building Effects on Post-earthquake Recovery 5-11

Table 5-1 Height Precedents in Past and Current Codes and Policies (continued)
Critical
Height Issue Source and Description
50 – 70 ft Fire Safety Current Code (2016 CBC Table 504.3): Maximum height for Type V construction, non-
sprinklered or sprinklered, respectively
65 ft Planning and SF Zoning Map Sheet HT01 (CCSF, 2017b): Height limit for Nob Hill neighborhood
Zoning
5 stories Seismic, Current Code (2016 SFEBC Chapter 4E): Façade inspection required for buildings of this
nonstructural height or taller
4 stories or 65 ft Fire Safety 1985 UBC (Table 5-D): Maximum height for Type II, III, or IV one-hour construction
1964 UBC (Table 5-D): Maximum height for Type III or IV one-hour construction
50 ft Planning and SF Zoning Map Sheet HT01 (CCSF, 2017b): Height limit in Chinatown neighborhood
Zoning
40 ft Planning and SF Zoning Map Sheet HT01 (CCSF, 2017b): Height limit in Russian Hill and North Beach
Zoning neighborhoods
3 stories or 50 ft Fire Safety 1964 and 1985 UBC (Table 5-D): Maximum height for Type V one-hour construction
2 stories plus Seismic, structural Current Code (2016 SFEBC Chapter 4D): “Soft story” buildings of this height or taller must
basement be retrofitted.
* Essential facilities corresponding to Risk Category IV (RC IV) and Seismic Design Category F (SDC F). See International Building
Code (ICC, 2016) Table 1604.5 for RC and Section 1613 for SDC classifications.

5-12 Part 5: Tall Building Effects on Post-earthquake Recovery ATC-119-1

Chapter 3

Safety and Emergency Response

3.1 Cordoning

Following moderate to severe earthquakes, there is risk of falling hazards due

to structural and nonstructural building component damage. An earthquake
damaged building with significant permanent deformations may pose a
significant collapse risk under the threat of aftershocks. Similarly,
nonstructural components, such as façades and building appendages, could
be further damaged and pose a life-safety hazard on surrounding areas due
to risk of falling.

These public safety reasons can lead to imposing block-long barricades, or

cordons, around large areas. In addition to falling hazards and the collapse
risk of one or more tall buildings, other reasons that may warrant such
cordons include: shoring, stabilization, and demolition activities; extensive
debris; damaged utilities; unstable geological features (e.g., landslides or
liquefaction); and vandalism.

Recent earthquake disasters have highlighted gaps in safety and awareness

for cordoning practices. For instance, in the Canterbury earthquake
sequence in 2010-2011 in Christchurch, New Zealand, previously damaged
buildings that had been partially stabilized or cordoned, subsequently
collapsed during an aftershock, killing four pedestrians and eight people in a
bus (CALBO, 2013).

3.1.1 Current Guidance

California Building Officials (CALBO) have developed interim guidelines for

barricading, cordoning, emergency evaluation and stabilization of buildings
with substantial damage in disasters (CALBO, 2013). This section provides a
review of the guidance.

After damaging events, government personnel will deploy to districts within

the jurisdiction that are reported to have experienced damage (or districts
known to be vulnerable). Those personnel first to arrive at a scene of severe
damage should take steps to initiate the evacuation of people out of harm’s
way and prevent further access. As outlined in the CALBO (2013) guidelines,
because first responders may be initially overwhelmed in the immediate

ATC 119-1 Part 5: Tall Building Effects on Post-earthquake Recovery 5-13

 aftermath of a damaging earthquake, Department of Building Inspection (DBI)
personnel, must be trained to have appropriate levels of knowledge, skills
and experience to:
• Identify structures that are obviously or suspected to be damaged
• Reduce the public’s exposure to risks
• Set up temporary cordons at safe horizontal distances of falling or
collapse hazards

Safe distances for initial barricades, as illustrated in Figure 5-1, are generally
conservatively set at 1.5 times the height of falling hazards to allow for the
possibility that falling items could bounce and shatter (NIOSH, 2009). Once
the nature and extent of building damage is further investigated, shorter safe
distances may be justified. While these safe distances for initial barricading
have limited implication on surrounding areas when applied to low to mid-rise
buildings, they could imply closing off entire city blocks when applied to tall
buildings, as illustrated in Figure 5-2. Cordons around tall buildings are more
likely to interfere with San Francisco’s emergency priority routes.
Furthermore, if cordons around multiple tall buildings are required, this could
lead to large areas within the Financial District being closed off, where most
of the city’s tall buildings are clustered together.

Figure 5-1 Initial cordoning safe distances (CALBO, 2013).

It is important that the City establish a lead agency, as well as clear roles and
responsibilities associated with post-earthquake evaluations and the ensuing
cordons around damaged buildings. Furthermore, guidelines for cordoning
around tall buildings should be developed prior to the earthquake, in
coordination with relevant groups, including building department, lifelines, fire,

5-14 Part 5: Tall Building Effects on Post-earthquake Recovery ATC-119-1

police, public works, and emergency management staff. The City should also
carry out table-top exercises to familiarize staff with the challenges and
procedures likely to follow disasters. Per CALBO (2013), the absence of a
lead agency appointed by ordinance places the Building Official in the lead
for these actions and requires emergency stabilization, debris removal, and
barrier installation work to be initially paid by the jurisdiction and later
recovered through legal action against building owners.

Figure 5-2 Illustration of a tall building cordon around a

hypothetical 520-ft tall building with a cordon radius
equal to 1.5 times the building height.

Following CALBO recommendations for complex buildings and critical

facilities, the City should supplement available staff and expertise through
mutual aid to speed emergency response and recovery by expediting
barricading, emergency stabilization and falling hazard removal activities. In
addition, establishment of a Tall Buildings Task Force to conduct the
following work supplement other available resources as follows:
1. Assist in conducting rapid assessments of tall buildings.
2. Manage detailed investigations of damaged tall buildings, their potential
fall zones, and how they might be prevented from further collapse.
3. Refine and confirm safe distances from tall building barricades and
cordons.

Emergency access to archives of plans for major existing structures may be a

critical source of intelligence even in the first hours after disasters (CALBO,
2013). This observation further highlights the relevance of maintaining and

ATC 119-1 Part 5: Tall Building Effects on Post-earthquake Recovery 5-15

 expanding the existing tall building database presented in Part 1 to include
building drawings that may benefit emergency inspections and response.

Depending on the severity of the earthquake, cordons may remain in place

for extended periods of time (potentially years). The expected duration of
temporary stabilization may influence decisions about design criteria and
permit processing to develop permanent repairs. Plans should be put in
place to enable permit applications for demolitions, long term barricade
locations, and stabilization works.

Tall building façade systems that suffer significant damage during an

earthquake are an obvious life-threatening falling hazard. Stiff curtain wall
panels attached to the exterior of a building may have insufficient lateral
deformation capacity to accommodate the lateral interstory drift imposed on
the building by the earthquake ground motion input (Filiatrault et al., 2002).
This problem is mostly likely relevant to buildings with flexible seismic-force-
resisting systems, such as steel moment-resisting frame buildings, which are
the most prevalent system in San Francisco’s tall building inventory.

Following a review of damage suffered by heavy cladding panels during the

1994 Northridge earthquake, Cohen (1995) indicated that efforts should be
undertaken to provide improved engineering details for these elements, which
have been previously ignored in design. Following the 1985 Mexico City
earthquake, Goodno et al. (1989) found that in many cases, cladding systems
increased the initial stiffness of the building before suffering extensive
damage. While cladding systems in modern high-rise buildings are designed
to be isolated from the structure, this may not be the case in older buildings.
Where the cladding is not isolated, it may initially stiffen the building, which
can affect (and potentially improve) the dynamic response. But, once the
cladding strength is exceeded, the imposed forces and deformations can
result in significant damage and falling hazards. In light of these
observations, the City should consider utilizing inspections and reports
resulting from San Francisco’s façade ordinance implemented in 2016 San
Francisco Existing Building Code to characterize the façade vulnerability and
risks of falling hazards. A related effort should evaluate the impact of the
associated debris on emergency priority routes (in coordination with the San
Francisco Lifelines Council).

The decision to cordon parts of a jurisdiction is not one to be made lightly, as

it can have disproportionate impacts on the urban community, and the
broader socio-economic factors that affect resilience. While the overriding
issue of the safety of the public must be sufficient to warrant the closure of a
road or section of a city, these effects must be considered when developing

5-16 Part 5: Tall Building Effects on Post-earthquake Recovery ATC-119-1

cordoning guidelines to enable evaluation methods for progressive reduction
of the cordoned area in order to expedite recovery.

3.2 Evacuation and Fire Following Earthquake

Most high-rise buildings have relatively small floor plan areas with travel
distances to exit stairways that are well within the code requirements for
evacuation time. However, emergency egress for high-rise buildings
presents special challenges for vertical evacuation, including the cumulative
occupant load leading to stairway congestion and the physical ability for
people to descend many flights of stairs. Current design standards require
buildings to be designed such that occupants are able to evacuate the
building without outside assistance in a time not exceeding half of the
required fire resistance time of the primary structural frame. This equates to
evacuation times of 1 to 1.5 hours based on fire resistance ratings of 2 to 3
hours for Type IB or IA construction (Bukowski, 2009).

The Current Code requires local water storage in buildings that is sufficient to
operate automatic suppression systems for 30 minutes. The rationale behind
this is that the responding fire department is expected to be able to access a
fire on any floor and begin suppression operations within 30 minutes of the
transmission of the original fire alarm (Bukowski, 2009). The expected 30-
minute fire department response time has an expectation of travel time to the
building and building access through a fire service elevator that has
emergency power backup. In the event of a fire following an earthquake in a
high-rise building, there are reasons to question the assumed 30-minute
response time, due to: (1) competing demands for fire department services
in the City; (2) travel time delays due to road blockages; and (3) building
damage that may affect access through the fire service elevator and
stairwells. Therefore, to the extent that the automatic building fire
suppression systems may become a primary line of fire defense following an
earthquake, the minimum local water supply requirement to provide 30
minutes service should be re-evaluated to: (1) control fire spread; and
(2) ensure safe building evacuation following a large earthquake.

A related topic concerns recent changes in building codes that permit the use
of specially equipped passenger elevators for evacuation of tall building
occupants for during a fire emergency (Kinateder et al., 2014). The seismic
risk in San Francisco (and other west coast cities) raises questions as to
whether elevators will function as intended after earthquakes. While ASME
A17.1 (2014) provides earthquake design requirements for elevators,
including seismic and counterweight displacement switches, and additional
visual and audible information systems, without period testing, these systems

ATC 119-1 Part 5: Tall Building Effects on Post-earthquake Recovery 5-17

 may not perform as intended following a large earthquake (Kinateder et al.,
2014). To the extent that elevators are relied upon for fire emergency
evacuation of mobility impaired occupants (or building occupants in general),
it is suggested to evaluate the risk that earthquake damage will cause
elevators to become inoperable and implications on the safety of occupants
in residential and office buildings.

3.3 Search and Rescue

Collapse prevention is the key underlying goal of earthquake design

provisions in building codes. Over the past century, building code
requirements for seismic safety have evolved based on observations of
building performance in past earthquakes combined with laboratory tests of
building components and consensus judgements of engineers, researchers,
and building code officials. Historically, the collapse risk has not been
explicitly quantified but instead is assumed to be sufficient for structures
designed to standards specified by building codes. Modern design guidelines
have recently become more explicit in the expected collapse safety of
buildings. For example, ASCE/SEI 7-16 (ASCE, 2017) indicates that building
structures designed and constructed in accordance with requirements for
Risk Category II buildings (which applies to most new tall building designs),
should have less than a 10% probability of collapse under the code-specified
Maximum Considered Earthquake (MCE) ground motions.

While it is generally assumed that the actual collapse risk in tall buildings is
less than the maximum permitted risk, there is very little information to
validate the collapse risk of buildings. For example, studies such as Haselton
and Deierlein (2007) have shown that buildings designed for the same site,
using the same code, materials, and structural system can have different
collapse risks due to differences in design decisions, such as structural layout
or the distribution of strength and stiffness over the height. This means that
without a detailed analysis, the collapse risk of a particular structure remains
largely unknown (Eads, 2013).

Collapse risk studies of buildings designed per modern building codes

indicate that collapse safety margins set by the building code are generally
met. On the other hand, a comparison of older non-ductile concrete buildings
(representative of California construction between 1950 and 1975) and
modern buildings, Liel and Deierlein (2013) found that the average mean
annual risk of collapse was about 40 times greater for older, non-ductile
concrete buildings as compared to new concrete buildings with respect to
modern collapse safety margins. Similar studies applied to tall steel moment-
resisting frame buildings suggest that the mean annual frequency collapse

5-18 Part 5: Tall Building Effects on Post-earthquake Recovery ATC-119-1

risk of older (1970s era) buildings is 28 times greater than equivalent modern
designs (Molina Hutt et al., 2018). Insofar as the tall building database
indicates the building inventory to include many older steel moment-resisting
frame buildings and several non-ductile concrete buildings, emergency
personnel, including Urban Search and Rescue (USAR) teams, should be
prepared to respond to building collapses under a strong earthquake.

Collapse is still the principal cause of death in the most large-scale

earthquake disasters. Although the collapse of modern tall buildings can be
generally regarded as a low probability-high consequence event, due to high
occupant loads and because older design codes did not explicitly quantify
protection against collapse, the collapse risk of older tall building cohorts
identified in the Part 1, should be investigated in further detail. More
importantly, emergency personnel, including Urban Search and Rescue
(USAR) teams, should be prepared to respond to such event.

USAR personnel and equipment required for any collapsed building depends
mainly on the construction and the damage type of the building, the building
size, the degree of the collapse and the number of casualties (Schweier and
Markus, 2004). One of the key factors in allocating search and rescue
resources is knowledge about the location, the extent, and the damage
characteristics of the totally or partially collapsed building. Different damage
types require different resources, as inferred by the damage types illustrated
in Figure 5-3. At large building collapses, the required number of rescue
personnel is generally dependent on the probable number of trapped
persons. Prior knowledge of expected occupancies of potentially vulnerable
tall buildings could help an effective mobilization of USAR resources.

Figure 5-3 Damage types by Schweier and Markus (2004).

ATC 119-1 Part 5: Tall Building Effects on Post-earthquake Recovery 5-19

 In carrying out search and rescue operations in a large collapsed building
within a dense city center, USAR teams may also face challenges in carrying
out the necessary structural stabilization tasks. FEMA USAR training
guidelines (FEMA, 2009) generally specify lateral forces for shoring at a
minimum of 2%, but ideally up to 10% of the weight to be stabilized.
Depending on the collapsed condition of a tall building, these guidelines could
result in prohibitively high lateral loads. It should also be noted that longer
term stabilization structures should have more comprehensive design criteria
than those established by search and rescue teams, since the latter may be
inadequate for long term exposure by the public.

A related question is whether search and rescue resources are sufficient to

cope with a large disaster in the dense urban downtown of San Francisco.
Currently in California there are 8 USAR task forces: two in Los Angeles, one
in Orange County, one in the City of Riverside, one in Menlo Park, one in
Oakland, one in Sacramento, and one in San Diego. Although teams
throughout the state will be able to assist in the event of a devastating
earthquake, the City of San Francisco should study the benefits of
establishing its own USAR task force as teams in closest proximity (Oakland
and Menlo Park) may be needed in their own jurisdiction in the event of a
major earthquake.

International guidelines for USAR activities are well established in INSARAG

(OCHA, 2012). However, these guidelines include no explicit discussion on
the recommended procedures in the event of tall building collapse. The City
should establish recommended criteria and propose relevant training
exercises that address these unique challenges, at least for USAR task force
teams anticipated to respond within San Francisco.

5-20 Part 5: Tall Building Effects on Post-earthquake Recovery ATC-119-1

Chapter 4

Consequences from Damage

4.1 Direct Costs to Owners and Tenants

Direct losses from earthquake damage include the financial costs of post-
earthquake repair, demolition, and reconstruction work. Given that the
primary objective of seismic building codes is to ensure life-safety under
extreme earthquake events, significant damage to structural and
nonstructural building components is likely to occur under moderate to
extreme earthquakes. While new buildings are expected to perform better
than older buildings, damage is likely to occur in both.

Structural components of a building include the structural framing (beams,

columns, bracing, shear walls, and floor systems; typically consisting of steel
or concrete) to resist gravity, earthquake, wind, and other types of loads. The
structural system in a building is typically analyzed and designed by a
structural engineer. Nonstructural building components include all of the
architectural walls and interior finishes, cladding, along with the mechanical,
electrical, and plumbing systems. Building contents, including furniture,
moveable partitions, and equipment, are a third distinct component of
buildings. The nonstructural components and contents are normally specified
by architects, mechanical engineers, electrical engineers, or interior
designers, but could also be purchased and installed directly by owners or
building occupants. Many of the nonstructural building components and
contents are vulnerable to earthquake damage (even if the structural system
performs as intended) and they often account for a significant portion of direct
financial losses under earthquakes. To put their relative costs in perspective,
the structural framing and foundation systems typically comprise about one-
quarter of the building construction cost, with the remainder being
nonstructural components and contents. Thus, in evaluating the risk of
damage and direct financial losses due to earthquakes, it is important to
include the potential damage to nonstructural components and contents and
ways that these can be reduced through measures to design or improve their
ruggedness against earthquakes (FEMA, 2012b).

Following the 1971 San Fernando earthquake, a survey of 50 damaged high-

rise buildings, which were located far away from the fault rupture and
experienced only mild shaking, showed that whereas none had major

ATC 119-1 Part 5: Tall Building Effects on Post-earthquake Recovery 5-21

 structural damage, 43 of the buildings suffered damage to drywall or plaster
partitions, 18 suffered damaged elevators, 15 had broken windows, and 8
incurred damage to their air-conditioning systems (Steinbrugge and Schader,
1973). These observations highlight that even under mild shaking, tall
buildings may suffer considerable losses due to damage to nonstructural
components.

Researchers have developed approaches to improve loss-estimating

methods for individual buildings and the repair times associated with a given
level of damage. The FEMA P-58 methodology and models provide the
means for estimating direct economic losses and repair time as a result of
damage to both structural and nonstructural components (FEMA, 2012a).
Recent studies on the seismic performance evaluation of tall buildings
indicate that older existing tall steel moment-resisting frame office buildings
are expected to suffer losses in the order of 34% of building replacement cost
under design earthquake ground motions (Molina Hutt et al., 2016) and 11%
of building replacement cost under a magnitude-7.0 earthquake on the
Hayward fault (USGS, 2018). Similarly, modern reinforced concrete shear
wall residential buildings are expected to experience losses under a design
earthquake ground motions on the order of 15% of building replacement cost
(Tipler and Deierlein, 2014) and 5% of building replacement cost under a
magnitude-7.0 earthquake on the Hayward fault (USGS, 2018).

Designing buildings to sustain less damage in earthquakes significantly

decreases the uncertainty in the behavior of the building and increases the
confidence that the building will perform as intended (Arup, 2013). The
FEMA P-58 approach to estimate building losses enables separating losses
into their constituent parts, making it possible to determine the contribution of
each component to the overall loss, and permits comparison of different
structural and nonstructural mitigation strategies. Performance-based
seismic design, as currently implemented through San Francisco’s AB-083
for the design of tall buildings requires minimum safety criteria for the
structural system and building cladding, but it does not address the design
and performance verification of nonstructural components. Discussions with
engineers who design the building mechanical, electrical, and plumbing
systems highlight the need to establish guidelines and criteria to address the
effect that damage to nonstructural components on building occupancy and
functionality. Where nonstructural components are designed for earthquake
effects, typically the design criteria are limited to addressing the risk of falling
hazards (e.g., by seismic anchorage or bracing), and the criteria do not
address functionality of the components (e.g., whether they can remain
operational or be repaired promptly after an earthquake). Part 3 addresses

5-22 Part 5: Tall Building Effects on Post-earthquake Recovery ATC-119-1

the need for greater emphasis on the design and installation of nonstructural
components in tall buildings to ensure enhanced overall seismic
performance. Part 3 also addresses some of the challenges in implementing
existing methods such as the FEMA P-58 methodology to loss assessments
in tall buildings.

4.2 Indirect Costs to Owners and Tenants

While seismic loss estimates associated with direct economic losses enable
discussions with building owners and investors about how individual retrofit
interventions can move buildings in the direction of becoming more resilient,
they do not provide a quantitative measure of resilience. In addition to direct
economic losses, there is great vulnerability to indirect economic losses due
to downtime, defined as the time required to achieve a recovery state after an
earthquake. Bonowitz (2011) defines three recovery states as follows: (1) re-
occupancy of the building; (2) functional recovery; and (3) full recovery. Re-
occupancy occurs when the building is deemed safe enough to be used for
shelter, although functionality may not be restored. Functional recovery
occurs when the building regains its primary function, i.e., it is operational but
perhaps with some limitations or inconveniences. Full recovery occurs when
the building is fully restored to its pre-earthquake condition.

Comerio (2006) identified rational and irrational components of downtime for

buildings. Rational components include ones that can be more readily
quantified, primarily repair construction costs and time. Irrational, situation-
specific components take into account the time needed to plan and mobilize
for repairs, and include financing, relocation of functions, workforce
availability, regulatory changes, and economic uncertainty. Comerio also
noted that delays due to owner decision making (indecision) is another
important irrational component, which is very difficult to quantify ahead of
time. To help quantify these factors, the pace of recovery can be linked to the
scale of damage in a stock of buildings in the affected region.

The Resilience-based Earthquake Design Initiative (REDi) for the Next

Generation of Buildings (Arup, 2013) guidelines provide a detailed downtime
assessment methodology for individual buildings and identify the likely
causes of downtime such that these can be mitigated to achieve a more
resilient design. The methodology identifies the extent of damage and
criticality of building components that may hinder achieving a recovery state.
If the damage in any component hinders achieving a certain recovery state,
the component needs to be repaired before such recovery state can be
achieved. Once the components that need repairing to achieve a certain
recovery state have been identified, the methodology includes delay

ATC 119-1 Part 5: Tall Building Effects on Post-earthquake Recovery 5-23

 estimates associated with impeding factors, defined as those factors which
may impede the initiation of repairs. Impeding factors include post-
earthquake inspection, engineering mobilization, contractor mobilization,
financing, permitting, and long-lead-time components.

Recent studies on the seismic performance evaluation of tall buildings, which

employ the methods discussed above, provide estimates of downtime to
functional recovery. Analyses of a 1970s design of 40-story tall steel
moment-resisting frame office building, modeled after existing buildings in
San Francisco, estimate that it could experience downtime to functional
recovery on the order of 87 weeks under a design earthquake (Molina Hutt et
al., 2016) or 41 weeks due to ground motions from a magnitude-7.0
earthquake on the Hayward fault (USGS, 2018). A significant portion (about
three-quarters) of the expected downtime is due to impeding factors, as
opposed to the duration of repairs. Similarly, studies of a 42-story modern
reinforced concrete shear wall residential building indicate that it could
experience downtime to functional recovery on the order of 84 weeks under a
design earthquake (Tipler and Deierlein, 2014) or 33 weeks under a
magnitude 7.0 earthquake on the Hayward fault (USGS, 2018). In both of
these studies, the ground motions experienced in San Francisco from the
magnitude-7.0 Hayward earthquake are about half as strong as those of the
design earthquake ground motion in San Francisco due to a larger magnitude
earthquake on the San Andreas fault. Thus, what appeared to be moderate
economic impacts due to direct losses, as presented in Section 4.1, could be
potentially catastrophic to owners and tenants due to extensive downtime and
the associated indirect costs. Part 3 addresses these disproportionate
downtime estimates in tall buildings with possible policy recommendations to
address them. Part 3 also addresses some of the challenges in
implementing existing downtime assessment methods to tall buildings.

4.3 Consequences to City

Both direct and indirect losses, as identified in Sections 4.1 and 4.2
respectively, can lead to long-term impacts including loss of jobs, housing or
human migration due to a re-alignment of economic activity. In addition to
the loss of population and income, other evidence of long-term impact in the
economic fortune of an affected region includes fluctuations in real estate
values. These higher-order impacts can have long-term and far-reaching
consequences, and therefore, they should be considered in post-disaster
analysis and policy formulation (Nov and DuPont, 2016).

Earthquake resilience is the ability of an organization or community to quickly

recover after a future large earthquake and the current building codes do not

5-24 Part 5: Tall Building Effects on Post-earthquake Recovery ATC-119-1

focus on it. The consequences of extended downtime and the inability of
people to return to their homes, jobs, and schools are difficult to quantify:
including the loss of culture, sense of community, and quality of life can
impact communities for years and even decades after an earthquake (Arup,
2013). Because it is not possible to control these higher-order impacts in a
post-earthquake disaster context, policy recommendations should address
both direct and indirect economic losses to ensure business continuity and
liveability of communities after the earthquake. Of particular relevance for tall
buildings are cordoning (Section 3.1 of this Part) and the development of
alternative habitability standards (as later described in Section 5.2 of this
Part) as two critical considerations in the transition between emergency
operations and recovery. The presence of extended cordons has
implications for immediate (impacted) neighbors and the broader community,
and alternative habitability standards are important for safely reoccupying tall
buildings. While individuals should not be exposed to unnecessary risk,
cordoning and reoccupancy posting decisions must carefully consider the
implications on displaced individuals and businesses.

ATC 119-1 Part 5: Tall Building Effects on Post-earthquake Recovery 5-25

Chapter 5

Accelerating Recovery

5.1 Inspection and Tagging

Following damaging earthquakes, post-earthquake safety evaluations of

buildings are typically carried out according to the procedures presented in
ATC-20-1, Field Manual: Postearthquake Safety Evaluation of Buildings,
(ATC, 2005), by which each inspected building is posted with one of three
placards: Inspected (Green), Restricted Use (Yellow) or Unsafe (Red). San
Francisco’s application of this posting procedure for tall buildings is discussed
in Part 6.

In addition to posting a building, it may be necessary to designate areas for

which access is restricted. These hazardous areas may be inside or outside
the building. According to ATC-20 procedures, areas outside the building
and within a potential striking distance from a falling hazard must be
barricaded to prevent entry and if necessary, the area can be formally posted
with an Unsafe placard. The procedures suggest that barricades to establish
a safety cordon, as previously discussed in Section 3.1 of this Part, may be
prompted by post-earthquake safety evaluations, as seen in the third item of
guidelines for barricading illustrated in Figure 5-4. The guidelines explicitly
state that when a building is in danger of collapse, in danger from collapse of
adjacent structures, or has falling hazards present, the building must be
posted as Unsafe and barricades should be set up.

Even though ATC-20 guidance includes barricades within its scope, the
evaluation may not be possible until a few days after the earthquake,
depending on the extent of local damage and the number of available
qualified inspectors and engineers. To prevent potential injuries that may
occur due to collapse or other falling hazard conditions in tall buildings, which
could occur unexpectedly or as a result of earthquake aftershocks, these
evaluations can be expedited through the Building Occupancy Resumption
Program (BORP), also discussed in Part 6.

BORP, which is administered by the City and County of San Francisco’s

Department of Building Inspection, allows San Francisco building owners to
pre-certify post-earthquake inspection of their buildings by qualified private
engineers and specialty contractors. The program requires the engineer to

ATC 119-1 Part 5: Tall Building Effects on Post-earthquake Recovery 5-27

 be familiar with the building’s systems, have access to relevant drawings and
that a copy of the completed emergency inspection plan be stored on site
along with inspection supplies and pertinent construction drawings of the
building’s architectural, structural, and life-safety systems for use in future
post-earthquake inspections. The minimum standard for this BORP
assessment is the ATC-20 Detailed Evaluation. Some owners or engineers
may choose a more complete engineering evaluation. Because the program
spans beyond a structural evaluation and requires sign-off by the elevator
firm and life-safety system maintenance personnel, its implementation could
further expedite building reoccupancy and recovery.
Table 2-4. Guidelines for Barricading
1. Use caution tape or cones only initially. Chain link fences and wood
or metal barricades make better long-term restraints.
2. Do not set barricades too close. Glass and brick walls can shatter on
impact. Initially set the barricades wide. After a Detailed Evaluation
or a period of stability (e.g., several days or a week or more), they
often can be moved closed to the building.
3. When a downtown area (e.g., several city blocks) has extensive dam-
age, recommend temporarily cordoning off the entire area. This
avoids the need to barricade individual buildings and helps prevent
theft.
4. Recommend the posting of guards at structures in imminent danger
of collapse.
5. Recommend scaffolding and planking be installed over sidewalks
and entrances to protect pedestrians when falling hazards are rela-
tively small (e.g., a few bricks).
6. If safe to do so, try to keep one lane open for traffic when barricad-
ing street and sidewalk areas.

Figure 5-4 Guidelines for barricading (from ATC, 1995).

The liability in this type of evaluation is no different from that of regular

building placarding. As outlined in the Limits of Liability section of the BORP
Guidelines for Engineers (CCSF, 2017c), the structural engineer charged with
performing the BORP inspection is acting as a City-authorized building
inspector to assist building owners to obtain timely building inspection in the
event of an emergency post-earthquake situation. As such, the structural
engineer is intended to have the same exemptions from liability as is provided
to the City building inspectors for emergency inspection.

Due to the complex nature of post-earthquake safety evaluationof tall

buildings and the benefits provided by the voluntary BORP program, the City
should consider expanding or mandating the program (or some variant of the

5-28 Part 5: Tall Building Effects on Post-earthquake Recovery ATC-119-1

program) for all tall buildings. Revisions to the program should be considered
as outlined in Part 6, which covers both ATC-20 and BORP more extensively.

As an additional means to facilitate post-earthquake assessment, the City

should evaluate the benefits of strong motion instrumentation of tall buildings
for reoccupancy. With modern instrumentation technology, building response
information is usually available immediately after an earthquake, which can
provide the earthquake engineer with data to help quantify the condition of
the building and, thereby, provide more assurance on the building safety to
city officials along with building owners or tenants. Per Administrative Bulletin
(AB) 058 (CCSF, 2008), adopted in 2008 all new buildings in San Francisco
over 10 stories (~120 ft) in height (or 6 stories with floor area > 60,000 sf)
should have strong motion instrumentation. Minimum instrumentation is 3
triaxial units, which should be located in the basement, mid-height, and near
the top of the building. While BORP allows the use of recorded motions to
guide building inspection, it currently does not include specific
recommendations on how to use such data. Moreover, thought should be
given to technologies for improving timely (real-time) access to and
interpretation of strong motion data, e.g., such as the USGS program to
facilitate evaluation of Veterans Administration hospitals (Kalkan et al., 2012).
Thus, in addition to promoting (or requiring) more instrumentation in
buildings, the City should consider an effort to facilitate consistent and
effective use of the strong motion data. In addition to facilitating post-
earthquake assessment, these data can also serve a longer term benefit of
validating and improving models used to design and evaluate the
performance of tall buildings.

5.2 Reoccupancy and Functional Recovery

Reoccupancy is allowed when the building is deemed safe enough to be

used for shelter. According to the Cal OES Safety Assessment Program
(SAP, see Part 6), reoccupancy can occur once a Green placard is posted
following inspection by a qualified professional on the basis that any damage
to structural and nonstructural components does not pose a threat to life
safety and if egress paths are undamaged (ATC, 2005). A Green placard
allows unrestricted access and reoccupancy to all portions of the building.
Clean-up and/or minor repairs to some nonstructural components (such as
fallen ceiling tiles) may be required so as not to impede egress in some areas
of the building. If life-safety hazards to occupants are evident (which may
include significant structural damage, exterior falling hazards due to damaged
cladding and glazing, interior hazards from damaged components hung from
the floor above or severely damaged partitions), these must be removed or
repaired before a Green placard is awarded.

ATC 119-1 Part 5: Tall Building Effects on Post-earthquake Recovery 5-29

 Since reoccupancy requires only the removal and repair of dangerous
conditions, it can, and often does, occur before functionality is restored.
Functional recovery represents the time required to regain capacity to serve
the facility’s primary function. For all occupancy types, this would require
restoring power, water, fire sprinklers, lighting, and heating, ventilation, and
air-conditioning (HVAC) systems while also ensuring that elevators are back
in service. Back-up systems can help achieve functional recovery until the
municipal utilities are restored and able to provide resources for full capacity.
In residential buildings, functional recovery is related to regaining habitable
conditions, which typically rely on the ready availability of power, water
(potable or not), and heat.

Even so, SPUR (2012) recommended alternative habitability standards for

houses to supersede regular code requirements (California Health and Safety
Code and the San Francisco Housing Code) during a housing-emergency
period declared by the City after a major earthquake. Such an emergency
period might extend for days, weeks, or longer. Relaxing any habitability
standard during some defined recovery period to facilitate reoccupancy and
recovery implies additional risk. Tall buildings rely on sophisticated systems,
such as HVAC, elevators, and fire suppression, for basic habitability and
SPUR-recommended relaxation of habitability standards, developed for
houses and small wood-frame apartment buildings to enable sheltering in
place, are not applicable to tall buildings. Therefore, the City should consider
developing alternative habitability standards for tall buildings after an
earthquake, considering minimum requirements for fire suppression and
safety systems, vertical transportation, water services, and electricity for pre-
determined periods of time following an earthquake. Part 6 Section 2.2
presents furtehr discussion on this topic.

Another critical consideration in enabling building reoccupancy and functional

recovery is monitoring and evaluating cordons, as discussed in detail in
Section 3.1 of this Part, particularly in the event of aftershocks.

5.3 Insurance

Timely repair and recovery following earthquakes requires that tall building
owners (and possibly tenants) have access to financial capital to pay for
inspections, repairs, and other costs. Funds for repair and reconstruction
can come from a variety of sources, including: bank accounts (either general
accounts or reserve funds), insurance, or loans. The availability and access
to funds for tall building repairs can vary significantly depending on the
building ownership and stakeholders. In particular, the financial options are
dramatically different for a single-owner commercial building versus a multi-

5-30 Part 5: Tall Building Effects on Post-earthquake Recovery ATC-119-1

owner residential condominium building. Due to the number of buildings and
diversity of owners, it is difficult to establish whether San Francisco’s tall
building inventory is financially well-prepared to recovery from a large
earthquake. Available information suggests that some commercial buildings
(especially those part of large real estate holdings) are financially well
prepared, whereas others, including residential condominiums may not be.

The California Earthquake Authority (CEA) was established following the

1994 Northridge earthquake to manage and facilitate making earthquake
insurance available from member insurance companies to residential
homeowners, mobile home owners, condo unit owners, and renters. Despite
efforts by the CEA to promote the value and benefits of earthquake
insurance, earthquake insurance penetration in California is low. As of
December 2017, there were just over 1 million earthquake insurance policies
in the state compared to its over 12 million households (less than 10% market
penetration). For tall residential condominium buildings, the earthquake
insurance and repair financing situation is further complicated by shared
responsibilities between homeowner associations (HOAs) and individual unit
owners.

For commercial buildings, a recent NY Times article (Fuller, 2018) and other
sources suggest that insurance penetration rates for commercial buildings is
similarly low, on the order of 10%. This is contrasted by other anecdotal
evidence from commercial tall building owners and insurance companies,
which suggests that most tall building owners in San Francisco have some
earthquake insurance. Based on these discussions, buildings owned by
large developers as part of a larger building portfolio are more likely to be
insured. For buildings with outstanding loans, the lending institution may
require earthquake insurance, but even if required at the time when a loan is
taken, the earthquake insurance may not remain over the life of the loan or
building. Thus, without any formal procedure to report insurance information,
the actual situation is difficult to verify.

Both for residential and commercial tall buildings, available evidence

suggests that even when earthquake insurance is present, the insured
amounts may be limited to a small fraction of the building asset value.
Commercial owners may only insure for the expected loss under the design
earthquake (probable maximum loss), and individual condominium owners
may only insure for damage within their unit.

In summary, the available information raises doubts about the availability of

capital to repair and recover from building damage after a strong to severe
earthquake.

ATC 119-1 Part 5: Tall Building Effects on Post-earthquake Recovery 5-31

 Earthquake insurance for commercial buildings is generally regarded as
specialty insurance. Whether a building is insured purely for earthquake
losses or through a multi-peril policy depends on how the contract with each
insurer is structured. Retrofit or other measures taken to reduce the risk of
earthquake damage can in concept be used to negotiate reduced insurance
premiums, but the degree to which premiums are reduced depends on the
market, the owner/risk manager, the broker and the insurance company. Due
to lack of requirements, a direct and transparent link between retrofit or other
seismic performance measures and insurance premiums does not exist.

Corelogic (2018) estimates that under a magnitude-7.0 earthquake

mainshock on the Hayward fault, residential earthquake insurance would
enable recovering 9% of the losses incurred due to damage to residential
buildings, whereas commercial earthquake insurance would enable
recovering 20% of the losses incurred due to damage to commercial
buildings. “The difference between the estimated property damage and the
amount recovered by insurance is driven largely […] by the lack of insurance
for most properties in California and, to a lesser degree, the effect of
insurance deductibles and limits.” This realistic earthquake scenario
highlights the considerable insurance gap, both for commercial and
residential properties, in the San Francisco Bay Area (RMS, 2017; CEA,
2011).

Another important consideration for building owners and tenants is business

interruption insurance, which covers loss of income suffered by a business
after a disaster, such as loss of rent. In the event of cordons, business
interruption insurance could be triggered by a city-ordained or owner-initiated
building closure, unless a specific clause excludes this specialty insurance.
In the case of cordons caused by a single damaged tall building restricting
access to surrounding buildings, this would be dealt with through general
liability insurance, as opposed to earthquake or business disruption
insurance.

Another situation unique to tall buildings is the disproportionate effect their

damage could have on nearby neighbors and the City. For example,
extended cordons around a damaged tall building would inflict losses on
nearby neighbors and possibly require the City to step in with emergency
measures to stabilize (or potentially demolish) a building. The extent to which
tall building owners are responsible and able to compensate costs borne by
neighbors and the City remains an important question.

Given the uncertainty regarding earthquake insurance coverage or other

means to provide financial resources for repair and recovery, it is

5-32 Part 5: Tall Building Effects on Post-earthquake Recovery ATC-119-1

recommended that the City identify potential limitations to the availability of
capital after a damaging earthquake and, if necessary, require minimum
levels of (self-) insurance for tall building owners to ensure recoverability.
Alternatively, other solutions, such as pooled recovery funds should be
explored.

ATC 119-1 Part 5: Tall Building Effects on Post-earthquake Recovery 5-33

Chapter 6

Issues and Recommendations

A characterization of building tallness in relevant building codes suggests that

while there is no single accepted height threshold is valid to distinguish
between “tall” versus “non-tall” buildings, the fact that different thresholds do
exist for fire safety, seismic, or planning (amongst other considerations), is an
indication that tall buildings present unique challenges. This Part provides an
overview of seismic risks that are disproportionately associated with tall
buildings, referred to as tall building issues, as they relate to post-earthquake
recovery. The following list provides a summary of tall building issues
identified within this Part, recommendations for new policy considerations and
further research to mitigate these effects, and a brief discussion on their
relevance to tall versus all buildings, and future versus existing buildings. For
each recommendation, references to relevant sections are provided.
1. Issue(s). Following a damaging earthquake, safe evacuation and
inspection of tall buildings involve many considerations that are not
addressed by existing post-earthquake inspection and response
measures. In particular, where it becomes necessary to cordon around a
damaged building, the standard safe cordoning distance of 1 to 1.5 times
the building height (or the height of the falling hazard), could be extremely
disruptive to the dense downtown neighborhoods of San Francisco.
Review of inspection requirements and discussions with local authorities
in the City of San Francisco reveal gaps in the understanding issues
associated with inspection of tall buildings and tall building cordons.
Recommendation(s). Define clear roles and responsibilities associated
with post-earthquake inspection and evaluations of tall buildings as
related to establishing cordon areas around damaged tall buildings and
coordinating with the San Francisco priority routes.
Study the benefits of establishing San Francisco’s own Urban Search and
Rescue team or develop plans to leverage regional, statewide, and
national teams (e.g., Oakland, Menlo Park) for evacuating residents of
damaged tall buildings and/or surrounding buildings that are at risk from
falling or collapse hazards. Ensure that training exercises for teams
anticipated to respond in San Francisco address the unique challenges
associated with tall buildings.

ATC 119-1 Part 5: Tall Building Effects on Post-earthquake Recovery 5-35

 Identify and anticipate liability issues related to cordons, including the
following: (1) the entity that declares cordon, i.e., the City or others;
(2) the owner of the cordoned building; and (3) neighbors of the cordoned
building that are impacted by the cordon.
Utilize inspections and reports resulting from San Francisco’s façade
ordinance to facilitate a detailed study of risks of falling debris.
Relevance. The disproportionate effect on the urban community
associated with adopting existing cordoning practices is a tall building
effect. While this applies to future and existing buildings, the greatest
concerns are with existing buildings, due to their greater risk of structural
and cladding damage.
2. Issue(s). The complexity associated with barricading, emergency
stabilization, and falling hazard removal activities for tall buildings may
require supplementing City staff and expertise to speed emergency
response and recovery after a damaging earthquake. Related to this are
questions as to whether USAR and other regional and national search
and rescue teams, which may be called in to assist the City, have
sufficient guidelines and training for conditions that may be encountered
with heavily damaged tall buildings.
Recommendation(s). Establish a Tall Building Task Force to: (1) help
effectively conduct rapid assessments of tall buildings; (2) manage
detailed investigations of damaged tall buildings, their potential fall zones,
and how they might be prevented from further collapse; and (3) help
refine and confirm safe distances from barricades and cordons.
Relevance. These complexities are relevant to tall buildings, but also
applicable to critical facilities. They apply to future and existing buildings,
but are of greater concern for existing buildings.
3. Issue(s). High-rises have a heavy reliance on the building’s automatic
fire-suppression systems to control the fire. The current building and fire
codes include a requirement for 30 minutes of local water supply for
automatic fire suppression systems, in the event that City water supply to
the building is disrupted. The 30 minutes is generally based on the
expected time for the fire department to arrive to combat the fire. This 30-
minute requirement has been in place for several decades and may not
be adequate in the event of a large earthquake where arrival times for the
fire department are impacted by: (1) competing demands for emergency
response; (2) disruptions to transportation routes; and (3) building
damage to fire service elevators. A related concern is whether the 30-
minutes is sufficient to allow for safe evacuation of the tall building (or

5-36 Part 5: Tall Building Effects on Post-earthquake Recovery ATC-119-1

 affected floors of the building) that has experienced stair or evacuation
elevator damage during the earthquake.
Recommendation(s). The City should engage the fire department along
with other agencies and experts to evaluate whether the local water
storage requirement (30-minute supply) is sufficient following expected
large earthquake in San Francisco. The study should also consider
current practices and policies for evacuation of tall buildings, including the
use of elevators for emergency evacuation. Based on this study, develop
recommendations for changes to local water supplies or other measures
to reduce the risk of post-earthquake fire spread and safe evacuation of
tall buildings.
Relevance: The reliance on the building’s fire suppression system and
challenges in evacuation are unique to tall buildings. Reliance on water
supply applies to both future and existing buildings. The use of elevators
for evacuation applies to future buildings.
4. Issue(s). Following a damaging earthquake, even if a building is safe
from structural failure and nonstructural hazards, occupancy is not
normally allowed unless habitability standards (unrelated to seismic
safety) are also met. Tall buildings rely on sophisticated systems, such
as HVAC, elevators, or fire suppression, for basic habitability. Relaxation
of habitability standards, such as those proposed by SPUR (2012)
developed for houses and small wood-frame apartment buildings to
enable sheltering in place, are not applicable to tall buildings.
Recommendation(s). Develop alternative habitability standards for tall
buildings after an earthquake, considering minimum requirements for fire
barriers, suppression and safety systems, vertical transportation, water
services, and electricity.
Relevance: The need for alternative habitability standards applies to all
buildings. However, the emphasis here is that these need to be different
for tall buildings. These standards are applicable to both future and
existing buildings, but should focus primarily on those with residential use
to enable sheltering in place.
5. Issue(s). Tall buildings pose challenges to quickly inspect following a
damaging earthquake. While programs such as Building Occupancy
Resumption Program (BORP) exist to expedite post-earthquake
inspection, these are voluntary. Similarly, while strong motion
instrumentation of tall buildings is required, these are not well utilized for
real-time post-earthquake evaluations of building safety.

ATC 119-1 Part 5: Tall Building Effects on Post-earthquake Recovery 5-37

 Recommendation(s). Consider expanding or mandating BORP for all
tall buildings. A detailed list of recommendations related to BORP can be
found in Part 5.
Evaluate ways to expand and improve the use of strong motion
instrumentation of tall buildings to facilitate rapid building evaluation for
informing decisions about building tagging, re-occupancy, and recovery
planning.
Relevance. The complexities of inspection are unique to tall buildings,
but generally applicable to large buildings. This applies to future and
existing buildings, but is of greater concern for existing buildings.
6. Issue(s). Insurance penetration for commercial and residential buildings
is low, and the insured amounts are often limited to a small fraction of the
building value. These raise questions about the availability of capital for
repair and recovery operations after a damaging earthquake.
Recommendation(s). Identify potential limitations to the availability of
capital after a damaging earthquake and explore ways to facilitate and
encourage minimum levels of (self-)insurance for tall building owners to
ensure recoverability.
Relevance. The large amount of capital required to recover from a
damaging earthquake is generally applicable to all buildings but is of
particular concern for tall buildings given: (1) the large number of building
occupants; and (2) potential financial liability of tall buildings on their
nearby neighbors and city services. This challenge applies to both future
and existing buildings, but it is of greater concern for existing buildings
that are at greater risk of damage.

5-38 Part 5: Tall Building Effects on Post-earthquake Recovery ATC-119-1

Appendix A

References

Arup, 2013, Resilience-based Earthquake Design Initiative (REDi) for the

Next Generation of Buildings, Arup, San Francisco, California.
ATC, 2005, Field Manual: Postearthquake Safety Evaluation of Buildings,
Second Edition, ATC-20-1, Applied Technology Council, Redwood
City, California.
ATC, 1995, Addendum to the ATC-20 Procedures for Postearthquake Safety
Evaluation of Buildings ATC-20-2, Applied Technology Council,
Redwood City, California.
ASCE, 2017, Minimum Design Loads and Associated Criteria for Buildings
and Other Structures, ASCE/SEI 7-16, American Society of Civil
Engineers Structural Engineering Institute, Reston, Virginia.
ASME, 2014, Safety Code for Elevators and Escalators, ASME A17.1,
American Society of Mechanical Engineers, New York, New York.
Bonowitz, D., 2011, Resilience Criteria for Seismic Evaluation of Existing
Buildings, 2008 Special Projects Initiative Report to Structural
Engineers Association of Northern California, San Francisco,
California.
Bukowski, R.W., 2009, Emergency Egress from Buildings, NIST Technical
Note 1623, National Institute of Standards and Technology,
Gaithersburg, Maryland.
CALBO, 2013, Interim Guidance for Barricading, Cordoning, Emergency
Evaluation and Stabilization of Building with Substantial Damage in
Disasters, Structural Safety Committee, California Building Officials,
Sacramento, California.
CBSC, 2016a, 2016 California Building Code, California Code of Regulations,
Title 24, California Building Standards Commission, Sacramento,
California.
CBSC, 2016b, 2016 California Existing Building Code, California Code of
Regulations, Title 24, Part 10, California Building Standards
Commission, Sacramento, California.

ATC 119-1 Part 5: Tall Building Effects on Post-earthquake Recovery 5-39

 CCSF, 2008, Procedures for Seismic Instrumentation of New Buildings,
Administrative Bulletin 058, Department of Building Inspection, the
City and County of San Francisco, California.
CCSF, 2016a, The San Francisco Building Code, the City and County of San
Francisco, California.
CCSF, 2016b, The San Francisco Existing Building Code, the City and
County of San Francisco, California.
CCSF, 2017a, San Francisco Property Information Map, Planning
Department, the City and County of San Francisco, California.
Available at: http://propertymap.sfplanning.org/, last accessed: August
2018.
CCSF, 2017b, SF Zoning Map - Sheet HT01, the City and County of San
Francisco, California. Available at: http://sf-planning.org/zoning-maps,
last accessed September 4, 2018.
CCSF, 2017c, San Francisco Building Occupancy Resumption Program –
Guidelines for Engineers, Department of Building Inspection, the City
and County of San Francisco, California. Available at: https://sfdbi.org
/borp-guidelines-engineers, last accessed: September 4, 2018.
CEA, 2011, Affordable & Available Earthquake Insurance: Underscoring the
Need for Risk‐Transfer Diversification, California Earthquake
Authority, Sacramento, California. Available at:
https://www.earthquakeauthority.com/Press-Room/Media-Resources
/Public-Affairs-and-Testimony/OECD-Affordable-Available-Earthquake
-Insurance, last accessed October 30, 2018.
Cohen, J.M., 1995, “Seismic performance of cladding: Responsibility
revisited,” Journal of Performance of Constructed Facilities, Vol. 9,
No. 4.
Comerio, M., 2006, “Estimating downtime in loss modeling,” Earthquake
Spectra, Vol. 22, No. 2, pp. 349-365.
Corelogic, 2018, Financial Implications of the HayWired Scenario. Available
at: https://www.corelogic.com/downloadable-docs/corelogic_haywired-
scenario_r5.pdf, last accessed October 30, 2018].
Eads, L., 2013, Seismic Collapse Risk Assessment of Buildings: Effects of
Intensity Measure Selection and Computational Approach, Technical
Report 184, John A. Blume Earthquake Engineering Center, Stanford
University, California.

5-40 Part 5: Tall Building Effects on Post-earthquake Recovery ATC-119-1

FEMA, 2009, National Response System Shoring Basics Module 2a, Shoring
Construction Module, Federal Emergency Management Agency
Washington, D.C.
FEMA, 2012a, Seismic Performance Assessment of Buildings, FEMA P-58,
prepared by the Applied Technology Council for the Federal
Emergency Management Agency, Washington, D.C.
FEMA, 2012b, Reducing the Risks of Nonstructural Earthquake Damage: A
Practical Guide, FEMA E-74, prepared by the Applied Technology
Council for Federal Emergency Management Agency, Washington,
D.C.
Filiatrault A., Christopoulos C., and Stearns C., 2002, Guidelines,
Specifications, and Seismic Performance Characterization of
Nonstructural Building Components and Equipment, PEER Report
No. 2002/05, Pacific Earthquake Engineering Research Center,
University of California, Berkeley, California.
Fuller, T., 2018, “In quake-prone California, alarm at scant insurance
coverage,” The New York Times. Available at: https://www.nytimes
.com/2018/08/31/us/california-earthquake-insurance.html, last
accessed September 4, 2018.
Goodno, B.J., Craig, J.I., and Zeevaert Wolff, A., 1989, “Behavior of heavy
cladding components in the Mexico earthquake,” Earthquake Spectra,
Vol. 5, No. 1, pp. 195- 222.
Haselton C., and Deierlein G.G., 2007, Assessing Seismic Collapse Safety of
Modern Reinforced Concrete Frame Buildings, PEER Report 2007/08,
Pacific Earthquake Engineering Research Center, University of
California, Berkeley, California.
ICBO, 1964, 1985, Uniform Building Code, International Conference of
Building Officials, Whittier, California.
ICC, 2016, International Building Code, International Code Council, Country
Club Hills, Illinois.
Kalkan, E., Banga, K., Ulusoy, H.S., Fletcher, J.P.B., Leith, W.S., and Blair,
J.L., 2012, Helping Safeguard Veterans Affairs’ Hospital Buildings by
Advanced Earthquake Monitoring, U.S. Geological Survey Fact Sheet
2012–3094. Available at: https://pubs.usgs.gov/fs/2012/3094/.
Kinateder, M.T., Omori H., and Kuligowski E.D., 2014, The Use of Elevators
for Evacuation in Fire Emergencies in International Buildings, NIST

ATC 119-1 Part 5: Tall Building Effects on Post-earthquake Recovery 5-41

 Technical Note 1825, National Institute of Standards and Technology,
Gaithersburg, Maryland.
Liel A., and Deierlein G.G., 2013, “Cost-benefit evaluation of seismic
mitigation alternatives for older reinforced concrete frame buildings,”
Earthquake Spectra, Vol. 29, No. 4, pp. 1391-1411.
Molina Hutt C., Almufti I., Willford M., and Deierlein G.G., 2016, “Seismic loss
and downtime assessment of existing tall steel-framed buildings and
strategies for increased resilience,” Journal of Structural Engineering,
Vol. 142, No. 8.
Molina Hutt C., Rossetto T., and Deierlein G.G., 2018, “Comparative risk-
based seismic performance assessment of 1970s vs modern tall steel
moment resisting frames,” manuscript submitted for publication.
NIOSH, 2009, Preventing Deaths and Injuries of Fire Fighters when Fighting
Fires in Unoccupied Structures, The National Institute of Occupational
Safety and Health, Cincinnati, Ohio.
Noy, I., and duPont, W., 2016, “The long-term consequences of natural
disasters,” Working Papers from the School of Economics and
Finance, Victoria University of Wellington, New Zealand.
NZPA, 2011, Christchurch Earthquake: Cordon around Grand Chancellor
Narrows, New Zealand Press Association. Available at:
www.stuff.co.nz/national/christchurch-earthquake/4716748
/Christchurch-earthquake-Cordon-around-Grand-Chancellor-narrows,
last accessed: September 2017.
OCHA, 2012, INSARAG Preparedness-Response, International Search and
Rescue Advisory Group Guidelines and Methodology, United Nations
Office for the Coordination of Humanitarian Affairs, New York, New
York.
PEER, 2017, Guidelines for Performance-based Seismic Design of Tall
Buildings, Version 2.03, Report No. 2017/06, Pacific Earthquake
Engineering Research Center, University of California, Berkeley,
California.
RMS, 2017, 25 Years After Loma Prieta, San Francisco Bay Area Faces
Increased Earthquake Risk; Next “Big One” Could Cause $200 Billion
in Losses, Risk Management Solutions, Newark, California. Available
at: https://www.rms.com/newsroom/press-releases/press-detail/2014-
10-13/25-years-after-loma-prieta-san-francisco-bay-area-faces
-increased-earthquake-risk-next-big-one-could-cause-200-billion-in
-losses, last accessed October 30, 2018.

5-42 Part 5: Tall Building Effects on Post-earthquake Recovery ATC-119-1

Schweier C., and Markus M., 2004, “Assessment of the search and rescue
demand for individual buildings.” Proceedings, 13th World Conference
on Earthquake Engineering, Vancouver, Canada.
SPUR, 2012, Safe Enough to Stay, San Francisco, California.
Steinbrugge, K.V., and Schader, E.E., 1973, Earthquake Damage And
Related Statistics in San Fernando, California, Earthquake of
February 9, 1971, edited by Leonard Murphy. Volume 1A, pp.
709-710 and 713, National Oceanic and Atmospheric Administration,
Washington D.C.
Tipler, J., and Deierlein G.G., 2014, Seismic Resilience of Tall Buildings –
Benchmarking Performance and Quantifying Improvements,
Department of Civil and Environmental Engineering, Stanford,
California.
USGS, 2018, The HayWired Earthquake Scenario - Engineering Implications,
USGS Scientific Investigations Report 2017-5013-I-Q, US Geological
Survey, Menlo Park, California.

ATC 119-1 Part 5: Tall Building Effects on Post-earthquake Recovery 5-43

 PART 6:
Post-earthquake Occupancy
and Safety Inspection
Chapter 1

Introduction

1.1 Background

This Part reviews and explains current San Francisco policy and practices for
post-earthquake safety inspection of buildings, especially as they relate to the
City’s existing tall buildings. The recommendations presented correspond to
Recommendations 3A, 3B, 3C, 3F, and 4B presented in Summary
Recommendations.
This Part addresses aspects of the following Tasks described in the CAPSS
Earthquake Safety Implementation Program Workplan 2012-2042 (CCSF,
2011):
• Task A.4.a: Develop and adopt shelter-in-place policies and procedures
• Task A.4.d: Adopt disproportionate damage trigger
• Task A.4.f: Update post-earthquake inspection policies and procedures
• Task B.1.b: Develop non-structural upgrade program for businesses
• Task B.2.b: Mandatory evaluation of 5+ dwelling unit residential buildings
and hotels/motels
• Task B.4.b: Develop post-earthquake repair and retrofit standards
• Tack C.2.b: Mandatory evaluation and retrofit of critical stores, suppliers,
and service providers
• Task C.2.c: Mandatory evaluation and retrofit of larger assembly
buildings

1.2 Organization

Chapter 2 presents the current San Francisco practice, as well as discussion

and recommendations related to safety, re-occupancy, functional recovery,
and resilience.

Chapter 3 presents discussion and recommendations related to the Safety

Assesment Program (SAP) administered by the California Office of
Emergency Services.

ATC-119-1 Part 6: Post-earthquake Occupancy and Safety Inspection 6-1

 Chapter 4 presents discussion and recommendations related to San
Francisco’s Building Occupancy Resumption Program (BORP).

Chapter 5 presents discussion and recommendations related to post-

eartqhuake safety evaluation of buildings.

Chapter 6 presents discussion and recommendations related to available

guidance on evaluation and repair of welded steel moment-resisting frame
buildings.

Chapter 7 presents a brief discussion on available additional techniques.

A list of references is provided at the end of the Part.

6-2 Part 6: Post-earthquake Occupancy and Safety Inspection ATC-119-1

Chapter 2

Background

2.1 Current San Francisco Practice

Reoccupancy is the ability to re-enter a building after an earthquake and

either resume normal operations or start on clean-up and repairs. For any
building, safe reoccupancy is the first milestone on the path to functional
recovery and eventually full recovery (Bonowitz, 2011). Since the Loma
Prieta earthquake, safe reoccupancy in San Francisco—for a building of any
height—has been assured by a process of safety inspections.

As part of this process, the City relies on the state-run Safety Assessment
Program (SAP), which implements the evaluation procedures presented in
ATC-20-1, Field Manual: Procedures for Postearthquake Safety Evaluation
(CCSF, 2008, Section 6.5.1; ATC, 2005). For a tall building, this process can
be complicated and delayed by the sheer size of the structure, to say nothing
of its architectural complexity and variety of tenant spaces. To help address
these challenges, San Francisco developed the Building Occupancy
Resumption Program (BORP) as a way of outsourcing the inspection process
to building owners themselves (CCSF, 2011). BORP also uses ATC-20
procedures, but for certain steel frame structures, it allows an alternative
guideline developed after the 1994 Northridge earthquake, known as FEMA
352 (FEMA, 2000).

This Part discusses each of these four components of San Francisco’s

current practice—SAP, BORP, ATC 20, and FEMA 352—and recommends
possible improvements.

2.2 Safety, Reoccupancy, Functional Recovery, and City

Resilience

SAP and BORP are both focused on safety, in the service of reoccupancy
and functional recovery. These three measures of building performance are
related, as each is a function of building damage. In general, the lower the
damage, the higher the safety, and the lower the damage, the faster the
reoccupancy and functional recovery (Bonowitz, 2016). Yet they are not
identical.

ATC-119-1 Part 6: Post-earthquake Reoccupancy and Safety Inspection 6-3

 For example, a building can perform safely but still not be ready to reoccupy
because of hazards posed by adjacent buildings or infrastructure, or because
of restrictions set by local authorities (perhaps for public safety or debris
removal). This was illustrated by the years-long closure of the entire central
business district in Christchurch, New Zealand after a 2011 earthquake.
Thus, while the poor performance of a tall building is likely to affect its
neighbors, it is also possible that damage to San Francisco’s downtown
neighborhoods might delay the recovery of an otherwise safe tall building.

From a tenant’s perspective, it is also possible to achieve a measure of

organizational recovery and resilience even without a safe building. Many
organizations use continuity-of-operations planning to limit downtime through
remote data storage, online services, and even alternate or backup facilities.
These are often intended as temporary measures, but in cases of severe
building damage, permanent relocation is sometimes unavoidable; if
anticipated and planned, it can also be an effective recovery strategy.

Setting aside relocation and continuity-of-operations, reoccupancy is also

different from functional recovery (Bonowitz, 2011). Both require a safe
structure, but functional recovery also requires working utilities and building
services, as well as clean-up or repair of contents. In a tall building, this will
almost always include elevators, mechanical and plumbing systems, life
safety systems, electricity and telecommunications, and building security.
SPUR (2012) recommends that in houses, and even in typical San Francisco
apartment buildings and small commercial buildings, in order to speed
reoccupancy and neighborhood recovery, the City should waive temporarily
certain habitability standards that rely on these nonstructural systems. In tall
buildings, however, these systems are more sophisticated and more essential
than in small, older buildings. For example, few San Francisco houses and
apartments have air conditioning, and for much of the year it is even possible
to live temporarily without heat or electricity. But in a tall building (especially
one without operable windows) the HVAC system is essential to any level of
functionality. Section 5.2 of Part 5 presents additional discussion of post-
earthquake habitability standards.

Typical post-earthquake safety evaluations using ATC-20 procedures must

be quick and focused; they therefore do not look beyond safety. In
Christchurch, however, officials eventually found it beneficial to modify their
post-earthquake inspection forms to consider what they called “3S”: safety,
sanitation, and security (ATC, 2014). In San Francisco, habitability issues
could be worth considering within the context of a more detailed or pre-
planned inspection such as that contemplated by BORP. For a tall building,
these issues might need to be considered for individual tenant spaces.

6-4 Part 6: Post-earthquake Reoccupancy and Safety Inspections ATC-119-1

Functional recovery can also be affected by a building owner’s rational
choices. For example, an owner may choose to delay recovery in order to
complete repairs with the building vacant. From a tenant’s perspective, this
indirect, externally-imposed delay has the same effect as a recovery delay
stemming directly from damage. To the extent that San Francisco tall
building tenants are more likely to be subject to such lease provisions, this
might be considered a tall building issue. Similarly, while an owner or tenant
might be willing to resume pre-earthquake functions in a space with repairs
still incomplete, the same space might not be ready to serve as a shop,
restaurant, bank, public parking, or public open space due to liability or
security issues. This concern is not limited to tall buildings. Indeed, in most
of San Francisco’s tall buildings, public accommodations such as these are
typically limited to the ground floor or basement levels.

Earthquake resilience is not an attribute of individual buildings, tall or

otherwise, but an attribute of community functions and, in the aggregate, of
the community as a whole (ATC, 2018). Because buildings, individually and
in groups, provide shelter and vital resources for these resilience-critical
functions, their timely reoccupancy and functional recovery makes an
essential contribution to the City’s resilience. To the extent that the City’s tall
buildings account disproportionately for its housing, or jobs, or revenue, then
the tall buildings deserve special attention within the broader resilience policy.
That said, while the City can and should set reoccupancy and recovery goals
for buildings, San Francisco’s earthquake resilience is ultimately measured
by the return of functions and services, not buildings.
This study addresses the effect of tall building damage on the tall buildings
themselves and, to a lesser extent, on the downtown neighborhoods where
tall buildings are densely clustered and on the City overall. But it does not
explicitly address the likely interactions between the tall buildings, the non-tall
buildings that still comprise most of the Financial District, and the critical
infrastructure that serves the neighborhood. Nor does it explicitly consider
resource demands and capacities of the Financial District and adjacent
neighborhoods’ businesses, residents, workers, and other stakeholders. A
separate recovery plan, drawing on the present study’s findings, would bring
these ideas together in a practical way to support a neighborhood and its
functions, as opposed to just individual buildings with certain characteristics.

Recommendation: Develop a comprehensive recovery plan for the

Financial District and adjacent neighborhoods. While the recovery plan is
being developed, the City Administrator should develop an interim recovery
plan that outlines at least the basic issues. By necessity, the interim plan will

ATC-119-1 Part 6: Post-earthquake Reoccupancy and Safety Inspection 6-5

 need to make many assumptions. The study recommended here would
confirm or correct those assumptions and fill in the critical knowledge gaps.
• Supplement the newly developed inventory of tall buildings with additional
inventory and analysis of specific building uses by industry or
employment sector, in coordination with development of recovery goals
(ATC, 2018) and with the recommendations presented in Part 1.
• Consider the combined effects of tall buildings, non-tall buildings, and
infrastructure, including liquefaction effects.
• Develop the recovery curve for one or more scenario earthquakes. The
recovery curve shows the level of immediate functional loss and the
extent of reoccupancy and recovery over time. Such a study would
provide a valuable basis for neighborhood recovery plans and for new
programs aiming to improve the reoccupancy and functional recovery
times of individual tall buildings.

2.3 Related Issues

The focus of this report is post-earthquake reoccupancy, specifically the

safety inspection protocols used by SAP and BORP to assess whether
shaking damage has rendered a tall building unsafe to reoccupy.

To be sure, earthquake effects other than shaking damage to an individual

tall building (or its neighbors) can affect that building’s reoccupancy.
Although outside the scope of this report, many of these are addressed by
San Francisco’s Emergency Response Plan (CCSF, 2008, Section 6.5) as
follows:
• Emergency response (CCSF, 2008, Section 6.1)
• Law enforcement (CCSF, 2008,Section 6.3)
• Large-scale evacuation (CCSF, 2008, Section 6.3.1)
• Perimeter control (CCSF, 2008, Section 6.3.2; see also cordoning
discussion in Part 5)
• Traffic control and route recovery (CCSF, 2008, Section 6.4)
• Emergency shoring of adjacent properties (CCSF, 2008, Section 6.5.2)
• Debris removal (CCSF, 2008, Section 6.5.3)
• Fire or tsunami damage

Part 5 discusses some administrative and logistical issues related to post-

earthquake reoccupancy, including owner-tenant relationships, the availability
of financing and insurance, and the feasibility of expanding BORP.

6-6 Part 6: Post-earthquake Reoccupancy and Safety Inspections ATC-119-1

Part 4 discusses the application of building code provisions to damaged
buildings. The safety inspections that inform reoccupancy decisions are
focused on the immediate safety of potential building occupants. Separate
from that concern, the San Francisco Existing Building Code requires
engineering evaluations to determine whether seismic retrofit will be required
in addition to damage repair. These evaluations are outside the scope of the
state’s Safety Assessment Program.

ATC-119-1 Part 6: Post-earthquake Reoccupancy and Safety Inspection 6-7

Chapter 3

The Cal OES Safety Assessment

Program

3.1 SAP as Applied in San Francisco

The Safety Assessment Program (SAP), administered by the California Office

of Emergency Services (Cal OES), was created to supplement local
government resources with mutual aid and volunteers from elsewhere in the
state. All SAP volunteers must be registered Disaster Service Workers
subject to California Code of Regulations Title 19. The program has been in
continuous development since its creation in the late 1970s, with significant
procedural improvements as recent as 2007 (Cal OES, 2015).

SAP responds to three accepted facts of post-earthquake recovery:

• Quick, efficient reoccupancy is essential to citywide recovery. Safety is
the focus, but speed is fundamental to the effort. As SAP training
materials note, “The primary goal of the Safety Assessment Program is to
get as many people as possible back into their buildings as quickly and
safely as possible” (SEAONC, 2013). Using a similar program, volunteer
inspectors in Christchurch inspected 72,000 buildings in 10 days and
130,000 in three weeks (ATC, 2014).
• City staff and resources should be assigned first to inspect and assess
essential infrastructure and facilities, especially those needed for public
safety and to support emergency response (Cal OES, 2015).
• In a large, damaging earthquake, practically every affected jurisdiction will
therefore need outside help to meet its inspection goals.

San Francisco’s priorities are consistent with SAP’s intent. The San
Francisco Emergency Response Plan (CCSF, 2008) states in Section 6.5.1:

Assessments will be necessary immediately to determine

whether public and private buildings can be safely entered or
used and to develop priorities for implementing repairs. The
primary goals of this process are to protect the public and to
expedite resumption of residence, business, services, and
community activity as soon as possible. ... The number of

ATC-119-1 Part 6: Post-earthquake Reoccupancy and Safety Inspection 6-9

 facilities and structures requiring immediate safety inspection is
far greater than the number of trained inspectors. Supplemental
resources will be required immediately.

The plan therefore calls for designated City representatives to “immediately

request support from ... volunteer building Evaluators and Coordinators under
the post-disaster SAP.” Table 6-1 of the plan’s Earthquake Annex states an
objective of completing all necessary building safety inspections and
achieving building reoccupancy within a week of a hypothetical earthquake
(of unspecified intensity).

For technical criteria and on-site procedures, SAP relies on ATC-20

procedures (ATC, 2005), which offer a Rapid evaluation method suited to the
objectives of SAP. ATC-20, discussed further in Chapter 4 of this Part,
results in a public “posting” with a colored placard stating the building’s
eligibility for reoccupancy. The San Francisco Building Code (Section
101A.21) endorses ATC-20, customizes the placard text, and gives the
building official discretion over other changes to placard design.

3.2 Recommendations Regarding SAP

Most of the following can apply to non-tall buildings as well as tall ones.

3.2.1 Activation

San Francisco activates SAP on the basis of instructions given in the

Earthquake Annex to the City’s Emergency Response Plan (CCSF, 2008).
However, the instructions are incomplete, somewhat ambiguous, and in some
respects, out of date. For example, current Earthquake Annex Section 6.5.1
charges the Department of Building Inspection (DBI) with implementing SAP
for private buildings, but activation responsibility is given to the EOC Logistics
Section, and Department of Public Works (DPW) is the lead agency for
implementing that Earthquake Annex section overall.

Recommendation: Clarify and update roles and responsibilities associated

with post-earthquake emergency response and safety inspection.
• The Department of Emergency Management (DEM) should clarify
Earthquake Annex Section 6.5.1 regarding who acts on behalf of the City
as the contact with Cal OES, making the request to activate SAP. The
Annex should be coordinated with Cal OES requirements (Cal OES,
2015).
• DEM should undertake a general update of the Earthquake Annex to
account for new technologies and recent construction (including major
public projects), as well as substantial mitigation over the last decade.

6-10 Part 6: Post-earthquake Reoccupancy and Safety Inspection ATC-119-1

 How DEM prioritizes inspection will affect how DBI allocates its own staff
and SAP volunteers.

3.2.2 Early Recovery for Critical Building Groups

The primary focus of SAP is safe reoccupancy, but SAP has a secondary
goal regarding functional recovery, especially with respect to infrastructure
(SEAONC, 2013, p.131): “Rapidly clearing for use vital services and
infrastructure that will impact the public at large.” To the extent that San
Francisco’s tall buildings (or any other particular building cohort) represent a
recovery-critical function or service, their recovery should also be prioritized,
or at least emphasized together with basic reoccupancy. Christchurch
officials recognized the benefit of classifying and inspecting buildings based
in part on their use and occupancy and in part on height, and they used the
post-earthquake safety inspections also to consider habitability and security
(ATC, 2014). BORP is a promising vehicle by which to introduce these ideas,
but to be effective for an entire functional sector, they will need to be
implemented more aggressively than through that voluntary program.

Recommendation: Plan to assign SAP volunteers to teams dedicated to

recovery-critical facilities, including pre-selected tall buildings.
• DBI should plan to use specially-qualified SAP volunteers to inspect pre-
selected groups of buildings.
• The building groups of interest should be related to the City’s tentative
recovery goals (ATC, 2018) but might also be associated with certain
building characteristics (such as structural system or material) or
particular occupancies (grocery stores, for example) that might have
especially vulnerable mechanical systems or contents. In concept, tall
buildings not covered by BORP could be one such group.

3.2.3 Indicator Buildings for Aftershock Tracking

The need to re-inspect buildings after significant aftershocks can challenge

SAP implementation in terms of resource allocation and documentation.
Officials in Christchurch identified a few buildings of specific structural types
and used them as indicators of the need to re-inspect the larger population
(ATC, 2014). Indicator buildings would be especially valuable for San
Francisco’s most common vulnerable building types (such as house-over-
garage or Victorian “soft story” conditions) and perhaps for tall or large
buildings where re-inspection would be especially demanding of scarce
resources. Identifying indicator buildings, however, would require tracking
early SAP inspections by structure type, and adding a documentation and
database task for DBI.

ATC-119-1 Part 6: Post-earthquake Reoccupancy and Safety Inspection 6-11

 Recommendation: DBI should plan to track common building types,
perhaps including tall buildings with specific characteristics, in order to
identify indicator buildings that will help manage re-inspection demands after
significant aftershocks.

3.2.4 Consistency

Experience from the Northridge, Humboldt County, and Napa earthquakes

shows inconsistent decision-making by SAP volunteers, leading to
incomplete or unaccountable implementation. For example, different
volunteer inspectors can often assess unfamiliar damage patterns in different
ways, posting similar buildings differently.

Recommendation: Develop clear procedures and policies, and document

them in San Francisco-specific training and guidance materials.
• DBI, as the agency responsible for implementing SAP for private
buildings, should identify areas of potential inconsistency and establish
procedures in coordination with DBI procedures and resources.
• DBI should encourage San Francisco-specific training of SAP volunteers
addressing issues likely to arise from San Francisco’s unique building
stock, some of which are related to the City’s tall buildings.

6-12 Part 6: Post-earthquake Reoccupancy and Safety Inspection ATC-119-1

Chapter 4

Building Occupancy Resumption

Program

4.1 Background

After the 1989 Loma Prieta earthquake, the San Francisco Department of
Building Inspection (DBI) developed the Building Occupancy Resumption
Program (BORP) to offer an optional “precertified emergency inspection
program” that helps address this challenge (CCSF, 2011).

BORP allows an owner—of any building—to apply for pre-approval by DBI to

use expert inspectors directly contracted with the owner to evaluate and post
the building using ATC-20 procedures. Through the pre-approval process,
the owner’s experts become familiar with the building’s structural and
mechanical systems, compile documents, and develop plans for post-
earthquake inspection. The experts, including structural engineers, are paid
on retainer and the building owner applies for recertification every two years.
Advance planning with BORP allows a more informed and detailed inspection
in the same time as a typical SAP inspection, while freeing SAP volunteers to
inspect other buildings. As an incentive, DBI charges the owner no fee to
participate in BORP.

The primary purpose of BORP is the same as that of SAP, to expedite

reoccupancy with safety inspections by experts other than City staff:

This private emergency inspection could facilitate rapid

decisions regarding the closure or reoccupancy of building
areas. Prearranged emergency inspection could reduce
inspection delays, as City inspection personnel typically are
dispatched first to areas of greatest damage or public hazard,
which may not include the building in question. (CCSF, 2011)

But BORP has a second goal beyond basic reoccupancy: to facilitate

functional recovery, especially of commercial buildings.

[SAP’s priorities are] geared toward public safety rather than

expeditious business resumption. Some building owners may
wish to develop programs of private inspection for their buildings

ATC-119-1 Part 6: Post-earthquake Reoccupancy and Safety Evaluations 6-13

 to permit rapid, individualized emergency response. (CCSF,
2011)

The recognition of business resumption as a benefit of BORP is consistent

with the City’s tentative recovery goals (ATC, 2018). In fact, the Department
of Public Works (DPW) uses a BORP-like process to facilitate inspection,
reoccupancy, and recovery of the City’s own facilities. The implication that
BORP can facilitate functional recovery suggests opportunities and rationales
for extending the program. Generally, however, BORP, like SAP, is focused
on safe reoccupancy.

4.2 Issues and Recommendations Regarding BORP

While BORP is currently used primarily by owners of large office buildings, it

has benefits for all buildings where SAP procedures would be slow or
inconclusive. Most of the following discussion apply to non-tall buildings as
well as tall ones.

4.2.1 Training Simulation

In the decades since it was developed after the Loma Prieta earthquake,
BORP has never been implemented after a significant damaging event. Each
BORP plan is updated every two years, but the simultaneous execution of
dozens of BORP plans has not been tested.

Recommendation: Test and improve current BORP procedures through

training simulations with the following considerations:
• DBI should organize and implement regular training simulations with two
purposes: (1) to identify building-specific logistical issues for the
engineers; and (2) to test DBI procedures for receiving, approving, and
tracking input from BORP engineers.
• The training should be simultaneous for at least a majority of enrolled
buildings.
• Each enrolled building should be required to participate in a training
simulation, perhaps once every four years, as a condition of continuing
BORP approval.

4.2.2 Program Extension

For a tall building, an ATC-20 Rapid evaluation performed by a randomly

assigned SAP volunteer is unlikely to be feasible, reliable, or conclusive.
BORP, in concept, solves many of the anticipated challenges, but DBI could
still face resource shortfalls if only a minority of the taller, larger, and more
complicated buildings are enrolled.

6-14 Part 6: Post-earthquake Reoccupancy and Safety Evaluations ATC-119-1

Recommendation: DBI together with the Office of Resilience and Capital
Planning (ORCP) should study the feasibility, benefits, and costs of extending
BORP through legislative mandates, triggers, or funded incentive programs
with the following considerations:

• Participation in BORP should be required for all new tall buildings.

Further, DBI, together with the Building Inspection Commission (BIC),
should consider requiring a version of BORP for all new buildings meeting
certain conditions related to their size, complexity, use and occupancy,
occupant load, or location within a recovery-critical neighborhood. This
would impose a small additional cost on building developers for planning
and documentation, but it would be marginal if produced as part of
construction documentation by the design team or by building
management during commissioning.

• Buildings identified as recovery-critical, or especially vulnerable, or

located in a neighborhood (perhaps the Financial District) especially
vulnerable to interaction and adjacency effects—that is, where many
buildings will be unable to reoccupy or recover because of damage to
other properties, should be subject to an extended BORP.

• For a mandatory or triggered program, a streamlined or “light” version of

BORP would likely be beneficial, though this would need to be balanced
with the recommendation to extend BORP beyond reoccupancy to
functional recovery (see Section 4.2.10 of this Part).
• To encourage more voluntary participation in BORP, DBI should increase
its outreach and marketing for the program and should consider additional
incentives.
• DBI should extend its BORP outreach and marketing to major employers
and tenants in buildings and neighborhoods where high BORP
participation would be especially beneficial to the City.

4.2.3 Personnel and Financing

In general, BORP engineers are deputized by the City and are thus
authorized to post ATC-20 placards (CCSF, 2011, Section III). However,
they are not eligible for the liability protection afforded to SAP volunteers
under Business and Professions Code Section 5536.27 because they are
engaged and compensated by the building owners. It is unclear whether
BORP engineers are eligible for state-funded workers compensation as SAP
volunteers are. Administrative aspects of BORP are discussed further in
Part 5.

ATC-119-1 Part 6: Post-earthquake Reoccupancy and Safety Evaluations 6-15

 4.2.4 BORP Report Organization

The BORP guidelines (CCSF, 2011) serve two purposes: (1) they describe
what the final BORP plan, kept at the building with copies held by the
retained engineers, should include; and (2) they describe what the BORP
engineer must submit to the City for certification. The latter purpose tends to
determine each BORP report’s organization, and the result is that most
reports look more like submittals to the City than practical tools to be used
with urgency in the hours after a damaging earthquake. The following
recommendations are based on a review of selected BORP reports made
available by DBI during this study.

Recommendation: Modify the BORP instructions to ensure that the

materials submitted for DBI approval are separate from the materials
intended to be kept at the building and used to conduct the post-earthquake
inspection with the following considerations:

• Inspector qualifications and certifications and submittal checklists should

be put in a separate section for building staff reference, while technical
materials needed to conduct the inspection should be in a separate front
section for inspector convenience.
• DBI should clarify and enhance Sections IV.D.2 and IV.D.3 of its program
guidelines (CCSF, 2011), which currently call for “detailed instructions”
regarding the scope and conduct of the inspection. In addition to
requiring compliance with ATC-20 Detailed evaluation requirements,
BORP guidelines should call for inspection instructions to follow the
outline of the two-page ATC-20 Detailed Evaluation Safety Assessment
Form. The blank form in the BORP plan should be pre-filled with building-
specific information—as should the DBI database that will hold the
inspection findings. In addition, each line of the form should have a
corresponding section in the BORP plan with information regarding the
relevant locations in the building, their built condition, possible damage
modes, and inspection procedures and tips.
• DBI should modify its instructions to require the listing of clear instructions
for initiating the inspection process on the first page of each BORP plan.
Each building’s team may decide whether the building staff or the
engineer will initiate the inspection, but the plan’s first page should be
clear and appropriate for the responsible party.
• DBI should recommend that each BORP plan should have separate pull-
out sections for each expert inspector, e.g., structural, mechanical, life
safety, or elevator. This would facilitate each inspection by avoiding

6-16 Part 6: Post-earthquake Reoccupancy and Safety Evaluations ATC-119-1

 duplication and other material not needed to complete the inspector’s
assigned scope.

4.2.5 BORP Report Content

A review of selected BORP reports made available by DBI during this study
show that content and completeness varies within the program. More
uniform content requirements, together with the more pragmatic report
organization recommended above, should improve reliability and could
reduce costs by eliminating unnecessary or vague requirements. As
recommended above, BORP report content related to specific building
components or conditions should be organized to correspond to the ATC-20
Detailed Evaluation Safety Assessment Form.

Recommendations. Update and clarify BORP instructions to ensure that

each BORP report contains all relevant material, considering current
technologies and practices.
• BORP instructions should require information on potential pounding, plan
irregularities, and vertical irregularities. This information should be linked
to the ATC-20 form sections on building description and other structural
hazards.
• DBI should update its program instructions to require information
regarding the building’s compliance with Chapter 4E, Building Façade
Inspection and Maintenance, of the San Francisco Existing Building Code
(CCSF, 2016) and Administrative Bulletin (AB) 110, Building Façade
Inspection and Maintenance (CCSF, 2017a).
• DBI should update its program instructions to clarify requirements
regarding photographs. “Before” photographs should not be necessary
for typical conditions, but they can be useful to describe a damage-prone
pre-earthquake condition or to document pre-earthquake damage; if
used, they should be linked to the appropriate line in the ATC-20 form.
“After” photos should be provided for cases where damage is perishable
or to document the change relative to a “before” photo, but the current
DBI instructions do not require them. “After” photos can also be useful to
document the inspection process. Wherever photos are included, they
must be detailed enough to convey the condition of interest, and
accompanied by text captions as needed.
• DBI should update its program instructions regarding the required building
information. As recommended above, the required information should be
keyed to the ATC-20 form, and the form itself should be pre-filled
wherever possible. In describing the building, the instructions should

ATC-119-1 Part 6: Post-earthquake Reoccupancy and Safety Evaluations 6-17

 distinguish the seismic-force-resisting system(s) from the gravity
system(s); should distinguish the seismic-force-resisting system in each
principal direction; and should specify combinations and changes in the
lateral system(s) over the height of the building. DBI should consider
recommending a tabular format to improve completeness and
consistency.
• DBI should clarify its instruction (Section VII.F) to “take preventive
measures regarding gas leaks, release of hazardous materials, or other
life-safety mitigation.” The intention of this instruction is not clear. The
instruction should be coordinated with Section VII instructions regarding
notification of DBI.
• DBI should clarify its instruction (Section IV.G): “At owner’s and
inspector’s discretion, non-structural hazards may be mitigated per DBI
procedures.” The intention of this instruction is not clear. It should be
clarified whether this instruction is regarding pre-earthquake mitigation or
whether it is related to eliminating dangerous conditions related to
earthquake damage to nonstructural components. Also, the specific DBI
procedures referenced should be clarified.
• DBI should consider requiring use of the ATC-20 Fixed Equipment
Checklist. This checklist is typically used only for essential facilities, but it
could be a basis for expanding the BORP equipment list and for ensuring
consistency and completeness. ASCE/SEI 41-17, Seismic Evaluation
and Retrofit of Existing Buildings, (ASCE, 2017) and FEMA E-74,
Reducing the Risks of Nonstructural Earthquake Damage: A Practical
Guide, (FEMA, 2012) also provide checklists that should be considered.

4.2.6 Use of Instrumentation

BORP guidelines explicitly allow the use of ground motion records to assist
damage finding and posting, and even to reduce the amount of inspection
(CCSF, 2011, Section IV.D.5 and Appendix A). The guidelines do not,
however, provide any technical procedures or criteria to follow. In particular,
they do not require a pre-earthquake structural analysis to calibrate the
instruments and establish baselines. See also the discussion of inspection
techniques in Chapter 7 of this Part.

Recommendation: DBI, perhaps in collaboration with the Structural

Engineers Association of Northern California (SEAONC) and the state’s
Strong Motion Instrumentation Program (SMIP), should establish criteria for
the use of recorded motions for finding damage and assessing building
safety. In the interim, if recorded motions are to be allowed for BORP, DBI

6-18 Part 6: Post-earthquake Reoccupancy and Safety Evaluations ATC-119-1

should establish minimum requirements for documentation and approval of
the findings and conclusions by BORP engineers.

4.2.7 Cordons and Barricades

BORP instructions (CCSF, 2011, Sections VII.B and VII.C) give unclear
guidance on barricading, suggesting that the BORP team should “arrange for
barricading of all unsafe areas” but should also “contact DBI immediately” so
that City staff can erect barricades or establish cordons for the public areas.
Meanwhile, ATC-20 procedures suggest that volunteer SAP inspectors can
undertake this work in public areas as well.

Recommendation: Clarify and update both BORP instructions and any San
Francisco-specific SAP procedures regarding the division of responsibility for
establishing cordons and barricades.
• DBI, together with the Department of Public Works (DPW), should clarify
and distinguish the roles of City staff and deputized inspectors.
• BORP instructions should also distinguish the roles of SAP volunteers
and BORP inspectors, as the latter also have contractual responsibilities
on behalf of their individual clients.
• The instructions, which currently discuss only the barricading of public
areas, should also recommend or instruct appropriate practices for cases
where an adjacent building threatens the building being inspected (both of
which are private spaces), and where the building being inspected
threatens its neighbors.

4.2.8 Use of FEMA 352

BORP instructions (CCSF, 2011, Section IV.D.1) allow and “recommend” the
use of FEMA 352, Recommended Postearthquake Evaluation and Repair
Criteria for Welded Steel Moment-Frame Buildings (FEMA, 2000), but “BORP
does not require connection inspection.” Inspection of individual connections
is fundamental to FEMA 352 and is the main difference between that system-
specific document and the more generic ATC-20. Thus, BORP instructions
appear to conflict with ATC-20, which references FEMA 352 for the Detailed
evaluation that BORP anticipates, and with FEMA 352 itself. Further, Section
IV.D.5 of BORP instructions allows instrument recordings to be used “as a
means of reducing the percentage of joints required to be inspected after an
earthquake,” and Appendix A makes a similar allowance. This instruction is
confusing because if inspection of the welded connections is not required,
there would be no need for instrument recordings to reduce the scope.

ATC-119-1 Part 6: Post-earthquake Reoccupancy and Safety Evaluations 6-19

 Recommendation: Develop BORP instructions for the use of FEMA 352 in
coordination with a general FEMA 352 policy.
• Even if FEMA 352 is not required, DBI should improve its BORP
instructions by adding specific provisions for acceptable voluntary use.
Consistent use of FEMA 352 within BORP will require DBI instructions
regarding the inspection triggers, minimum connection inspection scope
(pre- and post-posting), and connection selection methodology, as well as
administrative instructions regarding posting terminology, inspection
documentation, and coordination with ATC-20.
• DBI should revise its program instructions (CCSF, 2011, Sections IV.D.1,
IV.D.5, and Appendix A) for greater consistency with FEMA 352
principles. Even where FEMA 352 is not used specifically, DBI should
require at least a specified number of connection inspections (pre- and
post-posting).
• To facilitate connection inspections, DBI should revise its BORP
instructions to require pre-selection of connections to be inspected post-
earthquake, in order to address questions regarding access, fireproofing
removal, or necessary tools. Pre-selection will also facilitate
documentation and reporting.
• Since welded joints can include pre-earthquake flaws that become
confusing or controversial when found after an earthquake, DBI should
recommend pre-earthquake inspection and testing to set a baseline for
comparison with post-earthquake findings.

4.2.9 Concrete Structures

BORP guidelines reference FEMA 352 for welded steel structures but do not
reference similar technical documents for other structure types. For
Engineering evaluations, however, ATC-20 does reference FEMA 306,
Evaluation of Earthquake Damaged Concrete and Masonry Wall Buildings
(FEMA, 1998). For BORP projects, where engineers have the opportunity to
use recorded motions, review plans, perform destructive investigations, and
run structural analyses, use of FEMA 306 would be appropriate.

Recommendation: DBI, with input from SEAONC, should consider

supplementing BORP instructions with guidance on the use of FEMA 306 for
concrete structures.

4.2.10 Functional Recovery

DBI recognizes BORP as a way to support “expeditious business resumption”

in a way that SAP cannot (CCSF, 2011). With some enhancements, and at

6-20 Part 6: Post-earthquake Reoccupancy and Safety Evaluations ATC-119-1

the owner’s expense, BORP procedures might be enhanced to serve
functional recovery goals in addition to safety and reoccupancy goals. The
Department of Public Works already uses BORP-like procedures to facilitate
recovery of City-owned facilities.

Recommendation: Require certain existing buildings to file recovery plans.

• DBI, together with ORCP, should consider enhancing the BORP
guidelines with a module geared to functional recovery of the building and
of individual tenant spaces. The module should be optional and designed
not to interfere with the first priority of safe reoccupancy. For recovery-
critical buildings, DBI and ORCP should consider requiring use of the
functional recovery module as a condition of BORP certification.
• Beyond recovery-critical buildings, DBI and other relevant agencies
should consider a requirement or incentive for selected buildings to
produce and file a recovery plan. The required plan would include much
of the information currently required by BORP, and would need regular
updating, but it would omit the requirement to keep engineers and
inspectors on retainer. Post-earthquake safety inspection would likely still
be done by SAP volunteers, but the plan would include information that
might assist them, as well as pre-compiled resources for owners and
tenants to address recovery issues beyond the scope of SAP and
ATC-20. The requirement to produce a recovery plan would nudge
owners to think through the effects of potential damage and even to
consider enrolling voluntarily in BORP.
• Alternatively, DBI and other relevant agencies should study a requirement
for selected buildings to report building data for use by DBI in prioritizing
post-earthquake inspections by City staff or by SAP volunteers. The data
would include much of the structure and system information currently
required in BORP plans but not normally available on short notice to SAP
inspectors. The data would be used to support the typical ATC-20
process, enabling a more informed inspection and ideally resulting in a
more reliable and permanent posting. Since document review is not part
of the normal ATC-20 Rapid evaluation procedure, this system would be
most effective if paired with a program of specialized inspectors assigned
to specific building groups, as recommended for early recovery in Section
3.2.2 of this Part.

ATC-119-1 Part 6: Post-earthquake Reoccupancy and Safety Evaluations 6-21

Chapter 5

ATC-20

5.1 ATC-20 as Criteria for SAP and BORP

ATC-20, Procedures for Postearthquake Safety Evaluation of Buildings,

(ATC, 1989) was developed with California and Federal funding in the late
1980s and published just in time to be used after the 1989 Loma Prieta
earthquake. The ATC-20 procedures and criteria have been used by SAP
ever since.

The most current version of the ATC-20 procedures are documented in

ATC-20-1, Field Manual: Postearthquake Safety Evaluation of Buildings
(ATC, 2005), and are applicable to buildings of all heights and offer three
evaluation methods:
• A Rapid Evaluation is expected to take only 10 to 30 minutes per building
and is designed to be conducted without access to the building interior,
though interior inspection is recommended if feasible. The Rapid
Evaluation is most suitable where small buildings need to be inspected in
high numbers, and it is most effective for identifying obviously unsafe
conditions. As such, it is often inconclusive but still valuable as a triage
measure.
• A Detailed Evaluation is expected to take one to four hours per building,
including a full interior inspection and greater emphasis on details of the
structural system. The Detailed Evaluation is used to resolve an
inconclusive Rapid Evaluation (Cal OES, 2015).
• An Engineering Evaluation is expected to be used only when structural
analysis or destructive investigation are deemed necessary to resolve an
inconclusive Rapid or Detailed Evaluation.

Rapid and Detailed Evaluations both involve completion of a standard report

form. The Engineering Evaluation is customized and expected to yield a
more thorough report. SAP guidelines expect that Engineering evaluations
will be performed at the owner’s expense but may be requested by the
jurisdiction (Cal OES, 2015). When requested, a building is slated for
Engineering evaluation based on the results of a preceding Rapid and/or
Detailed evaluation, not based on its height or complexity alone.

ATC-119-1 Part 6: Post-earthquake Reoccupancy and Safety Inspections 6-23

 While ATC-20 is focused on safe reoccupancy, the Rapid and Detailed
evaluation forms do ask the inspector to estimate the observed damage as a
percentage of the building’s replacement cost. This is understood as a
response to the jurisdiction’s need to comply with Federal government
requirements regarding disaster declarations (SEAONC, 2013).

SAP training and implementation are premised on the Rapid evaluation.

Even so, an SAP inspector is expected to attempt to inspect the interior,
including basements and “every floor, including roof and penthouse(s),” as
well as nonstructural components and potential hazardous material spills
(SEAONC, 2013).

For a tall building, such an evaluation would no doubt take more than the
typical 20 minutes expected for high-volume Rapid evaluations. Even a
Detailed evaluation of a relatively modern tall building might be inconclusive if
the structural elements are hidden and the details are unknown in advance.
San Francisco’s Building Occupancy Resumption Program (BORP) has been
motivated to resolve these issues. Even though BORP inspections are based
on ATC-20 and are permitted to use the Rapid method, the advanced
planning makes a BORP inspection more like the Detailed method by default.
If analysis is performed in advance (and because the BORP engineer is
selected and compensated by the building owner), BORP can even achieve
some of the benefits of the Engineering method even if only a Rapid or
Detailed evaluation form is submitted.

Regardless of which method is used, the ATC-20 procedure is intended to

result in a “posting” indicating the building’s safety and readiness for
reoccupancy. Table 6-1 shows the three available postings with the ATC
descriptions of their intended meaning.

The RESTRICTED USE placard has multiple uses. The Yellow placard is
typically used for an inconclusive Rapid or Detailed evaluation, as might be
expected for a tall building outside BORP. The same Yellow placard might
be used to indicate a partial restriction, by time or location, to allow as much
use as is deemed safe even while the further evaluation is being planned.

ATC-20 identifies six categories of damage that, if deemed “severe” with

respect to the whole building, should be posted UNSAFE (ATC, 2005, Table
3-1):
• Collapse, partial collapse, or building off foundation
• Building or story significantly out of plumb
• Severe damage to primary structural members

6-24 Part 6: Post-earthquake Reoccupancy and Safety Inspections ATC-119-1

• Chimney, parapet, or similar falling hazard
• Mass ground or slope movement
• Other: Hazardous material release, etc.

Table 6-1 ATC-20 Posting Classifications (ATC, 2005)

Posting Classification
(Placard Color) Descriptive Text on Placard Description of Safety Assessment
INSPECTED “Lawful Occupancy Permitted. This structure has “No apparent hazard is found, although
(Green) been inspected (as indicated below) and no repairs may be required. The original seismic
apparent structural hazard has been found.” resistance is not significantly decreased. No
restriction on use or occupancy.”
RESTRICTED USE “This structure has been inspected and found to “A hazardous condition exists (or is believed
(Yellow) be damaged as described below. Entry, to exist) that requires restrictions on the
occupancy, and lawful use are restricted as occupancy or use of the structure. Entry and
indicated below.” use are restricted as indicated on the placard.”
UNSAFE “Do not enter or occupy. (This placard is not a “Extreme structural or other hazard is present.
(Red) demolition order.) This structure has been There may be imminent risk of further damage
inspected, found to be seriously damaged and is or collapse from creep or aftershocks. Unsafe
unsafe to occupy, as described below. Do not for occupancy or entry, except as authorized
enter, except as specifically authorized in writing by the local building department.”
by jurisdiction. Entry may result in death or
injury.”

Each of the six categories is possible in a tall building, and most are easily
discerned. For a tall building, however, assessment of “severe damage to
primary structural members” can require some expertise and ideally access
to the building’s structural plans (both of which BORP would provide). Even
so, ATC-20 outlines the most common damage patterns for engineered
concrete and steel structures. Table 6-2 summarizes those most applicable
to tall buildings.

In addition to highlighting certain damage patterns, ATC-20 emphasizes

inspection of building features that contribute to plan and vertical structural
irregularities. Irregularities are common to San Francisco’s tall buildings, in
which the plan dimensions at upper levels often differ from those at grade.
Although changes in the architectural floor plan do not always indicate shifts
in the structure, tall building lateral systems do often feature step-backs
toward the top and transfers or offsets toward the base. These structural
irregularities are often hidden even from satellite imagery. This is another
reason why review of the structural plans, perhaps through a BORP process,
is helpful, if not essential, for post-earthquake safety evaluation of tall
buildings.

ATC-119-1 Part 6: Post-earthquake Reoccupancy and Safety Inspections 6-25

Table 6-2 Concrete and Steel Damage Patterns Associated with an ATC-20 UNSAFE Posting
Potential Damage Patterns Relevant to Tall Potential Damage Patterns Relevant to
Structural Concrete Structural Systems Tall Steel Structural Systems
element (ATC, 2005, Chapter 8) (ATC, 2005, Chapter 9)
Foundation Bowing of underground walls Bowing of underground walls
“Fractured” foundation “Fractured” foundation
“Fractures” 1/2" wide in basement slab
Gravity system, and Flat slab punching shear cracking Failure or severe damage to gravity load-
floor or roof framing Failure or severe damage to gravity load-carrying carrying element or connection
element or connection Failure at shear connection between diaphragm
and beam
Diaphragms “Broken” floor-to-wall connection “Bowed, broken, or seriously damaged”
“Broken or seriously damaged” diaphragm diaphragm
Damaged chord or collector
Horizontal cracks 1/2" wide in concrete slab
Columns Buckled or fractured column Buckled or failed column
Spalling with exposure of column vertical reinforcing Shear failure at column base connection
Moment frames “Seriously degraded” moment frames “Seriously degraded” moment frames
Panel zone cracking “Significant” weld or other failure at moment joint
Flange buckling near moment joint
Column splices (by reference to FEMA 352)
Panel zones (by reference to FEMA 352)
Other lateral force Shear walls Diagonal bracing
resisting systems “Failed shear wall” Buckled or stretched brace
Diagonal cracks 1/8 wide extending between floors Broken brace or connection
Failure or slippage of horizontal construction joints Failed chord
Spalling with exposure of vertical reinforcing at
boundary elements
Horizontal cracks 1/8 wide extending through
boundary elements
Wall-pier systems
“Several failed piers at any one story”
Failed spandrel beams

As alternatives to its general inspection guidance, ATC-20 also cites, but

does not require, two FEMA documents for specific structure types: FEMA
306, Evaluation of Earthquake Damaged Concrete and Masonry Wall
Buildings, (FEMA, 1998) and FEMA 352, Recommended Post-earthquake
Evaluation and Repair Criteria for Welded Steel Moment-Frame Buildings, for
Detailed Evaluation of welded steel moment frame structures (FEMA, 2000).

6-26 Part 6: Post-earthquake Reoccupancy and Safety Inspections ATC-119-1

5.2 Issues and Recommendations Regarding ATC-20

ATC-20 has its own development and update process; an update is currently
underway. Further, SAP trainings are produced for broad audiences and
must cover more than San Francisco’s particular conditions. Therefore, it
might be difficult for the City to implement its own customization of ATC-20.
Nevertheless, DBI should establish its own understanding of how it will
interpret and implement ATC-20 guidance, especially through BORP, which
is a San Francisco program. Thus, the following interpretations and
suggested edits of ATC-20 materials might be developed for BORP
guidelines.

Most of the following can apply to non-tall buildings as well as tall ones, and
to implementation of ATC-20 through either SAP or BORP.

5.2.1 Limits on Exterior-Only Inspection

Exterior-only inspection, allowed by ATC-20’s Rapid evaluation method, will

be inconclusive for most tall buildings and for many non-tall buildings. BORP
requires a Detailed evaluation to solve this problem, but any building not
enrolled in BORP will be eligible to use the general ATC-20 procedures. That
said, ATC-20 does acknowledge that certain structure types should only be
posted after interior inspection. Unfortunately, this guidance is not as clear
as it should be.

Recommendation: DBI should clearly list the structure and building types
that may be posted INSPECTED only after an interior inspection.
• For these buildings, unless the building is posted UNSAFE, the posting
must be RESTRICTED USE, and the placard must call for interior
inspection. ATC-20 already recognizes that tilt-up structures are in this
group and that steel braced frames might be.
• DBI should consider defining this group by structural system (with
consideration for cases where the system type is unknown), height or
size, use or occupancy that would be sensitive to interior nonstructural
damage, or other attributes.

5.2.2 Limits on Rapid Evaluation

San Francisco developed BORP as an alternative to SAP in part because the

proper application of ATC-20’s Rapid evaluation method would be
inconclusive for many large and relatively modern buildings in which the
structural system is either not discernible or not observable from even an
interior inspection. Similarly, FEMA 352 was developed as an alternative to
ATC-20 in response to evidence that in welded steel moment frames,

ATC-119-1 Part 6: Post-earthquake Reoccupancy and Safety Inspections 6-27

 significant structural damage is not revealed by obvious story drift or by
nonstructural damage. The same observation has been made in other
structure types (ATC, 2005, p72), and the principle likely applies to many
large steel and concrete structures, including most tall buildings. Further,
even where some structural damage is found, ATC-20 notes that a posting of
UNSAFE applies only where “severe” damage renders the entire structure
dangerous (ATC, 2005, p53). In tall buildings, however, complicated multi-
modal response can make the critical locations harder to identify without an
advance review of the structural design. For these reasons, ATC-20, which is
enormously valuable for most buildings, is less well suited to tall and complex
structures.

Recommendation: DBI should clearly list the structure and building types
that may be posted INSPECTED only after a Detailed evaluation.
• For these buildings, unless the building is posted UNSAFE, the posting
must be RESTRICTED USE, and the placard must call for Detailed
evaluation.
• DBI should consider defining this group by structural system (with
consideration for cases where the system type is unknown), height or
size, or other attributes.
• The Detailed evaluation should involve conditions throughout the building
that are similar to any found damaged in an initial inspection.
• For tall and large buildings, DBI should consider supplementing ATC-20
guidance regarding its limits on UNSAFE postings for severe but local
damage. Local damage might include partial collapse at one end or wing,
floor or floor framing damage at only a small number of the building’s
floors, or connection damage that changes the system characteristics but
does not indicate a distinct collapse mode.

5.2.3 Damage Estimates

For both Rapid and Detailed evaluations, ATC-20 asks for an estimate of
damage as a percentage of the building’s replacement cost. In its SAP
training materials, SEAONC (2013, p66) has noted, “This is a matter of
personal judgment; there is no set formula to calculate this information.” Cal
OES (p3), advises each jurisdiction to have “a system of cost estimation
developed prior to the disaster.”

Recommendation: Develop or update a protocol for damage estimation.

• DBI should ensure that its methodology for converting SAP and BORP
inspectors’ damage estimates and building valuations into loss estimates
is up to date.

6-28 Part 6: Post-earthquake Reoccupancy and Safety Inspections ATC-119-1

• For all buildings, but especially for tall or other large buildings, to ensure
rational and consistent damage estimates, DBI should consider
developing guidelines for SAP and BORP inspectors regarding the
characterization of observed damage in terms of percentage loss.

5.2.4 Placard Use and Tracking

The posting of San Francisco tall buildings by SAP volunteers is likely to

involve considerable uncertainty, for which a RESTRICTED USE placard is
appropriate. However, the RESTRICTED USE placard can be used also to
indicate a firm conclusion applicable to only part of a building. Since the
disposition of a tall building is likely to both affect and be affected by
neighboring properties, these multiple meanings can lead to confusion and
delays. In Christchurch, the ATC-20 forms were modified to include sub-
categories to clarify the reasons for the selected posting. This made it easier
for the jurisdiction to prioritize and track follow-up inspections (ATC, 2014). In
particular, it is useful for DBI’s status tracking to distinguish buildings where
RESTRICTED USE is due to damage to the building itself or due to an
unsafe condition in an adjacent (perhaps tall) building.

Recommendation: Develop or update a protocol for consistent placard use

and tracking, as contemplated by SFBC Section 101A.21.
• Because San Francisco customizes and prints its own placards, DBI
should consider adding check boxes to identify the particular basis for a
RESTRICTED USE or UNSAFE posting. The ATC-20 Safety
Assessment Forms should be modified to match.
• DBI should track the subcategories as well as the overall posting. In
particular, if a building’s posting status is linked to that of an adjacent
building, the linkage should be tracked so that when the controlling
building’s issue is resolved, the dependent buildings’ status is also
updated.

5.2.5 Placard Text

The default placard text can be clearer and more consistent.

Recommendation: Because San Francisco customizes and prints its own

placards, it should consider the following revised text:
• INSPECTED placard: “... no apparent structural immediate safety hazard
has been found.”
• RESTRICTED USE placard: “Entry, occupancy, and lawful use are
restricted as indicated below.”

ATC-119-1 Part 6: Post-earthquake Reoccupancy and Safety Inspections 6-29

Chapter 6

FEMA 352

6.1 Background

FEMA 352, Recommended Post-earthquake Evaluation and Repair Criteria

for Welded Steel Moment-Frame Buildings (FEMA, 2000a), is one of a series
of guideline and background documents produced in response to steel frame
damage observed in the 1994 Northridge earthquake. FEMA 352 is
applicable to welded steel moment-frame buildings of all heights.

FEMA 352 is premised on two Northridge lessons learned after the original
publication of ATC-20 in 1989. First, pre-Northridge welded steel moment
frames are prone to connection fracture, especially at welded joints. The
1989 version of ATC-20 had acknowledged the possibility of connection
damage, particularly for “semirigid moment connections made with angles,
tees, and plates in bolted or riveted joints,” which it characterized as
“nonductile.” It called for the evaluator to “spot-check several beam-column
joints,” but only those deemed “nonrigid.” For the welded connection
common at the time, ATC-20 noted that “ductile moment frames are expected
to sustain yielding in beams while columns remain elastic.”

Second, Northridge fractures were commonly found in buildings that showed

few visible signs of structural damage. The ATC-20 Rapid evaluation
method, in which “ordinarily, only the exterior of the building is inspected,”
had been based on the experience that even when the structure is not
exposed, serious structural damage will show as a visible lean or as
corresponding nonstructural damage. By contrast, FEMA 352 notes, “often,
damage to steel moment-frame buildings cannot be detected by rapid
evaluations” (Section 1.5) and “The absence of significant observable
damage ... in a preliminary evaluation ... should not be used as an indication
that detailed evaluations are not required” (Section 3.3.4.3).

The recognition that exterior-only Rapid evaluation might be inadequate is not

without precedent. ATC-20 already says that concrete tilt-up buildings should
not be posted without an interior inspection of the roof-to-wall connections
and cites examples of steel braced frame damage that was only observed
after nonstructural finishes and fireproofing were removed (ATC, 2005). By

ATC-119-1 Part 6: Post-earthquake Reoccupancy and Safety Inspections 6-31

 referencing FEMA 352, ATC-20 is, if obliquely, saying the same about pre-
Northridge welded steel moment frames.

FEMA 352 thus differs conceptually from ATC-20 in its fundamental attitude
toward non-visible damage. This leads to two pragmatic differences: FEMA
352 uses different terminology for its inspection protocols and sets a higher
minimum scope of inspection both before the building is posted and after it
receives a green INSPECTED placard.

Table 6-3 compares the different evaluation methods of ATC-20 and FEMA
352. The key differences:
• Terminology: What ATC-20 calls a Detailed evaluation, FEMA 352 calls
Preliminary evaluation Step 4, involving inspection of a specified number
of beam-column connections. What ATC-20 calls an Engineering
evaluation, FEMA 352 calls a Detailed evaluation. Thus, simplistic
reference to “Detailed” evaluation can be confusing.
• Required scope prior to initial posting: ATC-20 allows a Rapid evaluation,
even an exterior-only evaluation. In many cases the ATC-20 evaluator
will recognize that the unknown conditions merit a Yellow RESTRICTED
USE placard, but ATC-20 does not require it. ATC-20-2, Addendum to
the ATC-20 Postearthquake Building Safety Evaluation Procedures (ATC,
1995, Section 4.4.1) acknowledged that “Rapid Evaluations can be
inconclusive,” but still allowed that “[Inspectors] may choose to post the
structures UNSAFE or RESTRICTED USE until more detailed inspection
is performed.” By contrast, compliance with FEMA 352 requires actual
connection inspection prior to any posting. Thus, by FEMA 352, an
ATC-20 Rapid evaluation is always incomplete and should always yield a
Yellow placard, unless a Red placard is already justified from the quick
review (Bonowitz, 2013).
• Required scope after initial posting: If an ATC-20 Rapid evaluation
results in a Green INSPECTED posting, the ATC-20 scope is complete.
However, ATC-20-3 presents a case study that suggests, but does not
require, connection inspection: “[S]teel frame buildings that have
experienced severe seismic motions [should] be inspected for possible
connection ... failures, even if the building is apparently not structurally
damaged and has been posted INSPECTED.” FEMA 352 would adopt
this more thorough approach. Using FEMA 352, the inspection scope is
not complete until the FEMA 352 Detailed evaluation assesses the
significance of the inspection findings. Reoccupancy is allowed by
inspector judgment after the initial Yellow placard is posted, but the FEMA
352 Detailed inspection must still be completed.

6-32 Part 6: Post-earthquake Reoccupancy and Safety Inspections ATC-119-1

Table 6-3 Comparison of ATC-20 and FEMA 352 Terminology and Evaluation
Methods
ATC-20 Evaluation Method Equivalent FEMA 352 Evaluation Method
RAPID EVALUATION PRELIMINARY EVALUATION, Steps 1 – 3
Required for posting Required for posting
Exterior-only inspection allowed Interior and exterior inspection required
Review of architectural and structural plans recommended
DETAILED EVALUATION PRELIMINARY EVALUATION, Step 4
Not required for posting Required for posting
Interior inspection typical Visual inspection of specified number of connections
Connection inspection recommended * required *

ENGINEERING EVALUATION DETAILED EVALUATION

Not required for posting Required even if posted INSPECTED
Based on connection inspections *
May be prescriptive or analytical
* BORP does not require connection inspection even when FEMA 352 is used. See Section 4.2.9 of this Part.

Table 6-4 provides a further breakdown of differences between the minimum

inspection scope each document requires prior to an initial posting. FEMA
352 expands the ATC-20 Rapid evaluation scope by requiring an interior
inspection and at least visual inspection of a specified number of welded
connections. FEMA 352 offers different methods for selecting the
connections to be visually inspected in Preliminary Step 4, but they almost
always require four locations per floor. For a tall building (and even for a
short one), this is not feasible in the time normally budgeted for an ATC-20
Rapid evaluation.

Table 6-4 Comparison of Minimum Scope of ATC-20 Rapid and FEMA 352
Preliminary Methods
ATC-20 FEMA 352
Evaluation Scope Rapid Evaluation Preliminary Evaluation
Exterior visual inspection Required Required (Step 1)
Interior visual inspection Not required Required (Step 2)
Hazardous materials review required
Elevator inspection required
Permanent drift Required, visual only Required (Step 3)
measurement
Connection visual Not required Required (Step 4)
inspection Removal of fireproofing required only if
damage is indicated
Ultrasonic or other testing not required

ATC-119-1 Part 6: Post-earthquake Reoccupancy and Safety Inspections 6-33

 FEMA 352 has not yet been implemented after an earthquake. A simulation
by the SEAONC Existing Buildings Committee Steel Frame Subcommittee
(2004) showed that FEMA 352’s statistical sampling procedures can yield
widely inconsistent results depending both on the method used and on
choices left to the inspecting engineer. The study was based on real 5-story
buildings damaged in the Northridge earthquake.

6.2 Issues and Recommendations Regarding FEMA 352

Because FEMA 352 is not a standard, DBI can develop its own
implementation procedures. Specific recommendations for applying FEMA
352 within BORP are provided in Section 4.2.8 of this Part.

6.2.1 Implementation of FEMA 352

Although not an official standard, FEMA 352 is the only guideline specifically
addressing inspection and evaluation of pre-Northridge welded steel moment
frames. Nevertheless, it lacks a track record of actual implementation and
has not been condensed into practical and enforceable provisions. Further,
use of FEMA 352 as written requires some choices and discretion by the
engineer. As discussed above, inspection of individual connections is
fundamental to FEMA 352 and is the main difference between this system-
specific document and the more generic ATC-20. Without the connection
inspection protocol, FEMA 352 adds very little to ATC-20 except for modestly
different inspection triggers. DBI, together with the Structural Subcommittee
of the Code Advisory Committee (and perhaps with the SEAONC Existing
Buildings Committee) should study the feasibility, benefits, and costs of
requiring some form of FEMA 352 compliance for welded steel moment frame
structures.

Recommendation: Develop a new Administrative Bulletin (AB) for post-

earthquake inspection and evaluation of welded steel moment frames with
the following considerations:
• The AB should address both safety inspection and structural evaluation,
as FEMA 352 covers both topics. The AB should reference FEMA 352
and set specific requirements where the document allows options. See
also Part 4.
• The AB should set San Francisco policy regarding connection inspection
triggers, minimum connection inspection scope (pre- and post-posting),
and connection selection methodology. It should provide administrative
instructions regarding posting terminology, inspection documentation, and
coordination with SAP and ATC-20. DBI should consider simply clarifying
that given a certain shaking level or other inspection trigger, the building

6-34 Part 6: Post-earthquake Reoccupancy and Safety Inspections ATC-119-1

 must be posted RESTRICTED USE (indicating an incomplete inspection)
until the minimum number of connection inspections is complete.
• Review FEMA 352 for application to representative San Francisco
structures. In particular, review the findings of the SEAONC Existing
Buildings Committee Steel Frame Subcommittee (2004) showing that
results are sensitive to the joint sampling method.
• The AB should apply FEMA 352 to interpret generic SAP and ATC-20
guidance on appropriate posting of inspected buildings.
• DBI should develop tracking procedures to account for the multiple
phases (FEMA 352 “steps”) of a post-earthquake safety inspection and
posting process based on FEMA 352. The procedures should ensure
that required connection inspections are reported and completed, with
postings that properly reflect the interim stages.

6.2.2 Application of FEMA 352 to Loma Prieta Damage

The 1994 Northridge earthquake revealed unexpected damage to dozens of

welded steel moment frame structures throughout greater the Los Angeles
area. Some of these were tall buildings; most were not. In most cases,
connection damage did not reveal itself through obvious damage to finishes
or noticeable changes in behavior under everyday use. Five years earlier,
San Francisco’s steel buildings were subjected to the Loma Prieta
earthquake. Without the benefit of lessons that were later learned in
Northridge, steel buildings in San Francisco were never systematically
inspected for possible welded connection damage following the Loma Prieta
earthquake.

About 30 Bay Area steel moment frame buildings are known to have been
inspected for possible Loma Prieta damage in the few years following the
Northridge earthquake. At least five buildings, all between 30 and 60 miles
from the Loma Prieta epicenter, but none in San Francisco proper, were
found to have weld fractures. Four of these damaged buildings were 12
stories or taller. Although details from these voluntary inspections are
incomplete, the heaviest reported damage appears to have been in buildings
on soft soil where available information on peak ground accelerations (PGA)
exceeded 0.25g. For two of the damaged buildings, however, the closest
recorded ground motion had a PGA of only 0.18g (FEMA 355E, 2000b,
Section 5.9).

In 1996, the Structural Engineers Association of California, Applied

Technology Council, and Consortium of Universities for Research in
Earthquake Engineering (SAC) Joint Venture notified Bay Area building

ATC-119-1 Part 6: Post-earthquake Reoccupancy and Safety Inspections 6-35

 officials of these findings, but no Bay Area city is known to have adopted an
official policy or implemented a formal inspection program in response. Since
1996, it is likley that additional buildings with welded steel moment frames
have been inspected for Loma Prieta damage, either out of an owner’s self-
interest or in the course of typical pre-purchase investigations. Very likely,
some of these have been in San Francisco, and some have probably been
tall buildings, but the procedures for these voluntary inspections are not
regulated by DBI, and thus, the findings are not recorded by the City.

Almost 30 years after Loma Prieta, should San Francisco implement a

systematic inspection program for its pre-Northridge welded steel moment
frames? Inspection can be costly and disruptive. If there is no reason to
expect damage, or if any reasonably expected damage would have little
effect on future performance, why expend the resources? A key issue
involves the ground motion that should trigger a FEMA 352 inspection.

FEMA 352 itself addresses this question. FEMA 352 Section 3.2
recommends seven inspection triggers for a given building based on
estimated ground motion at the site and known damage nearby. Specifically,
for a city with high seismicity like San Francisco, FEMA 352 recommends a
protocol of joint inspections where the PGA exceeds 0.25g, where nearby
unreinforced masonry buildings show “prevalent” full or partial collapse,
where nearby nonstructural damage is at “high levels,” or where other nearby
buildings show “considerable” damage.

Except for the PGA trigger, these conditions allow for ample judgment.
Considering Loma Prieta building performance in or near downtown San
Francisco, examples of FEMA 352 indicators of potential damage were
observed in different parts of the city. Whether or not similar levels of
damage would trigger FEMA 352 inspections for downtown buildings after a
future earthquake would be a matter of building official judgment.

The PGA trigger in FEMA 352 is more quantitative, but it was based on
sound (but still incomplete) data from just one event (Northridge) and on
relatively sparse data, described above, from Loma Prieta. The majority of
Northridge steel frame damage occurred in Los Angeles neighborhoods with
PGA values of 0.50g or higher (USGS, 1994). In downtown Los Angeles,
PGA values were under 0.30g, with some under 0.20g, and very little damage
was found in buildings of any height. However, post-Northridge inspection
there, especially among tall buildings, was relatively rare, and it effectively
stopped when the scope of the city’s mandatory inspection ordinance
excluded most of downtown. “[I]t is possible that no damage was found in

6-36 Part 6: Post-earthquake Reoccupancy and Safety Inspections ATC-119-1

downtown structures because so little inspection had been done there early
in 1994” (FEMA 355E, Section 6.3.2).

Loma Prieta ground motion records in San Francisco are limited. Available
information includes data from recording stations and data inferred from
Modified Mercalli Intensity (MMI) and Did You Feel It (DYFI) observations
contained in Loma Prieta reconnaissance reports (USGS, 1989). Stations
recorded PGA values between 0.06g and 0.16g, which are below FEMA 352
triggers for inspection, but observational data at several downtown locations
suggest inferred PGA values of 0.24g, which is approaching the threshold
value of 0.25g in FEMA 352 for areas of high seismicity. Additional
observational data include 0.45g near the intersection of Market Street and
Van Ness Avenue and more than 0.8g at locations in the Marina and near the
Ferry Terminal. Although these data are outliers, they indicate a potential for
more significant shaking in selected areas.

As a policy matter, the San Francisco Existing Building Code is clear:

earthquake damage must be found and repaired. Beyond repair, seismic
retrofit might even be triggered, as discussed in Part 4. How to implement
these code provisions so long after the event raises interesting questions
worthy of a coherent and official reponse. Had FEMA 352 been available as
a guide in 1989, portions of San Francisco might have been considered to be
at, or above, the triggering PGA value for inspection, and the presence of
damage to other classes of buildings would have required the building official
to make a judgment regarding the need for inspection. Considering the
question again in 2018, the most useful information would come from
voluntary inspections performed—but not reported to the City—over the past
two decades. Information from voluntary inspections, in combination with
recorded or inferred ground motion data, could be used to confirm the
threshold indicators for possible damage in the Loma Prieta earthquake.

Recommendation: Apply the repair provisions of the San Francisco Existing

Building Code with respect to possible Loma Prieta damage to welded steel
moment frames.
• Criteria should be based on FEMA 352 and on the Administrative Bulletin
recommended in Section 6.2.1 of this report. DBI should consider
alternative criteria to accommodate voluntary inspection and repair work
already performed in good faith.
• Application of FEMA 352 and the AB should consider available
information and community consensus regarding appropriate ground
motion triggers and damage indicators. In particular, the City should seek
information from voluntary inspections performed since 1989 as a way of

ATC-119-1 Part 6: Post-earthquake Reoccupancy and Safety Inspections 6-37

 gauging the likelihood of weld damage in San Francisco buildings of
various heights.
• Enforcement should focus on finding and repairing damage. While the
SFEBC also triggers seismic work in cases of substantial structural
damage and disproportionate damage (see Part 4), the City might
consider waiving these triggers, given the unusual circumstances of post-
earthquake inspection so long after the damaging event.
• While enforcement is already supported by SFEBC provisions and should
not require new legislation, the City should consider implementing the
applicable code provisions through a special program to accommodate
the unusual circumstances. The program might involve specific
notifications and outreach to building owners and tenants, as well as
liberal deadlines to allow phased or flexible compliance.

6-38 Part 6: Post-earthquake Reoccupancy and Safety Inspections ATC-119-1

Chapter 7

Inspection Technologies

In addition to the policies and programs discussed in this part, the tools and
methods of inspection, damage-finding, and documentation, while outside the
scope of this report, are worthy of review as well. Some of these are new or
emerging technologies, while others are merely common sense applications
of existing ideas that might be required, funded, or encouraged by a building
code, a program such as the City’s Building Occupancy and Resumption
Program (BORP) or a City initiative to develop standards and best practices.
As City policy, some of these might benefit from (or be subject to) review by
the City’s Committee on Information Technology.
• Access openings built into the structure and architecture
• Building-specific social media data recorded or posted by tenants (who
will not be present during the safety inspection)
• Satellite or GPS imagery, for overhead or birdseye views, especially for
comparison of pre- and post-earthquake conditions
• Building instrumentation to record earthquake motions. This is more
feasible than ever, and BORP guidelines already allow (but do not
require) the use of motion records to assist damage-finding and posting.
However, standards for processing, interpreting, and applying recorded
motion data by non-experts do not yet exist. Also of interest are
questions about ownership and publication of data, which might be
private or privileged but have public benefit.
• Measurable building data, especially for comparison of post-earthquake
values with pre-earthquake baseline values. These might include:
o Three-dimensional laser scanning, as used to monitor cliff movements
and selected buildings in Christchurch (ATC, 2014)
o Laser measurements in elevator shafts, for gauging permanent lateral
deformations
o Ambient vibration testing to establish structural periods and mode
shapes
o Strain gauges on steel members

ATC-119-1 Part 6: Post-earthquake Reoccupancy and Safety Inspection 6-39

 o Movement gauges across existing joints, between floors, or between
nonstructural components and the structure.
• Use of unmanned aerial vehicles (UAVs, also known as drones) and
robots. Both were used in Christchurch (ATC, 2014). San Francisco has
a drone policy that authorizes specific uses by different City departments
(CCSF, 2017b), but neither DPW nor DBI—the two departments
responsible for most post-earthquake building safety inspections—is
authorized (DataSF, 2018).
• Computerized or online documentation and automated databasing to
reduce data management demands and improve speed, consistency, and
completeness of reporting.
• Barcoding of buildings to allow smartphone access to building information
for City staff or deputized Cal OES Safety Assessment Program (SAP)
volunteers performing post-earthquake safety inspections.
• Computer or smartphone apps to assist building owners and tenants with
their own reoccupancy and recovery, to reduce the post-earthquake
resource demand on the City.

Recommendation: Study new and emerging technologies and practices for

damage finding and documentation. From the preceding list, the City should
consider additional study and program development for the topics deemed
most likely to improve citywide resilience.

6-40 Part 6: Post-earthquake Reoccupancy and Safety Inspection ATC-119-1

Appendix A

References

ASCE, 2017, Seismic Evaluation and Retrofit of Existing Buildings, ASCE/SEI

41-17, American Society of Civil Engineers Structural Engineering
Institute, Reston, Virginia.
ATC, 1989, Procedures for Postearthquake Safety Evaluation of Buildings,
ATC-20, Applied Technology Council, Redwood City, California.
ATC, 1995, Addendum to the ATC-20 Postearthquake Building Safety
Evaluation Procedures, ATC 20-2, Applied Technology Council.
ATC, 2005, Field Manual: Postearthquake Safety Evaluation of Buildings,
Second Edition, ATC-20-1, Applied Technology Council, Redwood
City, California.
ATC, 2014, Building Safety Evaluation after the February 22, 2011
Christchurch, New Zealand Earthquake: Observations by the ATC
Reconnaissance Team, ATC-109, Applied Technology Council,
Redwood City, California.
ATC, 2018, Recommended Earthquake Performance Goals for San
Francisco's Buildings (ESIP Task A.6.h), final report submitted to the
City and County of San Francisco prepared by the Applied
Technology Council, Redwood City, California.
Bonowitz, D., 2011, Resilience Criteria for Seismic Evaluation of Existing
Buildings, Structural Engineers Association of Northern California,
August 5.
Bonowitz, D., 2013, ATC 20 Simulation, presented as part of SEAONC
ATC-20 Training, Santa Clara, June 22.
Bonowitz, D., 2016, Earthquake Resilience: A New Context for Engineering
Research, presentation to the PEER-SEMM Seminar, Berkeley,
March 14.

ATC-119-1 Part 6: Post-earthquake Reoccupancy and Safety Inspections 6-41

 Cal OES, 2015, Post-Disaster Safety Assessment Program: Guideline to the
Activation and Utilization of Program Resources, California Office of
Emergency Management, Sacramento, California. Available at
http://www.caloes.ca.gov/cal-oes-divisions/recovery/disaster
-mitigation-technical-support/technical-assistance/safety-assessment
-program, last accessed September 18, 2018.
CCSF, 2008, City and County of San Francisco Emergency Response Plan:
Earthquake Annex, Department of Emergency Management, the City
and County of San Francisco, California.
CCSF, 2011, Building Occupancy Resumption Program, Department of
Building Inspection, the City and County of San Francisco, California.
CCSF, 2016, The San Francisco Building Code, the City and County of San
Francisco, California.
CCSF, 2017a, Building Façade Inspection and Maintenance, Administrative
Bulletin 110, Department of Building Inspection, the City and County
of San Francisco, California.
CCSF, 2017b, Citywide Employee Drone Policy, Committee on Information
Technology, the City and County of San Francisco, California.
DataSF, 2018, Authorized Drone Uses, March 18, https://data.sfgov
.org/City-Management-and-Ethics/Authorized-Drone-Uses/5e3s-sj56
FEMA, 1998, Evaluation of Earthquake Damaged Concrete and Masonry
Wall Buildings, Basic Procedures Manual, FEMA 306, prepared by the
Applied Technology for the Federal Emergency Management Agency,
Washington, D.C.
FEMA, 2000a, Recommended Post-earthquake Evaluation and Repair
Criteria for Welded Steel Moment-Frame Buildings, FEMA 352,
prepared by the Structural Engineers Association of California,
Applied Technology Council, and Consortium of Universities for
Research in Earthquake Engineering (SAC) Joint Venture for the
Federal Emergency Management Agency, Washington, D.C.
FEMA, 2000b, State of the Art Report on Past Performance of Steel Moment-
Frame Buildings in Earthquakes, FEMA 355E, prepared by the
Structural Engineers Association of California, Applied Technology
Council, and Consortium of Universities for Research in Earthquake
Engineering (SAC) Joint Venture for the Federal Emergency
Management Agency, Washington, D.C.

6-42 Part 6: Post-earthquake Reoccupancy and Safety Inspections ATC-119-1

FEMA, 2012, Reducing the Risks of Nonstructural Earthquake Damage: A
Practical Guide, FEMA E-74, prepared by the Applied Technology
Council for the Federal Emergency Management Agency,
Washington, D.C.
SEAONC, 2013, Post-Disaster Safety Assessment Program (SAP) ATC-20
Training: Student Manual (Version 11), Structural Engineers
Association of Northern California, San Francisco, California.
SEAONC, 2004, “Post-earthquake evaluation of welded steel moment
frames: case studies using FEMA 352,” in SEAOC 2004 Convention
Proceedings, Structural Engineers Association of California, August.
SPUR, 2012, Safe Enough to Stay, San Francisco, California.

ATC-119-1 Part 6: Post-earthquake Reoccupancy and Safety Inspections 6-43

 PART 7:
Pre-earthquake Evaluation
Chapter 1

Introduction

This Part reviews and explains current San Francisco policy for pre-
earthquake seismic evaluation, especially as it relates to the City’s existing
tall buildings. It recommends modifications to the current policy and
proposes additional study or development of new regulations to address
identified issues. The recommendations presented correspond to
Recommendations 2B and 3H presented in the Summary Recommendations.

This Part addresses aspects of the following Tasks described in the CAPSS
Earthquake Safety Implementation Program Workplan 2012-2042 (CCSF,
2011):
• Task C.1.a: Mandatory evaluation on sale or by deadline of building
types not otherwise covered by the Workplan
• Task C.1.b: Evaluation of buildings retrofitted prior to 1994 or built to
non-conforming performance standards
• Task C.2.a: Mandatory retrofit of older non-ductile concrete residential
buildings
• Task C.2.d: Mandatory evaluation and retrofit of pre-1994 welded steel
moment frame buildings
• Task C.2.e: Mandatory evaluation and retrofit of other low-performance
buildings

1.1 Organization

Chapter 2 presents the current San Francisco practice.

Chapter 3 presents several alternative approaches.

Chapter 4 presents a summary of issues and recommended modifications to

the current policy and proposes additional study or development of new
regulations to address identified issues.

Appendix A presents an annotated set of current San Francisco Existing

Building Code provisions related to this topic.

Appendix B presents a brief discussion on seismic retrofit.

A list of references is provided at the end of the Part.

ATC-119-1 Part 7: Pre-earthquake Evaluation 7-1

Chapter 2

Current Policy for Pre-earthquake

Evaluation

Current San Francisco policy requires evaluation, and possibly seismic

retrofit, when a building undergoes a substantial alteration or a change of
use, but these cases are rare, especially for tall buildings. Other jurisdictions,
as well as some private sector stakeholders, have developed more proactive
policies that San Francisco could adopt.

The San Francisco Existing Building Code (SFEBC) regulates most aspects
of pre-earthquake structural evaluation and retrofit. SFEBC provisions
combine the California Existing Building Code (CEBC, which itself is an
amended version of the International Existing Building Code (IEBC), and San
Francisco amendments.

The SFEBC is organized by project type, recognizing five categories:

additions, alterations, repairs, changes of occupancy, and relocations. Of
these, only alteration and change of occupancy provisions are at all likely to
affect pre-earthquake evaluation of San Francisco tall buildings. In concept,
a tall building could be made even taller with an addition, but even in that rare
case, the current policy is not likely to change; the existing building would
have to be retrofitted to make its seismic performance equivalent to that of a
new building.

Appendix A of this Part presents the SFEBC provisions regarding seismic

evaluation triggered by either alteration or change of occupancy projects.

2.1 2016 SFEBC Alteration Triggers

Modern building codes consider whether an intended alteration project—

typically, a nonstructural renovation or tenant improvement—warrants
unintended, or “triggered,” upgrades to the rest of the building. Seismic
retrofit is one of the potential upgrades.

Codes used throughout California have veered back and forth on the issue.
Until the mid-1970s, model codes used a cost trigger: if the value of the
alteration exceeded 50% of the value of the existing building, a full-building
upgrade was required. When it was recognized that the cost trigger

ATC-119-1 Part 7: Pre-earthquake Evaluation 7-3

 discouraged renovation, especially in multi-family residential buildings, the
provision was weakened to a “do no harm” rule, with essentially no upgrade
trigger at all. A seismic retrofit trigger was restored to the California code
when it adopted the 2006 International Building Code. This version
measured structural effects with the intent of triggering at least a partial
seismic upgrade when the intended alteration would increase loads or
decrease member strength by 10% or more. Significant clarifications were
made in the 2009 edition, so the current alteration provision is still less than
ten years old (Bonowitz et al., 2014). The current CEBC provision is in
Section 403.4 (CBSC, 2016) and annotated in Appendix A of this Part.

In practice, this provision, known to engineers as “the 10% rule,” rarely

results in retrofit. Typical alteration projects do not touch the building’s
structural system and rarely add enough mass to affect the expected seismic
performance. Where the project would trigger seismic work, developers can
reduce the project scope to avoid it. Even when retrofit is triggered, the
current provision has the effect of upgrading only individual structural
elements, without requiring a thorough review of the full earthquake-resisting
structural system or any nonstructural elements. This provision focuses on
the size and impact of the proposed alteration, not on the adequacy of the
structure it is going into. Thus, it is common for alteration projects to be
permitted in seismically deficient or even collapse-prone buildings.

Meanwhile, during the “do no harm” era, some California cities, including San
Francisco, adopted their own alteration triggers or continued to enforce local
interpretations of the earlier ones. While this preserved a policy of well-
intended, “passive” seismic improvements, it led to inconsistent rules and
enforcement across the state (Hoover, 1992). San Francisco retains its own
triggers, developed in the mid-1970s, as amendments to the CEBC, in
SFEBC Sections 403.12.1 and 403.12.2 (CCSF, 2016). They are
summarized as follows:
• For nonstructural alterations, the SFEBC requires a full-structure seismic
retrofit when the intended work would add, remove, or modify
architectural elements in two-thirds or more of the building’s stories within
any two-year period. Thus, a typical tenant improvement on a single floor
would never trigger seismic upgrade in a multi-story building. In a tall
building, a dozen or more floors would all have to undergo significant
renovation within two years for the trigger to apply.
• For structural alterations, the SFEBC requires a full-structure seismic
retrofit when a significant portion of the floor framing, columns, and
bearing walls are altered by the intended project. This sort of structural

7-4 Part 7: Pre-earthquake Evaluation ATC-119-1

 alteration is necessary when an owner wants, for example, to create a
two-story space or an internal stair by removing part of an existing floor.
In a tall building, however, the triggering alteration would almost certainly
have to involve a column or bearing wall in the lower half of the structure.
Such alterations are rare even without triggered seismic improvements.
Especially at the ground level of a tall building where lobby or retail
spaces are frequently renovated, the structure is already so heavy,
expensive, and critical that designers are wise not to touch it.

Each of these San Francisco triggers is more aggressive than the CEBC,
because the “10%” exception that tends to limit triggered work is not offered.
For tall buildings, however, neither is more effective at prompting seismic
improvements, since each trigger applies only in rare cases and can almost
always be avoided by carefully scoping the intended project. Further, even
where the provision might apply, it triggers only structural review, ignoring
any nonstructural deficiencies.

On the other hand, the CEBC alteration trigger and the additional San
Francisco alteration triggers do apply in some cases (particularly in smaller
buildings), and they can be effective as part of a suite of policy tools that
includes mandates and incentives (Bonowitz et al., 2014). However, it should
be recognized that the purpose of the building code’s alteration triggers is not
to find seismically risky buildings and force their retrofit. Rather, the current
purpose of the code’s alteration triggers is to find a balance that encourages
modernization and adaptive reuse of the building stock while leveraging
reasonable opportunities to reduce risk. Further discussion is provided in
Appendix B.

This balancing approach makes sense especially for tall buildings where
retrofit can be extremely costly and disruptive. If the current policy is too
lenient or too slow, the challenge is to adjust the code and to consider
alternative risk reduction measures without necessarily mandating retrofit, as
discussed in Chapter 3 of this Part.

San Francisco’s amendments are due for update and coordination as the
CEBC adopts two new provisions in the 2018 International Existing Building
Code (IEBC, ICC, 2018). First, the 2018 IEBC will allow the use of reduced
loads for alteration-triggered seismic work, much like the reduced seismic
loads already allowed in the SFEBC. Second, the 2018 IEBC will have a new
trigger for “substantial structural alteration,” much like the one already in the
SFEBC. These two changes will bring the CEBC closer to the SFEBC, but
the new versions are not identical to the SFEBC versions, so some

ATC-119-1 Part 7: Pre-earthquake Evaluation 7-5

 coordination to clarify and retain San Francisco’s policy intent will be needed,
as discussed in Appendix A to this Part.

2.2 2016 SFEBC Change of Occupancy Triggers

As with alteration projects, the SFEBC’s provisions for change of occupancy

projects include both the basic trigger of the CEBC, which rarely applies, and
a more aggressive supplemental trigger unique to San Francisco. (Most
change of occupancy projects also involve some alterations. In these typical
cases, both the change of occupancy provisions discussed here and the
alteration provisions discussed in the previous section apply, with the more
demanding provision governing the combined project.)

The basic provision for change of occupancy is in CEBC Section 407.4: If a

change of use or occupancy also changes the assigned risk category, then
the structure is subject to a full seismic upgrade. San Francisco’s new and
existing tall buildings are typically assigned to Risk Category II, a broad group
that includes nearly all residential and commercial buildings. Thus, even if a
building’s primary use were to change from office to mercantile or from multi-
family residential to hotel, that would not change the Risk Category, so the
retrofit trigger would not be pulled.

Risk Category is assigned by CEBC Table 1604.5. An occupant load greater

than 5,000 puts a building in Risk Category III, but a change of use or
occupancy that increases the occupant load from under 5,000 to over 5,000
would be quite rare. Public assembly facilities are also assigned to Risk
Category III, but even if a theater or banquet hall were built into an existing
tall building, that would not change the building’s “primary occupancy,” so the
CEBC Section 407.4 trigger would still not apply.

If part of an existing tall building were to be converted to a residential care

facility (Occupancy I-2) with 50 or more patients, to a K-12 school with more
than 250 students and staff, or to post-secondary education with more than
500 occupants, such a change would move the building into Risk Category
III. In concept, such a change is possible. In practice, the cost of the
triggered seismic work would likely discourage the project.

San Francisco supplements this basic provision with a more likely trigger that
does not require a change of Risk Category. If a change of use involves an
increase in occupant load of more than 10% and more than 100 people, a
retrofit using San Francisco’s reduced seismic loads is triggered. With this
two-part trigger, the creation of a public assembly space or the conversion
from residential to office could actually lead to retrofit. It seems likely that for
San Francisco’s tall buildings, this supplemental change of occupancy trigger

7-6 Part 7: Pre-earthquake Evaluation ATC-119-1

might be pulled even more often than any of the alteration triggers discussed
in Section 2.1. Then again, the additional cost of a triggered retrofit might be
enough to discourage the change, as it often does with alteration projects.

ATC-119-1 Part 7: Pre-earthquake Evaluation 7-7

Chapter 3

Alternative Approaches

As discussed in Chapter 2 of this Part, current San Francisco policy almost

never requires a seismic evaluation or retrofit of an existing tall building even
as it undergoes significant changes and renovations over decades. Retrofits
do occur, but these rare projects are typically done voluntarily, and the
building is often vacant during the work. Few existing buildings present that
kind of opportunity. In all likelihood, the vast majority of seismic work
performed on San Francisco tall buildings is voluntary with respect to the
building code. Therefore, if San Francisco adopts the performance goals
presented in Recommended Earthquake Performance Goals for San
Francisco's Buildings (ATC, 2018) and better performance is required from
existing tall buildings (or any building cohort), alternative approaches should
be considered.

3.1 Mandatory Retrofit Ordinances

One alternative is to set the building code and its triggers aside and work
toward mandatory retrofit ordinances instead. After all, if a community is at
risk from an identifiable group of buildings, that risk is not likely to go down by
waiting for each building, one by one, to propose a substantial voluntary
alteration that happens to pull an arbitrary trigger. To address persistent
community-wide risk, it has always been more effective to mandate retrofit,
whether the concern is public safety, as in the case of unreinforced masonry,
or essential functions, as in the case of California hospitals. San Francisco
has been a leader in developing mandatory programs for specific building
groups that pose risks to the city, such as “soft-story” apartment buildings
(San Francisco Existing Building Code, SFEBC Chapter 4D) and private
schools not regulated by the state (SFEBC Section 329). Indeed, CAPSS
Earthquake Safety Implementation Program Workplan 2012-2042 (CCSF,
2011) already contemplates broad mandatory programs for most buildings
designed before 1994, including non-ductile concrete and welded steel
moment frame structures of any height.

If San Francisco’s tall buildings pose risks comparable to its 2,000

unreinforced masonry buildings or 5,000 soft-story wood-frame buildings, and
if voluntary work is not achieving the City’s reasonable safety and recovery
goals, then the case for a mandatory program should be easy to make—

ATC-119-1 Part 7: Pre-earthquake Evaluation 7-9

 provided the costs are bearable by the community. However, retrofit of a tall
building while it remains occupied and functional can be enormously
expensive and disruptive. Further, tall buildings do not comprise a uniform
class in terms of risk (unlike unreinforced masonry or soft-story buildings) or
function (unlike hospitals, schools, or multi-family residential), so a mandatory
program based only on height will have scattered effects, with benefits that
might be hard to demonstrate.

For these reasons, even if the Workplan 2012-2042 contemplates programs

for all structures of a certain age or type, subsets of tall (or otherwise large or
complex) buildings merit special attention, perhaps even different
prioritization or criteria. A study of non-ductile concrete buildings in Los
Angeles reached this conclusion, recognizing that mandatory retrofit of all
1,500 buildings would not be the most effective policy; in particular, the study
found that mandatory retrofit for only the tallest buildings (8 stories or taller)
would be significantly less effective than retrofit of about the same number of
smaller, older buildings with distinct deficiencies, largely because the retrofit
cost for the tall buildings was so much greater (Anagnos et al., 2016).

Based on these considerations, the most effective and feasible response is

not necessarily to mandate retrofits for dozens or hundreds of buildings.
Instead, it will be more effective to accommodate the resource constraints of
owners and tenants, to ensure that risks are properly valued, and to plan for
effective repair and recovery through programs such as the Building
Occupancy Resumption Program (BORP; discussed in detail in Part 6).

3.2 Aggressive Triggers with Focused Targets

Although San Francisco’s alteration trigger, “2/3”, is more aggressive than the
California Existing Building Code’s (CEBC) basic provision, it is found that for
a tall building it has practically no effect. Similarly, San Francisco’s “10% and
100 persons” change of occupancy trigger is rarely pulled. This section
presents the idea of triggering seismic work more frequently, but only for the
most critical buildings.

Some simple changes to SFEBC Section 403.12.1 or Section 407.4.1 that

might make the triggers apply in more cases include:
• Lowering the triggering value for alterations from “2/3” of a building’s
stories and supplementing it with an absolute number of altered stories.
For example, the trigger could apply whenever any four stories are
altered, so that a few substantial projects would trigger even in a tall
building.

7-10 Part 7: Pre-earthquake Evaluation ATC-119-1

• Extending the two-year accumulation period to five or ten years so that
the trigger cannot be as easily avoided by spacing projects over time.
• Lowering the triggering value for change of occupancy projects, perhaps
from “10 percent [and] 100 persons” to “5 percent [and] 50 persons.”

That said, these triggers are judgmental and largely arbitrary, so any such
alternative values merely reset the line for an owner seeking to avoid an
upgrade. More important, the concern for any aggressive upgrade trigger is
that it can backfire, discouraging otherwise beneficial renovations or even
prompting work to be done without permits (Bonowitz et al., 2014). If
necessary, the more aggressive triggers can be tempered by the following:
• Continuing to allow reduced seismic loads and age benchmarks, as the
SFEBC currently does.
• Allowing a “10% rule” or similar exception to ensure that only substantial
deficiencies need to be addressed.
• Revising the triggers to apply only to certain alterations or changes of use
that truly extend the life of the building. Building-wide mechanical
upgrades, information technology upgrades, and the addition of
amenities, more so than basic improvements to a single tenant space,
indicate a long-term investment that merits a seismic review. (Typical
tenant improvement projects often come when a space is newly leased,
so this alteration trigger could be replaced, or should at least be
coordinated with, a potential acquisition trigger discussed in Section 2.3
of this Part.)
• Revising the triggers to apply only (or more aggressively) to problematic
conditions identified in advance, such as vulnerable structure types or
geologic conditions.
• Revising the triggers to apply only to certain uses or occupancies critical
to city or neighborhood recovery. Of the uses typically found in tall
buildings, ATC (2018) has identified multi-family housing, hotels, and
major employers as deserving of “better than code” recovery goals.

The concept of “extended useful life” has precedents from other cities, from
institutional policies, and even from the CEBC itself. In high seismic areas
like San Francisco, the Federal Government requires seismic evaluation and
potential retrofit when an alteration “significantly extends the building’s useful
life through alterations ... that total more than 30 percent of the replacement
value of the building” (BSSC, 2011). The wording of the provision is meant to
give discretion to each Federal agency, but in practice, most agencies rely on
the quantitative replacement cost trigger. California retains a similar cost

ATC-119-1 Part 7: Pre-earthquake Evaluation 7-11

 trigger for State-owned buildings. CEBC Section 317.3.1 triggers seismic
work when the alteration’s “total construction cost, not including cost of
furnishings, fixtures and equipment, or normal maintenance, for the building
exceeds 25 percent of the construction cost for the replacement of the
existing building.”

Cost triggers were once common in California building codes (see Section
2.1 of this Part), but they are now mostly used where the authority having
jurisdiction is also the owner, or at least has a financial stake, as is the case
for public buildings. For private buildings regulated by a City, cost triggers
are hard to implement, and they have been removed from model codes.
Building officials prefer not to review financial calculations or replacement
cost estimates. Owners dislike the uncertainty of cost triggers; a complete
design and cost estimate is needed before one can know whether the trigger
will be pulled.

The Seattle Existing Building Code (City of Seattle, 2015a) applies the
concept of “extended useful life” with a subjective trigger for “substantial
alteration” and defines the term to include “remodeling ... that substantially
extends the useful physical or economic life of the building or a significant
portion of the building, other than typical tenant remodeling” as well as “a
significant increase in the occupant load of an unreinforced masonry building”
(Section 304.1.1). With this provision, the City of Seattle has replaced a strict
measure of cost or work area with ample building official judgment. In
practice, the provision typically requires a conference with the building official,
and the designation considers project cost, project scope, the vacancy rate,
building age, and other building-specific conditions (City of Seattle, 2015b).
What is unique about this Seattle provision is the desire to consider all of
these factors together and in context and the emphasis on building official
judgment. On one hand, this allows policy flexibility to accommodate special
circumstances; on the other, it introduces subjectivity and uncertainty.

Interestingly, one example given in Seattle’s guidance document (City of

Seattle, 2015b) involves a hypothetical tall building: “A large new restaurant
in a fully occupied high-rise building clearly is not a substantial alteration
project. However, a similar project in an older, partially-occupied, three-story
building is likely to be a substantial alteration.” As noted in Section 2.2 of this
Part, this very example of an assembly occupancy added in the ground floor
of a tall building could pull San Francisco’s current change of occupancy
trigger.

Recent proposals to amend the International Existing Building Code (IEBC)

further extended the idea of a customizable, targeted alteration trigger (ICC,

7-12 Part 7: Pre-earthquake Evaluation ATC-119-1

2016, proposal EB25-16 and similarly, 2012 proposal G213-12). The 2016
proposal, made by the National Council of Structural Engineers Associations
(NCSEA), offered a tabular format with which a jurisdiction would list the
specific building groups that would be subject to a trigger, slightly modified
and filled out in a table format. NCSEA described its proposal (ICC, 2016) as
beneficial to local resilience planning as follows:

The table format allows a jurisdiction to identify the buildings of greatest

interest to local mitigation and resilience plans. We find that in some
jurisdictions the concern is about a particularly vulnerable structure type
(like URM, or non-ductile concrete), in some it is about school safety and
recovery, in some it is about protecting senior or low-income housing, for
some it is about revitalizing a commercial district. A uniform, one-size-
fits-all approach no longer suits the needs of communities thinking about
natural disaster recovery and resilience.

Table 7-1 Tabular Format for Identifying Mitigation Priorities Triggered by Alterations*
Seismic Original
Risk Design Structural Permit
Priority Type Occupancy Cat. Cat. Size Location Attribute Date Other
Emergency — IV F — — — — —
Soft story R — — 2+ st, — Wood frame pre-1978 —
3+ units target story
Private school E — — — — — — 26+ students
Welded steel — — — 3+ st — WSMF pre-1996 —
moment frame
Non-ductile R — — 3+ st — Concrete pre-1980 —
concrete
Central business — — — 8+ st CBD — pre-1996 —
district
(downtown)
Tall B — — 12+ st — — pre-1996 Recovery-
critical
* (Based on ICC, 2016. See text for explanation.)

In its simplest form, the code change would be nothing but a change in
format, from code text to a table. In Table 7-1, for example, the first row,
priority type “Emergency,” reflects a requirement already in the IEBC, CEBC,
and SFEBC: Section 403.4.2 already requires seismic retrofit when a near-
fault Risk Category IV facility undergoes a major alteration. By the NCSEA
proposal, a jurisdiction could keep that row and leave the rest of the table
blank.

ATC-119-1 Part 7: Pre-earthquake Evaluation 7-13

 Or, a jurisdiction could amend the model code by adding rows to the table to
identify mitigation priorities in terms of use or occupancy, risk category,
building height or size, age, structural characteristics, soil profile type, or even
neighborhood. For example, the “Soft story” and “Private school” rows in
Table 7-1 would complement two of San Francisco’s current mandatory
programs by adding a related alteration trigger. The “welded steel moment
frame” and “non-ductile concrete” rows show how the City might use this
format to address other structure types already identified in Workplan 2012-
2042. The final two rows of Table 7-1 show how the City might target certain
tall buildings (shown, arbitrarily, as 8 stories or 12 stories and taller) based on
concerns discussed elsewhere in this report, such as buildings clustered in
the dense central business district (downtown) or those supporting business
and commerce functions deemed recovery-critical.

None of the completed rows in Table 7-1 represents a specific policy

recommendation of this report. Rather, it is the format of the table that
appears useful as a way to balance more aggressive triggers with more
focused intent. The tabular format by itself would not change the triggers, or
the triggered scope (structural only), or the criteria currently in the SFEBC.
But if the targets of retrofit triggers can be more thoughtfully identified and
justified, then the triggers can be adjusted so that they will apply to more
projects and be more effective as risk reduction tools.

This approach can still face feasibility hurdles. Even in a seismically deficient
and recovery-critical building, if the triggered work is simply unaffordable, the
owner can always choose to scale back or cancel the intended project, and in
the end, the City gets neither risk reduction nor building renovation.

3.3 Aggressive Triggers with Reduced Scope

Another alternative is to make the current triggers more aggressive, so they

apply to a wide range of intended projects, but limit the triggered scope to
work that is less costly and less disruptive. Three ways to limit the triggered
retrofit scope are:
• Set the structural retrofit criteria low, so that all but the riskiest structures
or most egregious deficiencies are exempt. This approach makes sense
for building groups that would justify a mandatory program if they existed
in large numbers, like unreinforced masonry buildings. For tall buildings,
or commercial buildings generally, the criteria could be based on a
Collapse Prevention performance level or a hazard level lower than the
code’s already reduced seismic loads, or both.

7-14 Part 7: Pre-earthquake Evaluation ATC-119-1

• Set the structural retrofit criteria to allow partial retrofits of collapse-prone
deficiencies only. This approach makes sense for buildings with known
and easily identifiable weak links, such as “soft-story” apartments, cripple
wall houses, or obsolete tilt-up structures. For tall buildings, this
approach might be useful for concrete or steel braced frame structures
with discontinuous systems or for pre-Northridge welded steel moment
frames.
• Set the retrofit criteria to cover only critical nonstructural components.
This approach would not reduce collapse risk, but in many commercial
buildings, including tall buildings, the collapse risk in a design earthquake
is already relatively low compared with other vulnerable structure types.
Since the current SFEBC triggers focus exclusively on structural retrofit,
this approach offers an opportunity to shift focus to the nonstructural
systems that affect functional recovery and that tend to be damaged even
in smaller events. New triggers could address critical nonstructural
systems, perhaps including elevators, fire suppression systems, and
exterior cladding, leaving the current code provisions to cover structural
issues (in the rare cases in which those current triggers apply).

Another way to reduce the triggered scope is to focus on evaluation and

disclosure, as opposed to actual retrofit. City of Portland, for example,
supplements the Oregon building code with an alteration cost provision that
triggers only an evaluation (City of Portland, 2015, Section 24.85.060) as
follows :

When an alteration for which a building permit is required has a value

(not including costs of mechanical, electrical, plumbing, permanent
equipment, painting, fire extinguishing systems, site improvements,
eco-roofs and finish works) of more than $175,000, an ASCE 41
evaluation is required. This value of $175,000 shall be modified each
year after 2004 by the percent change in the R.S Means Construction
Index for Portland on file with the Director. A letter of intent to have
an ASCE 41 evaluation performed may be submitted along with the
permit application. The evaluation must be completed before any
future permits will be issued.

The current Los Angeles program for non-ductile concrete structures offers a
related precedent. The program mandates retrofit, but the retrofit deadline is
25 years after notification by the building department. In the near term, only
evaluation is required, so the program functions as a mandatory evaluation
only (City of Los Angeles, 2015, Chapter IX, Division 95).

ATC-119-1 Part 7: Pre-earthquake Evaluation 7-15

 In both Portland and Los Angeles, the intent is that evaluation findings will
prompt owners to plan and execute their own voluntary retrofits, or at least
take steps to reduce their risks. This is also the logic behind Workplan 2012-
2042 tasks that call for mandatory evaluation followed a few years later by
mandatory retrofit. Many believe that the production of a formal evaluation
report should motivate an owner by confirming liability for damage if a retrofit
is not performed (Bonowitz, 2018).

The evaluation reports produced by these mandates and triggers are

presumably public documents available to potential buyers, tenants, and the
general public. If San Francisco were to develop an evaluation trigger, it
could include a provision requiring disclosure to current and prospective
tenants. Public disclosure is the purpose of placarding programs used by
some jurisdictions for unreinforced masonry buildings and was a feature of a
pioneering risk reduction program in Palo Alto (City of Palo Alto, 1986,
Section 16.42.070).

Having and sharing more information regarding expected earthquake

performance is beneficial to stakeholders and to the City in the long run, but
the disclosure process between owners and tenants can be fraught,
especially if one or both parties is unfamiliar with the technical issues. To
help address these concerns, government agencies (including the State of
California), institutions, and private organizations have developed various
“rating systems” over the years. San Francisco has its own, a four-level
Seismic Hazard Rating used by the Department of Public Works (Lui, 2018).
For communicating with non-expert stakeholders, however, San Francisco
would benefit from a rating system that uses a more pragmatic, non-technical
terminology; separates safety, repair cost, and recovery time into separate
ratings; accommodates multiple evaluation methodologies commonly used by
engineers; and is available for free public use, allowing the City to specify it
without a Request for Proposal (RFP) process (SEAONC BRC, 2015).

Other recommendations presented in this report can also work with codified
triggers. San Francisco can write an alteration or change of use trigger to
require development of a recovery plan or participation in the City’s Building
Occupancy Resumption Program (BORP,discussed further in Part 6). Either
of these triggers would represent a recovery-focused extension of the
triggered seismic evaluation. At the very least, the City can use a code
trigger to require the building owner to file basic information to enhance the
building database described in Part 1 and to inform the City’s own response
and recovery plans as discussed in Part 6.

7-16 Part 7: Pre-earthquake Evaluation ATC-119-1

Well-intentioned retrofit triggers can lead to unintended consequences when
owners decide they would rather cancel a project than undertake the
additional triggered work. This is especially likely in tall buildings where the
triggered retrofit might be disproportionately costly and disruptive. A reduced
scope can represent a balance of the City’s and the owner’s interests.

3.4 Acquisition Triggers

Codified retrofit triggers are meant to take advantage of the opportunity

provided when an owner invests in a major alteration or change of
occupancy. The intended project proves the building is a viable investment,
and the work often exposes the structure while the building is at least partially
vacant. As discussed above, however, the current SFEBC triggers simply do
not catch many tall buildings. More buildings might be triggered for seismic
work if San Francisco could also take advantage of the opportunity presented
when a building is bought or when a major tenant starts or renews a lease.
Task C.1.a of Workplan 2012-2042, “mandatory evaluation on sale or by
deadline for all building types not otherwise covered,” already contemplates a
program of “mandatory evaluation on sale” for broad groups of San Francisco
buildings.

Acquisition (by lease or sale) is not a project type within the CEBC or SFEBC
for private buildings. Many government agencies and institutions, however,
have internal policies for “seismic due diligence” that could be adopted to
supplement the codes. Private lenders and insurers also rely on seismic
evaluations at the time of sale, though their interests are different from those
of the City as a regulator and recovery planner.

As with cost-based alteration triggers, acquisition triggers are most feasible

when the regulating agency also has a stake in the building as the owner or
lessee. Federal government agencies require a seismic evaluation when a
building is “added to the Federal inventory through purchase or donation”
(NIST, 2011, Section 2.1) and when an agency starts or renews a lease in a
non-Federally owned building (NIST, 2011, Section 1.3.2). The University of
California and California State University both require a demonstration of
acceptable performance (using the State’s seven-level rating system) for
most leased facilities and for any building acquired by purchase or title
transfer (University of California, 2017; CSU, 2016).

For private property, California law requires the disclosure of “earthquake

fault zones” and mapped landslide and liquefaction hazards when real
property is transferred (Civil Code Section 1103). These disclosures say
nothing about expected performance of any structures on the property,

ATC-119-1 Part 7: Pre-earthquake Evaluation 7-17

 however, and can thus be misleading. For the buildings themselves,
California also requires the seller of a house built before 1960 (wood-frame
construction, up to four units) to provide the buyer with a bare bones seismic
evaluation in the form of a checklist (Government Code Section 8897.1). The
program is widely seen as ineffective, however, since sellers are allowed
(even encouraged) to check the “Don’t Know” option, the report is not
required for multi-family buildings, and owners are not required to share the
information with tenants. In San Francisco, title transfer through
condominium conversion is specifically exempt from seismic review or
disclosure, even while a general building inspection and other corrective work
is required.

For commercial buildings, California law addresses the transfer of concrete

“tilt-up” and similar reinforced masonry buildings (typically only one or two
stories tall), requiring the seller only to provide a copy of a Seismic Safety
Commission document. No building-specific information is required, no
retrofit is required, and the statue applies only to pre-1975 buildings, even
though tilt-up design provisions continued to improve even after the 1994
Northridge earthquake. The law does provide an interesting negative
incentive, however: any buyer who fails to retrofit within three years of the
purchase “shall not receive payment from any state assistance program for
earthquake repairs resulting from damage during an earthquake until all other
applicants have been paid” (Government Code Section 8893).

For the residential and commercial properties addressed by California law,

compliance is the responsibility of the seller or the seller’s agent.
Presumably, the compliance status could be noted on a title report, but
enforcement by the building department is not expected. Thus, precedents
for effective acquisition triggers exist, but they are more in institutional policy
than in enforceable legislation or building code provisions.

If San Francisco were to create an acquisition trigger for private buildings, the
process would probably require legislation, as opposed to just an
administrative bulletin or building code amendment. In addition,
implementation would require new coordination between the Department of
Building Inspection, which would set criteria and approve the triggered
evaluation reports, and the Office of Assessor-Controller, which would have
to initiate the process based on recorded title transfers. It is not clear which
City department would be responsible for triggering evaluations based on
new leases. Possibly, as with the California statutes for residential and tilt-up
transfers, compliance would be left to the private stakeholders themselves,
and would be initiated by their realtors and attorneys without much City
oversight.

7-18 Part 7: Pre-earthquake Evaluation ATC-119-1

The legislation would need to develop consensus on a number of questions.
The legislative process will need to address the following issues, among
others:
• Which buildings should be subject to the trigger? This study is about tall
buildings, but if the City goes through the effort to create an acquisition
trigger, it might consider a broader application. If a broad scope is
politically infeasible, some of the scope narrowing and target focusing
ideas in previous sections of this report might be helpful. Should
relatively new, or recently retrofitted buildings be exempted or
benchmarked?
• How should the sale trigger be set? For a residential building, should the
conversion to condominiums count? Should the sale of a single
condominium pull the trigger? Partial sales of large office buildings have
become more common throughout the country, largely to avoid transfer
taxes and, in California, property tax reassessment that applies when at
least half the building is sold (Li, 2017; Morris, 2018). In 2016, San
Francisco raised transfer taxes on sales over five million dollars, further
encouraging partial sales, including several in well-known tall buildings
(Littman, 2018; Li, 2018). An effective acquisition trigger would therefore
need to be set below 49 percent, but should also be adjustable to account
for future trends in the real estate market and likely adjustments to current
tax law.
• How should the lease trigger be set? Leases are very common, so an
effective acquisition trigger should apply to at least some of them as well.
That said, many new leases come with tenant improvements and
alterations, so this acquisition trigger will overlap to a degree with any
revised alteration triggers in the SFEBC.
• What scope and criteria should be used for the triggered evaluation?
Should it include nonstructural components and systems? Should it
consider only safety, or should it consider prospects for reoccupancy and
recovery?
• Who will be allowed to produce a triggered evaluation? Since there will
be no design or construction in an acquisition-triggered seismic
evaluation, will a licensed engineer be required? Should this depend on
the scope and criteria selected? California’s Board for Professional
Engineers, Land Surveyors, and Geologists has generally not considered
the production of seismic evaluations as within the practice of civil or
structural engineering. For a tall building, even the design requires only a
Civil Engineer, not a Structural Engineer, though SEAOC has proposed

ATC-119-1 Part 7: Pre-earthquake Evaluation 7-19

 legislation to require an SE for the design of buildings taller than just 45
feet (Schinske, 2015).
• What role should the Department of Building Inspection play? Should
acquisition-triggered evaluations be subject to approval by DBI? If the
requirement is in statue, but not in the building code, DBI’s regulatory
authority is less clear. A City approval will probably improve quality and
compliance rates, but it will also increase costs. Owners will probably
prefer the approach of the Government Code sections cited above, which
leave compliance largely to the parties and their selected consultants.
• Should the triggered evaluations be public? As the fruit of a City policy,
the reports would presumably be filed with the City and available for
public review. Alternatively, if the ultimate purpose is only to raise the
awareness of owners and tenants, leaving any actual risk reduction as
voluntary, an argument might be made for requiring disclosure to current
and prospective tenants but restricting access to the general public, as is
the case for reports required by the Government Code sections cited
above. See also the discussion of disclosure and communication issues
in Section 2.2 of this Part.

3.4.1 The PML Market as an Alternative to Public Policy

Another alternative is to leverage an existing practice from the private sector.

When a building is sold, the standard practice of most lenders is to call for an
analysis of its “probable maximum loss,” or PML. If the PML is too high, the
lender will either decline the loan or insist that the buyer purchase earthquake
insurance as a hedge against the lender’s risk. PML analyses are common
for commercial buildings but have been provided for some multi-family
residential buildings as well. Any San Francisco tall building being sold in the
near future would be expected to have a PML analysis performed as part of
the transaction.

If the City could compel building owners to submit or make public their PML
reports, or perhaps just to disclose them to tenants, that might meet some of
the goals of a legislated acquisition trigger. Owners might even prefer this
approach if it would mean they can preserve some confidentiality of the
findings and keep the transaction process free of City approvals.

But there are also reasons why current PML practices might not serve the
City’s interests, at least not as a replacement for broader public policy.
These include:
• Two ASTM standards (2007a; 2007b) define terminology and procedures
but the practice is largely unregulated. Because PML analyses are

7-20 Part 7: Pre-earthquake Evaluation ATC-119-1

 produced to support private transactions, they do not receive peer review
or building department approval. Even if the parties agree to abide by
them, they do not provide quality assurance. Since no engineering
design is involved, California licensure rules do not apply in a reliable
way. As a result, PML evaluation scopes, methods, and conclusions are
frequently inconsistent between evaluators (Searer et al., 2009).
• PML findings are used primarily to support financial decisions, which
differ in their nature and intent from policy decisions focused on public
safety, functional recovery, or community resilience. Where the financial
decision is tied to a fixed PML value, reported results tend to cluster
around that value, essentially saying “above” or “below” without providing
more useful detail. Indeed, typical PML practice has become so rote that
many do not even consider it especially valuable for its own stated
purpose of properly valuing the risk (Porter et al., 2004).
• PML reports are traditionally considered private and confidential, so much
so that owners have been known to commission the evaluations through
their lawyers to preserve attorney client privilege and avoid disclosing the
findings.
• PML analyses are not provided for lease transactions.

Thus, even if San Francisco would want to take advantage of the analyses
already being performed in the PML market, it would have to establish its own
standards that might interfere with current private sector practices.

3.5 Incentives for Voluntary Evaluation or Retrofit

Seismic risk reduction can be classified as mandatory, triggered, or voluntary.

Practically all current work on San Francisco’s tall buildings falls into the latter
category. Voluntary work can sometimes be incentivized, but experience has
shown that substantial benefits need to be provided to spur actual seismic
retrofit. Some incentives that already exist—a waiver on property tax
assessment (Revenue and Taxation Code Section 74.5), the 2017 CalCAP
loan guarantee program (CPCFA, 2017), and even the growing availability of
PACE financing—have had very little impact, especially on tall buildings.
Thus, feasible incentives for seismic improvements to San Francisco tall
building safety or recovery are likely to be less effective than focused
mandates or triggers.

That said, dozens of San Francisco tall buildings already participate in the
City’s post-earthquake reoccupancy program, BORP, discussed in detail in
Part 6. BORP can be considered an incentivized voluntary program that
benefits both the City and the building’s owner and tenants. In exchange for

ATC-119-1 Part 7: Pre-earthquake Evaluation 7-21

 City approval of a privatized and expedited safety inspection, a building
owner agrees to develop a reoccupancy plan and often commissions a
seismic evaluation as part of the process.

7-22 Part 7: Pre-earthquake Evaluation ATC-119-1

Chapter 4

Issues and Recommendations

4.1 Revisions and Amendments to the SFEBC

San Francisco supplements the CEBC with its own provisions that trigger
seismic upgrade when a major alteration or a change of occupancy is
proposed. Because of their size and typical uses, however, tall buildings are
almost never affected by these San Francisco amendments. Therefore, even
the most collapse-prone tall buildings only rarely receive the scrutiny intended
by the code. That said, since the retrofit of an occupied tall building is
especially expensive and disruptive, a more aggressive trigger provision
could discourage modernization or tenant improvement. Thus, the San
Francisco Existing Building Code’s generic provisions are problematic for tall
and similarly large or complex facilities. Still, these provisions can be
effective if they are made more aggressive, so that they apply to more
alteration and change of occupancy projects, but if they are also made more
focused, so they address the conditions of greatest concern in ways that
building owners can afford.

Recommendation: Amend the San Francisco Existing Building Code

provisions for alteration and change of occupancy projects.
• Lower the trigger levels, while also exempting buildings that do not
present well-recognized safety or recovery risks. Identify these subject
buildings in advance as particular combinations of occupancy or use,
building size, structural attributes, building age, and other critical
conditions.
• Lower the trigger levels, but for tall buildings, consider limiting the scope
of triggered work to allow partial retrofits, to focus on nonstructural
systems only, to require evaluation and disclosure only, or to require
participation in BORP or the filing of a recovery plan.
• Amend the evaluation criteria in SFEBC Section 301.1.4.2, and develop
an Administrative Bulletin as needed, to facilitate the use of FEMA 351,
Recommended Seismic Evaluation and Upgrade Criteria for Existing
Welded Steel Moment-Frame Buildings (FEMA, 2000), for eligible
structures. The findings of the SEAONC EBC Steel Frame Subcommittee
(2004) might be valuable in reviewing the options provided by FEMA 351.

ATC-119-1 Part 7: Pre-earthquake Evaluation 7-23

 • Update and amend the SFEBC to coordinate San Francisco amendments
with existing and forthcoming CEBC provisions.
o Revise the text that incorrectly says Section 301.2.1 applies only
when invoked by Section 403.
o Since Section 301.2.1 applies only to cases where reduced forces are
used (alterations and repairs), consider combining with Section
301.1.4.2 (to be Section 303.3.2 in the 2019 SFEBC).
o Delete the unnecessary pointer from Section 301.2.3 to Section
301.1.4.2. Pointers to Section 301.2.3 should instead point directly to
Section 301.1.4.2.
o Correct the reference in the first sentence of Section 403.4. It should
refer to Section 403.9 regarding voluntary seismic improvements, not
Section 403.5.
o Correct the numbering of Sections 403.12.1 and 403.12.2.
o Coordinate Sections 403.12.1 and 403.12.2 with new provisions on
similar topics coming in the 2019 CEBC. In particular, confirm the
intended use of the term “substantial structural alteration,” which will
have a different meaning in the 2019 CEBC.
• Clarify or revise limits on upgrade triggers imposed by the Department of
Housing and Community Development (HCD).

4.2 New Legislation

To supplement the SFEBC’s rarely applied alteration and change of

occupancy triggers, the City could take advantage of opportunities that arise
when a tall building is sold or a sizable portion of it is leased to a new tenant,
as contemplated by the CAPSS Earthquake Safety Implementation Program
Workplan 2012-2042 (CCSF, 2011). Even if only a private evaluation and
disclosure is triggered, that would at least ensure that buildings are properly
valued with respect to the risks they pose to owners and tenants. Because
acquisition is not a project type covered by the CEBC or SFEBC, and
because only evaluation and disclosure would be triggered, an acquisition-
based program would likely require legislation, and implementation would
likely require coordination with the Office of Assessor-Recorder.

Recommendation: Develop a program for evaluation and disclosure

triggered by sale or lease.

7-24 Part 7: Pre-earthquake Evaluation ATC-119-1

• With a working group or advisory panel, develop recommendations for
applicability and triggering conditions considering changing trends in real
estate practice and tax law.
• Develop recommendations for the technical scope and criteria for the
triggered evaluation. For purposes of resilience and recovery planning,
the triggered evaluation might be required to include an estimate of
recovery time as well as safety.
• Develop recommendations regarding City approval, disclosure, and public
posting of evaluation findings.

ATC-119-1 Part 7: Pre-earthquake Evaluation 7-25

Appendix A

Annotated Building Code

Provisions

This appendix presents the current building code provisions applicable to

triggered pre-earthquake seismic evaluation and retrofit. The current, 2016,
San Francisco Existing Building Code (SFEBC) amends the 2016 California
Existing Building Code (CEBC), which itself adopts and amends the 2015
International Existing Building Code (IEBC). This appendix shows the
relevant provisions from all three documents. In addition, it provides a
commentary on the combined provisions and their application to existing tall
buildings, and it offers notes on the forthcoming 2019 CEBC, which will adopt
the 2018 IEBC.

This appendix focuses on earthquake evaluation and retrofit provisions that

apply to pre-earthquake assessment of tall buildings associated with
alteration or change of occupancy projects. As such, it does not include all
existing building provisions or all tall building provisions. In particular, the
following specialty provisions might apply to certain tall buildings but are
omitted from this appendix for brevity:
• Historical Building Code provisions
• Provisions for state-owned buildings in CEBC Sections 317-322; Section
317.1.1 allows these sections to be adopted by local jurisdictions for use
on private buildings
• Regulations used by OSHPD and DSA
• San Francisco amendments regarding buildings “on a military base
selected for closure” (SFEBC Section 326) and homeless shelters
(SFEBC Section 401.7), which are not likely to affect tall buildings

California amendments from the Department of Housing and Community

Development (HCD) are included, since they are meant to cover apartment
buildings, condominiums, and hotels that might be in tall buildings.

The text is formatted to distinguish between the various source documents

using the following legend:

ATC-119-1 Part 7: Pre-earthquake Evaluation 7-27

 2015 IEBC. Model code. Plain text (except for section headings in bold).
2016 CEBC. California amendments, applicable in San Francisco. Italic
plain text (except for section headings in bold).
2016 SFEBC. Additional San Francisco amendments. Bold italic
text.
ATC Commentary. Plain italic text, indented.
2018 IEBC. Model code language already published and expected to
be adopted into the 2019 CEBC and SFEBC. Plain text, shaded and
indented.

CHAPTER 1
SCOPE AND ADMINISTRATION

1.8 Department of Housing and Community Development

1.8.10 Other Building Regulations

1.8.10.1 Existing structures. Notwithstanding other provisions of law, the

replacement, retention, and extension of original materials and the use of
original methods of construction for any existing building or accessory
structure, or portions thereof, shall be permitted in accordance with the
provisions of this code as adopted by the Department of Housing and
Community Development. For additional information, see California Health
and Safety Code, Sections 17912, 17920.3, 17922 and 17958.8.

Commentary: This and similar California amendments for HCD are

generally interpreted to prohibit upgrade triggers for residential
buildings. See the commentary at Section 401.2.1.

CHAPTER 2
DEFINITIONS

ADDITION. An extension or increase in floor area, number of stories, or

height of a building or structure.

ALTERATION. Any construction or renovation to an existing structure other

than a repair or addition. Alterations are classified as Level 1, Level 2 and
Level 3.

7-28 Part 7: Pre-earthquake Evaluation ATC-119-1

 Commentary: The 2018 IEBC will change this definition by removing
the second sentence about alteration levels. This change has no
substantive effect.
The 2018 IEBC (and presumably the 2019 CEBC) will also have a
new definition of Substantial Structural Alteration, a term that was
previously used only in the IEBC’s Work Area method but will, in the
new edition, be used also in the Prescriptive method adopted by
California. See the discussion of 2018 IEBC Section 503.11 in the
commentary to 2016 SFEBC Section 403.12.2. The new definition
will read as follows:
SUBSTANTIAL STRUCTURAL ALTERATION. An alteration in
which the gravity load-carrying structural elements altered within a 5-
year period support more than 30 percent of the total floor and roof
area of the building or structure. The areas to be counted toward the
30 percent shall include mezzanines, penthouses, and in-filled courts
and shafts tributary to the altered structural elements.

CHANGE OF OCCUPANCY. A change in the use of the building or a portion

of a building. A change of occupancy shall include any change of occupancy
classification, any change from one group to another group within an
occupancy classification or any change in use within a group for a specific
occupancy classification.

Commentary: The 2018 IEBC will revise this definition, but the change
should have no substantive effect.

SEISMIC REHABILITATION. Work conducted to improve the seismic lateral

force resistance of an existing building.

Commentary: 2018 IEBC will remove this definition, which is neither

used nor needed in the model code.

WORK AREA. That portion or portions of a building consisting of all

reconfigured spaces as indicated on the construction documents. Work area
excludes other portions of the building where incidental work entailed by the
intended work must be performed and portions of the building where work not
initially intended by the owner is specifically required by this code.

Commentary: This definition is not used by the 2016 CEBC. It will be

used by the 2018 IEBC (and presumably the 2019 CEBC). See the

ATC-119-1 Part 7: Pre-earthquake Evaluation 7-29

 discussion of 2018 IEBC Section 503.11 in the commentary to 2016
SFEBC Section 403.12.2.

CHAPTER 3
PROVISIONS FOR ALL COMPLIANCE METHODS

301.1 General. The repair, alteration, change of occupancy, addition or

relocation of all existing buildings shall comply with one of the methods listed
in Sections 301.1.1 through 301.1.3 as selected by the applicant. Sections
301.1.1 through 301.3.3 shall not be applied in combination with each other.
Where this code requires consideration of the seismic force-resisting system
of an existing building subject to repair, alteration, change of occupancy,
addition or relocation of existing buildings, the seismic evaluation or design
shall be based on Section 301.1.4 regardless of which compliance method is
used.

Exception: Subject to the approval of the code official, alterations complying

with the laws in existence at the time the building or the affected portion of
the building was built shall be considered in compliance with the provisions of
this code unless the building is undergoing more than a limited structural
alteration as defined in Section 907.4.4. New structural members added as
part of the alteration shall comply with the California Building Code.
Alterations of existing buildings in floor hazard areas shall comply with
Section 701.3.

Commentary: The 2016 CEBC references to different “methods” and

to Sections 701 and 907, though unchanged relative to the 2015 IEBC
model code, are inappropriate, as the CEBC uses only the
Prescriptive Method of 2016 CEBC Chapter 4.
The 2018 IEBC will change this general scoping provision to account
for new chapter organization and section numbering. It will also
replace the exception’s reference to “limited structural alteration” with
the following:
This exception shall not apply to the structural provisions of Chapter 5
or to the structural provisions of Sections 706, 806 and 906.

With this change, the 2018 IEBC (and presumably, the 2019 CEBC
and SFEBC) remove a discretionary waiver that might have applied to
typical alteration projects. The code official will retain discretion to
waive alteration-triggered requirements related to heights and areas,

7-30 Part 7: Pre-earthquake Evaluation ATC-119-1

 egress, etc., but will no longer be able to waive the normal structural
triggers.
301.1.4 Seismic evaluation and design procedures. The seismic
evaluation and design shall be based on the procedures specified in the
California Building Code or ASCE 41. The procedures contained in Appendix
A of this code shall be permitted to be used as specified in Section 301.1.4.2.

Commentary: In the 2018 IEBC, Section 301.1.4 has been edited for
clarity and to coordinate with ASCE 41-17 instead of ASCE 41-13.
The entire section has also been relocated and renumbered as
Section 303.3.
The 2016 CEBC adopts only three of the 2015 IEBC’s five Appendix A
chapters. None of the three adopted chapters applies to tall buildings
(or to steel or concrete structural systems of any size).

301.1.4.1 Compliance with International Building Code-level seismic

forces. Where compliance with the seismic design provisions of the
California Building Code is required, the criteria shall be in accordance with
one of the following:

1. One-hundred percent of the values in the California Building Code. Where

the existing seismic force-resisting system is a type that can be designated
as “Ordinary,” values of R, Ω0 and Cd used for analysis in accordance with
Chapter 16 of the California Building Code shall be those specified for
structural systems classified as “Ordinary” in accordance with Table 12.2-1 of
ASCE 7, unless it can be demonstrated that the structural system will provide
performance equivalent to that of a “Detailed,”
“Intermediate” or “Special” system.

2. ASCE 41, using a Tier 3 procedure and the two-level performance

objective in Table 301.1.4.1 for the
applicable risk category.

Commentary: Section 301.1.4.1 gives the “full” earthquake design

criteria applicable when seismic work is triggered by a change of
occupancy (See Section 407). The criteria vary by risk category: For
RC I or II: Structural Life Safety with a BSE-1N hazard and Collapse
Prevention with a BSE-2N hazard; for RC III: Structural Damage
Control with a BSE-1N hazard and Structural Limited Safety with a
BSE-2N hazard. Tall buildings would not be expected to be assigned
to Risk Category IV.

ATC-119-1 Part 7: Pre-earthquake Evaluation 7-31

 301.1.4.2 Compliance with reduced International Building Code-level
seismic forces. Where seismic evaluation and design is permitted to meet
reduced California Building Code seismic force levels, the criteria used shall
be in accordance with one of the following:

1. The California Building Code using 75 percent of the prescribed forces.

Values of R, Ω0, and Cd used for analysis shall be as specified in Section
301.1.4.1 of this code.

2. Except where these requirements are triggered by Section 403.12,

structures or portions of structures that comply with the requirements
of the applicable chapter in Appendix A as specified in Items 2.1
through 2.5 and subject to the limitations of the respective Appendix A
chapters shall be deemed to comply with this section.
...
2.5. Seismic evaluation and design of concrete buildings assigned to Risk
Category I, II or III are permitted to be based on the procedures specified in
Chapter A5.

3. ASCE 41, using the performance objective in Table 301.1.4.2 for the
applicable risk category.

Commentary: Section 301.1.4.2 gives the “reduced” earthquake

design criteria applicable when seismic work is triggered by certain
alteration projects. See Section 403.
The 2016 CEBC reference to Chapter A5 is inappropriate, as
California does not adopt that chapter of the model code. Also, the
2018 IEBC (and presumably the 2019 CEBC) will no longer have
Chapter A5; it was omitted because it is essentially identical to ASCE
41. As noted above, none of the other Appendix A chapters adopted
by the CEBC is applicable to tall buildings or to concrete or steel
structural systems of any size.
Section 301.1.4.2 is referenced by SFEBC Section 301.2, which
contains San Francisco’s traditional reduced load criteria with
benchmark dates. The CEBC allows three options where reduced
loads are allowed. Presumably, all three options are acceptable
within the intent of SFEBC Section 301.2.
• Option 1 is the traditional “75 percent” approach long embraced by
San Francisco.

7-32 Part 7: Pre-earthquake Evaluation ATC-119-1

 • Option 2 allows five prescriptive approaches for specific building
types. As noted above, however, California adopts only three of
the five Appendix A chapters, none of which is applicable to tall
buildings, and Chapter A5 (not adopted by the CEBC) has already
been removed from the 2018 IEBC. Note also that SFEBC
modifies Option 2 with an introductory phrase prohibiting its use
on retrofits triggered by major alterations (SFEBC Section
403.12).
• Option 3 allows the use of ASCE 41-13 with performance
objectives that vary by risk category: For RC I or II: Structural Life
Safety with a BSE-1E hazard; for RC III: Structural Damage
Control with a BSE-1E hazard. Tall buildings would not be
expected to be assigned to Risk Category IV.
In addition to ASCE 41, San Francisco should consider allowing the
use of FEMA 351, Recommended Seismic Evaluation and Upgrade
Criteria for Existing Welded Steel Moment-Frame Buildings, for
eligible structures (FEMA, 2000). An Administrative Bulletin to guide
its use will probably be needed.

301.2 Minimum Lateral Force for Existing Buildings

301.2.1 General. This section is applicable to existing buildings when

invoked by SFEBC Section 403. This section may be used as a
standard for voluntary upgrades.

Commentary: Section 301.2 gives additional criteria for seismic work

triggered by alteration projects in Section 403. Despite the wording of
Section 301.2.1, these provisions are also invoked for change of
occupancy projects by SFEBC Section 407.4.1.

An existing building or structure which has been brought into

compliance with the lateral force resistance requirements of the San
Francisco Building Code in effect on or after the dates shown in Table
301.2.1 [Table 7-2 below], shall be deemed to comply with this section
except when a vertical extension or other alterations are to be made
which would increase the mass or reduce the seismic resistance
capacity of the building or structure. Where multiple building types
apply, the later applicable date shall be used. Where none of the
building types apply, compliance shall be at the discretion of the
Director. Building type definitions are given in ASCE 41-13, Table 3-1.

ATC-119-1 Part 7: Pre-earthquake Evaluation 7-33

 Commentary: Starting with the 2016 SFEBC, this section includes
benchmark dates for each structural system (as opposed to the prior
provision, which gave a single date for all buildings). Where the
benchmark table reprinted in Table 7-2 applies, it effectively
eliminates the requirement for alteration-triggered seismic evaluation
(and possible retrofit) in more recent buildings, some as old as 1995.
However, the waiver for recent structures does not apply to alterations
where the intended work “would increase the mass or reduce the
seismic resistance capacity.” This exception distinguishes seismic
work triggered by structurally significant alterations (already covered
by IEBC/CEBC Section 403.4 and SFEBC Section 403.12.2) from
architectural alterations covered by San Francisco’s own trigger in
SFEBC Section 403.12.1.

Table 7-2 Dates Required to Demonstrate Building Compliance (reproduced

from Table 301.2.1 in Section 301.2 of SFEBC)
Date of Model Code
Building Type Compliance (for reference)
Wood frame, wood shear panels (Types W1 & W2) 1/1/1984 UBC 1976
Wood frame, wood shear panels (Type W1A) 7/1/1999 UBC 1997
Floor areas greater than 3,000 ft2 per level
Steel moment-resisting frame (Types S1 & S1a) 12/28/1995 UBC 1994
Steel concentrically braced frame (Types S2 & S2a) 7/1/1999 UBC 1997
Steel eccentrically braced frame (Types S2 & S2a) 1/1/1990 UBC 1988
Buckling-restrained braced frame (Types S2 &S2a) 1/1/2008 IBC 2006
Light metal frame (Type S3) 1/1/2008 IBC 2006
Steel frame w/ concrete shear walls (Type S4) 12/28/1995 UBC 1994
Steel plate shear wall (Type S6) 1/1/2008 IBC 2006
Reinforced concrete moment-resisting frame (Type C1) 12/28/1995 UBC 1994
Reinforced concrete shear walls (Types C2 & C2a) 12/28/1995 UBC 1994
Tilt-up concrete (Types PC1 & PC1a) 7/1/1999 UBC 1997
Precast concrete frame (Types PC2 & PC2a) 1/1/2008 IBC 2006
Reinforced masonry (Type RM1) Flexible diaphragms 7/1/1999 UBC 1997
Reinforced masonry (Type RM2) Stiff diaphragms 12/28/1995 UBC 1994
Seismic isolation or passive dissipation 7/1/1992 UBC 1991

301.2.2 Wind forces. Buildings and structures shall be capable of

resisting wind forces as prescribed in San Francisco Building Code
Section 1609.

7-34 Part 7: Pre-earthquake Evaluation ATC-119-1

301.2.3 Seismic forces. Buildings and structures shall comply with the
reduced International Building Code-level seismic forces, as defined in
Section 301.1.4.2. The building separation limitations of Section ASCE
7-10 Section 12.12.3 need not be considered.

When upper floors are exempted from compliance by Section 403.12.2,

the lateral forces generated by their masses shall be included in the
analysis and design of the lateral force resisting systems for the
strengthened floor. Such forces may be applied to the floor level
immediately above the topmost strengthened floor and distributed in
that floor in a manner consistent with the construction and layout of the
exempted floor.

Commentary: Section 301.2.3 gives the “reduced” earthquake design

criteria for use when a seismic upgrade is triggered by certain
alteration or repair projects. The first part of this provision is not
actually needed; as shown, it merely points to criteria already
provided in CEBC Section 301.1.4.2. This sentence remains,
however, as a vestige of prior SFEBC provisions for San Francisco’s
own alteration triggers (SFEBC Sections 403.12.1 and 403.12.2) –
triggers the IEBC/CEBC does not have.
In addition to setting the force level, Section 301.2 provides
benchmarking (Section 301.2.1) and waives building separation
requirements (Section 301.2.3); these provisions are unique and
traditional to San Francisco. As such, they are perhaps meant to
apply only to work triggered by San Francisco amendments, but the
way the code is written, they would apply to any project subject to a
code section that invokes Section 301.2.

CHAPTER 4
PRESCRIPTIVE COMPLIANCE METHOD

401.1.2 Existing state-owned structures. [BSC] ... The provisions of

Sections 317 through 322 may be adopted by a local jurisdiction for
earthquake evaluation and design for retrofit of existing buildings.
...
401.2.1 Existing materials. ... [HCD 1] Local ordinances or regulations
shall permit the replacement, retention and extension of original materials,
and the use of original methods of construction [if the] structure complied with
the building code provisions in effect at the time of original construction and
... does not become or continue to be a substandard building. ...

ATC-119-1 Part 7: Pre-earthquake Evaluation 7-35

 Commentary: This and similar California amendments for HCD are
generally interpreted to prohibit upgrade triggers for residential
buildings. However, the final phrase of the provision, regarding
“substandard building,” alters that general interpretation with respect
to earthquake design. “Substandard building” is defined in Health and
Safety Code Section 17920.3 to include any residential building with
“inadequate structural resistance to horizontal forces.”
Model “soft story” provisions proactively classify targeted buildings as
substandard to avoid conflict over the HCD exception. SFEBC could
do the same for other buildings, including tall residential buildings, that
fail a triggered evaluation. This would be a helpful policy clarification
as well as a reasonable interpretation.

SECTION 403
ALTERATIONS

403.1 General. Except as provided by Section 401.2 or this section,

alterations to any building or structure shall comply with the requirements of
the California Building Code or California Residential Code, as applicable, for
new construction. Alterations shall be such that the existing building or
structure is no less conforming to the provisions of the California Building
Code or California Residential Code, as applicable, than the existing building
or structure was prior to the alteration. ...

403.1.1 Replacement, retention and extension of original materials.

[HCD 1] Local ordinances or regulations shall permit the replacement,
retention and extension of original materials, and the use of original methods
of construction [if the] structure complied with the building code provisions in
effect at the time of original construction and ... does not become or continue
to be a substandard building. ...

Commentary: Section 401.2 allows existing materials to remain unless

dangerous. The HCD 1 provision added as Section 403.1.1 is
consistent with the amendment to 401.2.1. See the commentary
there.
403.4 Existing structural elements carrying lateral load. Except as
permitted by Section 403.5, where the alteration increases design lateral
loads in accordance with Section 1609 or 1613 of the California Building
Code, or where the alteration results in a prohibited structural irregularity as
defined in ASCE 7, or where the alteration decreases the capacity of any

7-36 Part 7: Pre-earthquake Evaluation ATC-119-1

existing lateral load-carrying structural element, the structure of the altered
building or structure shall be shown to meet the requirements of Sections
1609 and 1613 of the California Building Code. For purposes of this section,
compliance with ASCE 41, using a Tier 3 procedure and the two-level
performance objective in Table 301.1.4.1 for the applicable risk category,
shall be deemed to meet the requirements of Section 1613 of the California
Building Code.

Exception: Any existing lateral load-carrying structural element whose

demand-capacity ratio with the alteration considered is no more than 10
percent greater than its demand-capacity ratio with the alteration ignored
shall be permitted to remain unaltered. For purposes of calculating demand-
capacity ratios, the demand shall consider applicable load combinations with
design lateral loads or forces in accordance with Sections 1609 and 1613 of
the California Building Code. For purposes of this exception, comparisons of
demand-capacity ratios and calculation of design lateral loads, forces and
capacities shall account for the cumulative effects of additions and alterations
since original construction.

Commentary: This is the key provision for alteration-triggered seismic

evaluation and upgrade. It sets three triggers by which the intended
alteration is to be measured
• Increase in lateral load, perhaps due to added mass,
• Creation of a prohibited irregularity, perhaps due to added mass
or alteration of diaphragms, and
• Decrease in lateral capacity, perhaps due to removal of columns
or wall elements.
If a trigger is pulled, the triggered scope involves the entire structure,
considering both wind and seismic loads. The exception typically
reduces the triggered scope; if the intended alteration (a tenant
improvement or mechanical upgrade, for example) does not touch the
structure and adds little mass, the scope will likely be reduced to the
point where no seismic work is required at all. In these common
cases, it can be quickly shown that the change in demand-capacity
ratio for any element of the lateral force-resisting system will not be
close to 10 percent. Thus, the addition of mass triggers a review of
the entire structure, but the exception then exempts every element,
one by one.
The intent of this trigger, and its exception, is not to identify
seismically deficient buildings and force them to be retrofitted.
Rather, it is to identify alteration projects (or the accumulation of

ATC-119-1 Part 7: Pre-earthquake Evaluation 7-37

 alteration projects, as noted in the last sentence of the exception) that
make a significant change to the building’s expected seismic
performance. By focusing on the effect of the intended alteration –
not on the adequacy of the pre-alteration structure – the provision
attempts to strike a balance between encouraging modernization and
adaptive reuse of existing buildings and reduction of unacceptable
risks. The result, well understood by the code writers, is that a highly
deficient or even a collapse prone building would not trigger a seismic
evaluation or retrofit if the intended alteration is small.
This basic trigger is supplemented in the SFEBC by a San Francisco
provision for explicitly structural alterations to which the “10 percent”
exception does not apply. See Section 403.12.2.
For any part of the structure not exempt by the exception, the
provision sets the criteria by reference to Table 301.1.4.1, requiring
“full” seismic loads. SFEBC Section 301.2, which allows reduced
loads and benchmarking waivers, does not apply to work triggered by
Section 403.4 because of the exception in Section 301.2 for triggers
based on added mass or reduced lateral capacity. However, the San
Francisco building official should still be able to allow the use of
reduced seismic loads by discretion, referencing any of three
alternative standards:
• In the 2018 IEBC, the section corresponding to 2016 CEBC
Section 403.4 has been changed to allow reduced seismic loads
for alteration-triggered seismic work.
• The IEBC model code on which the CEBC is based has three
methods. For simplicity and consistency with past practice,
California adopts only the Prescriptive method, but the IEBC’s
Work Area method allows the use of reduced seismic loads for
alteration-triggered seismic work. The previous edition of the
CEBC included a direct reference to the IEBC as an approved
alternative, effectively allowing the use of reduced seismic loads.
The 2016 CEBC does not have that explicit allowance of the Work
Area method, but the Building Standards Commission has stated
that in shifting from CBC Chapter 34 to the CEBC there was no
intent to change the substance of any requirement or allowance.
Therefore, the criteria of the IEBC Work Area method can be
reasonably cited as acceptable alternative criteria here.
• CEBC/SFEBC Section 401.1.2 allows the jurisdiction to adopt the
provisions for state-owned buildings in CEBC Sections 317 – 322.

7-38 Part 7: Pre-earthquake Evaluation ATC-119-1

 Those provisions allow the use of reduced seismic loads for
alteration-triggered seismic work.
The first sentence of CEBC Section 403.4 contains an error. It should
refer to Section 403.9, which covers voluntary seismic improvements,
not Section 403.5.
2016 CEBC Sections 403.4.2, 403.5, 403.6, and 403.7 trigger seismic
evaluation or retrofit for Risk Category IV facilities, unreinforced
masonry parapets, and unreinforced masonry walls in cases of certain
alterations. These provisions are not shown here because they are
unlikely to apply to tall buildings.

403.4.1 Existing structural elements carrying lateral load. [HCD] ...

Commentary: Section 403.4.1, for residential buildings regulated by

HCD, is the same provision and exception as Section 403.4, but
without the reference to Section 403.5 (perhaps because HCD did not
recognize the section numbering erratum) and without the allowance
to use ASCE 41.

403.9 Voluntary seismic improvements. Alterations to existing structural

elements or additions of new structural elements that are not otherwise
required by this chapter and are initiated for the purpose of improving the
performance of the seismic force-resisting system of an existing structure or
the performance of seismic bracing or anchorage of existing nonstructural
elements shall be permitted, provided that an engineering analysis is
submitted demonstrating the following:

1. The altered structure and the altered nonstructural elements are no less
conforming to the provisions of the California Building Code with respect to
earthquake design than they were prior to the alteration.

2. New structural elements are detailed as required for new construction.

3. New or relocated nonstructural elements are detailed and connected to
existing or new structural elements as required for new construction.

4. The alterations do not create a structural irregularity as defined in ASCE 7

or make an existing structural irregularity more severe.

Commentary: This provision has been significantly revised for the

2018 IEBC, but its basic intent is unchanged. As noted in the first
sentence of Section 403.4 (which should reference Section 403.9, not

ATC-119-1 Part 7: Pre-earthquake Evaluation 7-39

 Section 403.5), the normal triggers and the “10 percent” exception do
not apply to voluntary work done exclusively to improve seismic
performance.

403.12.1 Non-structural alterations. Whenever alteration work in a

building or structure involves substantial changes to elements such as
walls, partitions or ceilings, on 2/3 or more of the number of stories
excluding basements, the building or structure as a whole shall comply
with Section 301.2. The term “substantial change” includes the
addition, removal, repair or modification of such elements. All such
work included in alteration permits issued within two years of the date
of a permit application shall be included in the determination of whether
the application is proposing substantial change to the building or
structure. ...

Commentary: This San Francisco provision supplements the basic

alteration trigger of Section 403.4. (The SFEBC section numbering is
inappropriate, because CEBC Section 403.12 addresses an unrelated
topic, but this does not affect the substance of the provision.)
Unlike that basic provision, Section 403.12.1 has no “10 percent”
exception; once the trigger is pulled based on the extent of non-
structural work done over any two-year period, a seismic evaluation
and a retrofit of any elements found deficient must be performed. The
evaluation and retrofit criteria are the San Francisco criteria in Section
301.2, including the waivers for benchmarked buildings.
In practice, there has always been some code official judgment
involved in the counting of altered spaces toward the “2/3” trigger.
More significant for tall buildings, however, is the recognition that in an
occupied building, two-thirds of the stories will almost never be altered
by typical upgrades and tenant improvements within the course of two
years.

403.12.2 Structural alterations. When more than 30 percent of the floor

and roof areas of the building or structure have been or are proposed to
be involved in substantial structural alteration, the building or structure
shall comply with Section 301. The areas to be counted towards the 30
percent shall be those areas tributary to the vertical load carrying
components (joists, beams, columns, walls and other structural
components) that have been or will be removed, added or altered, as
well as areas such as mezzanines, penthouses, roof structures and
infilled courts and shafts.

7-40 Part 7: Pre-earthquake Evaluation ATC-119-1

Exceptions:
1. When such alterations involve only the lowest story of a wood frame
building ...
2. When such alterations involve the lowest story of a Type V building
or structure of R3 occupancy ...

Commentary: Like Section 403.12.1, this San Francisco provision

supplements the basic alteration trigger of Section 403.4. (And as
with Section 403.12.1, the SFEBC section numbering is inappropriate,
but has no substantive effect.) Two exceptions to Section 403.12.2
apply only to wood-frame structures.
Like Section 403.12.1, Section 403.12.2 has no “10 percent”
exception; once the trigger is pulled based on the extent of structural
work intended by any one project, a seismic evaluation and a retrofit
of any deficient elements must be performed. The evaluation and
retrofit criteria are the San Francisco criteria in Section 301.2,
including the waivers for benchmarked buildings. That said, it seems
clear that this provision is not intended to apply to voluntary seismic
work covered by Section 403.9; as such, this provision would be
clearer if it included a “but for” clause at the top, as Section 403.4
does.
Even though the intent of the provision is to trigger lateral system
evaluation and retrofit, the trigger is based on intended alteration to
the gravity load-carrying elements, not to the existing lateral system
elements. As in Section 403.4, the trigger is meant to measure the
extent of the intended work, not the adequacy of the existing lateral
system.
Section 403.12.2 uses the term “substantial structural alteration,”
which is not defined in the 2016 CEBC. Rather, the provision appears
to be self-referential. Its context suggests that a “substantial”
structural alteration is simply one where the area tributary to the
altered structural elements exceeds the 30 percent threshold. In fact,
this understanding is now confirmed by a definition in the 2018 IEBC,
as described in the commentary to the definition of Alteration, above.
The 2018 IEBC definition is close to the wording of this SFEBC
provision, but it adds a five-year timeframe to require consideration of
cumulative effects. Thus, while SFEBC Section 403.12.2 would apply
only to a single project or permit application, the 2018 IEBC would
consider any work done in the previous five years. The intent is to

ATC-119-1 Part 7: Pre-earthquake Evaluation 7-41

 prevent a project from avoiding the trigger by splitting the work into
multiple, but near simultaneous, permits.
The 2018 IEBC defines substantial structural alteration in order to use
it in a new trigger provision similar to SFEBC Section 403.12.2. 2018
IEBC Section 503.11 will read as follows:
503.11 Substantial structural alteration. Where the work area
exceeds 50 percent of the building area and where work involves a
substantial structural alteration, the lateral
load-resisting system of the altered building shall satisfy the
requirements of Sections 1609 and 1613 of the International Building
Code. Reduced seismic forces shall be permitted.
Exceptions:
1. Buildings of Group R occupancy with not more than five dwelling or
sleeping units used solely for residential purposes that are altered
based on the conventional light-frame construction methods of the
International Building Code or in compliance with the provisions of the
International Residential Code.
2. Where the intended alteration involves only the lowest story of a
building, only the lateral load-resisting components in and below that
story need comply with this section.

This coming model code provision is quite close to San Francisco’s

own current provision. The intent is to expand the basic alteration
trigger in current CEBC Section 403.4 to acknowledge projects with
intended structural work. Section 403.4 suggests the code’s intent
that when the intended project would not affect the structure, a wind
or seismic upgrade is not justified. This additional provision says that
when the intended work already involves substantial structural work,
that is the time to look at the lateral system as well. That said, there
are some differences between the SFEBC provision and the coming
IEBC/CEBC provision that will need to be worked out as San
Francisco adopts its next code:
• The exceptions are different. In particular, the second IEBC
exception would apply to all buildings with ground floor-only
alterations, not only to woodframe structures.
• By citing the new definition of substantial structural alteration, the
IEBC provision brings in a five-year timeframe for tracking
cumulative effects that is not in the original SFEBC provision.
Interestingly, however, because the SFEBC provision also uses
the term “substantial structural alteration,” it too could
inadvertently be invoking the new definition.

7-42 Part 7: Pre-earthquake Evaluation ATC-119-1

 • Most important, the IEBC provision applies only “where the work
area exceeds 50 percent of the building area,” while the SFEBC
has no such limitation. Thus, the IEBC version of this structural
trigger still only applies where the building is undergoing an
extensive (and typically non-structural) alteration project – one
that would be unlikely in an occupied tall building. The IEBC’s
work area wording reflects the intent of the new provision to match
an existing provision in the IEBC’s Work Area method.

SECTION 407
CHANGE OF OCCUPANCY

407.1 Conformance. No change shall be made in the use or occupancy of

any building unless such building is made to comply with the requirements of
the California Building Code for the use or occupancy. ...

Exception: The building need not be made to comply with the seismic
requirements for a new structure unless required by Section 407.4.

407.4 Structural. When a change of occupancy results in a structure being

reclassified to a higher risk category, the structure shall conform to the
seismic requirements for a new structure of the higher risk category. For
purposes of this section, compliance with ASCE 41, using a Tier 3 procedure
and the two-level performance objective in Table 301.1.4.1 for the applicable
risk category, shall be deemed to meet the requirements of Section 1613 of
the California Building Code.

Exceptions:
1. Specific seismic detailing requirements of Section 1613 of the California
Building Code for a new structure shall not be required to be met where the
seismic performance is shown to be equivalent to that of a new structure. A
demonstration of equivalence shall consider the regularity, overstrength,
redundancy and ductility of the structure.

2. Where a change of use results in a structure being reclassified from Risk

Category I or II to Risk Category III and the structure is located where the
seismic coefficient, SDS [sic], is less than 0.33, compliance with the seismic
requirements of Section 1613 of the California Building Code is not required.

3. [BSC] For state-owned buildings ...

ATC-119-1 Part 7: Pre-earthquake Evaluation 7-43

 Commentary: This is the key provision for change of occupancy-
triggered seismic evaluation and upgrade. The trigger is
straightforward: Any change of occupancy that involves an upward
change of Risk Category triggers a full-building seismic retrofit as
needed to match the requirements for a new structure of the higher
risk category. Essentially all existing tall buildings in San Francisco
are assigned to Risk Category II, and an existing tall building is highly
unlikely to undergo a change of use or occupancy that would move
the building to Risk Category III or IV.
Unlike the alteration trigger in Section 403.4, there is no “10 percent
exception,” and “reduced” loads are not allowed. Most changes of
use or occupancy that would pull this trigger would be accompanied
by architectural changes that must be treated as a simultaneous
alteration project. Thus, even if the simultaneous alteration does not
trigger seismic work, the change of occupancy or occupant load
might.
Exception 1 makes an allowance for obsolete detailing. However, the
provision uses undefined terms and is thus largely unenforceable.
For this reason, Exception 1 has been removed from the 2018 IEBC.
Also, ASCE 41 explicitly accounts for obsolete detailing, so Exception
1 is not necessary in any case as long as ASCE 41 is allowed.
Exception 2 does not apply in San Francisco. Exception 3 does not
apply to San Francisco buildings regulated by the SFEBC.
The 2018 IEBC will add an exception for changes of occupancy that
affect only small areas of a building. For San Francisco tall buildings,
this new exception might apply in the rare cases of a school or
residential care facility becoming a new tenant in an existing tall
building. For example, if two floors of a 25-story residential building
are converted to a care facility (Occupancy I-2), the building would be
reassigned to Risk Category III but would still be exempt from seismic
retrofit. 2018 IEBC Section 506.4.3, Exception 1, will read as follows:
Where the area of the new occupancy is less than 10 percent of the
building area and the new occupancy is not assigned to Risk
Category IV, compliance with this section is not required. The
cumulative effect of occupancy changes over time shall be
considered.

7-44 Part 7: Pre-earthquake Evaluation ATC-119-1

407.4.1 Structural. [HCD] ...

Commentary: Section 407.4.1, for residential buildings regulated by

HCD, is the same provision and exceptions as Section 407.4, but
without the allowance to use ASCE 41 and without Exception 3.

407.4.1 Change of occupancy. In addition to the other requirements of

this code, the
term “comply with the requirements of this code for
such division or group of occupancy,” as used in this section, shall
also mean compliance with the lateral force provisions of Section 301.2
when the change results in an increase of more than 10 percent in the
occupant load of the entire building or structure, and which also
increases the occupant load by more than 100 persons as compared to
the occupant load of the existing legal use or the use for which the
building was originally designed. A building changing occupancy to an
E occupancy, and is otherwise subject to Section 329, shall comply with
Section 329.

Exceptions:

1. When a change of occupancy or use involves only one story of a

building or structure, only the lateral force resisting elements in that
story and all lateral force resisting elements below need comply with
Section 301.2.

2. A change from a Group R, Division 3 to a Group R, Division 1 or

Division 2 Occupancy ...

Commentary: This San Francisco provision supplements the basic

change of occupancy trigger of Section 407.4. (The SFEBC section
numbering is inappropriate, because the CEBC already has a Section
407.4.1, as shown, but this does not affect the substance of the
provision.)
The first sentence of this San Francisco provision appears to
reference the first sentence of CEBC Section 407.1, though the
wording does not match exactly. The intent is to trigger seismic
evaluation (and possibly retrofit) with a significant increase in
occupant load even if the risk category does not change. Indeed,
depending on how one reads the term “the change results,” one could
read the provision to apply even if there is no change of use or
occupancy, but merely a higher occupant load. In any case, this

ATC-119-1 Part 7: Pre-earthquake Evaluation 7-45

 provision represents a more conservative and more likely trigger than
the basic provision in CEBC Section 407.4.
If the two-part trigger is pulled, the provision allows the reduced loads
and benchmark dates of SFEBC Section 301.2. Thus, even if this
provision is more conservative in intent than the basic provision in
Section 407.4, it is less conservative and less onerous in application.
This San Francisco provision also has two exceptions that do not
apply to the CEBC provision, but neither is likely to apply to a
triggered tall building.

7-46 Part 7: Pre-earthquake Evaluation ATC-119-1

Appendix B

Seismic Retrofit as an Alteration

Seismic retrofit is a type of alteration project, but it is not the intent of the
code that any retrofit should trigger a different or additional retrofit.
Nevertheless, a recent survey by the Structural Engineers Association of
California has revealed that the provisions of the California Existing Building
Code (CEBC; CBSC, 2016) are still commonly misunderstood and misapplied
in ways that affect retrofit projects (Zepeda et al., 2018). Partly the
misunderstanding is due to a typo in the CEBC (see the commentary to
Section 403.4 in Appendix A of this Part), but partly it is due to the history of
inconsistent local practices around the state. In any case, it is useful to clarify
the code’s intent as part of this review of current San Francisco policy.

Seismic retrofits can be classified as triggered, voluntary, or mandatory.

Triggered retrofits should be straightforward. Even if there are questions
about how to apply the “10% rule” or San Francisco’s own “2/3 rule,” it should
be clear that any retrofit triggered by the code is a self-contained structural
project. A triggered retrofit might in turn trigger other work (such as
accessibility or energy conservation improvements), and of course it cannot
be allowed to render another part of the structure noncompliant. But it is not
the intent of the code that a local retrofit in one part of the structure (for
example, one or two elements that miss the “10%” exception) should then be
considered a structural alteration that leads to a full-structure retrofit.

Might a retrofit triggered by addition, repair, or change of occupancy be

considered an alteration that triggers additional retrofit? In concept, yes. But
in practice, a retrofit triggered by one of the other project types would already
exceed what the alteration provisions might ever require, so the question
becomes moot.

Voluntary seismic retrofit is also an alteration, and many voluntary retrofits

use reduced criteria or address only certain critical deficiencies. But the code
explicitly prevents these alterations from triggering additional seismic work;
the first sentence of CEBC Section 403.4 says plainly that it does not apply to
voluntary seismic improvements, which are regulated only by Section 403.9.
(The current CEBC references the wrong section number, but this error
should be corrected in the 2019 edition.) Thus, neither the triggers in Section

ATC-119-1 Part 7: Pre-earthquake Evaluation 7-47

 403.4 nor the “10% rule,” which is an exception to Section 403.4, apply to
voluntary retrofit.

Mandatory retrofit is an alteration project as well. Like voluntary retrofits,

many mandatory retrofits are partial and could, in concept, pull the code’s
alteration trigger. Unlike voluntary retrofits, however, there is no code
provision that explicitly protects the intended project from additional triggered
work. Therefore, the clearest way to avoid confusion and scope creep is for
the mandating ordinance to include a provision that explicitly waives CEBC
Section 403.4.

7-48 Part 7: Pre-earthquake Evaluation ATC-119-1

Appendix C

References

Anagnos, T., Comerio, M., and Stewart, J., 2016, “Earthquake loss estimates
and policy implications for nonductile concrete buildings in Los
Angeles,” Earthquake Spectra, Vol. 32, No. 4.
ASCE, 2017, Seismic Evaluation and Retrofit of Existing Buildings, ASCE/SEI
41-17, American Society of Civil Engineers Structural Engineering
Institute, Reston, Virginia.
ASTM, 2007a, Standard Guide for Seismic Risk Assessment of Buildings,
ASTM E2026-07, ASTM International, West Conshohocken,
Pennsylvania.
ASTM, 2007b, Standard Practice for Probable Maximum Loss (PML)
Evaluations for Earthquake Due-Diligence Assessments, ASTM
E2557-07, ASTM International, West Conshohocken, Pennsylvania.
ATC, 2018, Recommended Earthquake Performance Goals for San
Francisco's Buildings (ESIP Task A.6.h), final report submitted to the
City and County of San Francisco prepared by the Applied
Technology Council, Redwood City, California.
Bonowitz, D., 2018, “Theories of owner liability for earthquake losses: A
debate by lawyers for non-lawyers,” Eleventh U.S. National
Conference on Earthquake Engineering: Integrating Science,
Engineering & Policy, June 25-29, Los Angeles, California.
Bonowitz, D., McCormick, D., Somers, P., 2014. “Alterations and seismic
upgrade: The building code as mitigation policy,” Earthquake
Engineering Research Institute, November 19, Oakland, California.
CBSC, 2016, 2016 California Existing Building Code, California Code of
Regulations, Title 24, Part 10, California Building Standards
Commission, Sacramento, California.
CCSF, 2011, CAPSS Earthquake Safety Implementation Program Workplan
2012-2042, the City and County of San Francisco, California.
CCSF, 2016, The San Francisco Existing Building Code, the City and County
of San Francisco, California.

ATC-119-1 Part 7: Pre-earthquake Evaluation 7-49

 City of Los Angeles, 2015, City of Los Angeles Municipal Code, Los Angeles,
California.
City of Palo Alto, 1986, Palo Alto Municipal Code, Palo Alto, California.
City of Portland, 2015, Portland City Code and Charter, Portland, Oregon.
City of Seattle, 2015a, 2015 Seattle Existing Building Code, Seattle
Department of Construction and Inspections, Seattle, Washington.
City of Seattle, 2015b, Seattle Building Code Requirements for Existing
Buildings that Undergo Substantial Alterations [Tip 314], Seattle
Department of Construction and Inspections, March 23, Seattle,
Washington.
CPCFA, 2017, California Capital Access Program (CalCAP) Seismic Safety
Financing Program, Office of the State Treasurer, California Pollution
Control Financing Authority, Sacramento, California.
CSU, 2016, CSU Seismic Requirements, California State University, Office of
the Chancellor, November 1, Long Beach, California.
FEMA, 2000, Recommended Seismic Evaluation and Upgrade Criteria for
Existing Welded Steel Moment-Frame Buildings, FEMA 351, prepared
by the Structural Engineers Association of California, Applied
Technology Council, and Consortium of Universities for Research in
Earthquake Engineering (SAC) Joint Venture for the Federal
Emergency Management Agency, Washington, D.C.
Hoover, C., 1992, Seismic Retrofit Policies: An Evaluation of Local Practices
in Zone 4 and their Application To Zone 3, Earthquake Engineering
Research Institute, Oakland, California.
ICC, 2016, 2016 Group B Committee Action Hearings, International Code
Council, April, Louisville, Kentucky.
ICC, 2018, International Existing Building Code, International Code Council,
Country Club Hills, Illinois.
Li, R., 2017, “By feasting on S.F. ‘partial stakes,’ big real estate investors are
saving millions in taxes,” San Francisco Business Times, October 26,
San Francisco, California.
Li, R., 2018, “SF tower leased to Facebook could set price record in sale.”
San Francisco Chronicle, September 10, San Francisco, California.
Littman, J., 2018, “Partial stake in transamerica pyramid for sale,”
Bisnow.com, January 24.

7-50 Part 7: Pre-earthquake Evaluation ATC-119-1

Lui, R., 2018, SHR and City-BORP Update, January 22, San Francisco,
California.
Morris, K., 2018, “Why sell a billion-dollar building when you can unload a
piece of it instead?” The Wall Street Journal, March 11, New York,
New York.
NIST, 2011, Standards of Seismic Safety for Existing Federally Owned and
Leased Buildings: ICSSC Recommended Practice 8 (RP 8), NIST
GCR 11-917-12 report, prepared by the Building Seismic Safety
Commission for the National Institute of Standards and Technology,
Gaithersburg, Maryland.
Porter, K., Beck, J., and Shaikhutdinov, R., 2004. “Simplified estimation of
economic seismic risk for buildings,” Proceedings of the 13th World
Conference on Earthquake Engineering, August 1-6, Vancouver,
British Columbia, Canada.
Schinske, D., 2015, “SEAOC Stumps for Significant Structures,”
seaoc.org/news, March 24.
SEAONC BRC, 2015, Earthquake Performance Rating System User’s Guide,
Structural Engineers Association of Northern California, Building
Ratings Committee, San Francisco, California.
SEAONC EBC Steel Frame Subcommittee, 2004, “Pre-earthquake evaluation
of welded steel moment frames: Case studies using FEMA 351,” in
SEAOC 2004 Convention Proceedings, Structural Engineers
Association of California, August 25-28, Monterey, California.
Searer, G., Paret, T., Kehoe, B., and Hachem, M., 2009, “The problems with
PMLs,” Proceedings of the 2009 ATC & SEI Conference on Improving
the Seismic Performance of Existing Buildings and Other Structures,
December 9-11, San Francisco, California.
University of California, 2017, Seismic Safety Policy, University of California,
Office of Capital Asset Strategies and Finance, May 22.
Zepeda, D., Hagen, G., O’Connell, K., McLellan, R., Buckalew, J., and
Sumer, A., 2018, “Existing buildings and the ‘10% rule’: Are we in
agreement?” Proceedings of the 2018 SEAOC Convention, Structural
Engineers Association of California, Sacramento, California.

ATC-119-1 Part 7: Pre-earthquake Evaluation 7-51

 Project Participants

City and County of San Francisco Office of Resilience and Capital Planning
Brian Strong Danielle Mieler (Project Manager)
Chief Resilience Officer and Director Principal Resilience Analyst
Office Resilience and Capital Planning Office Resilience and Capital Planning
1 Dr. Carlton B. Goodlett Place 1 Dr. Carlton B. Goodlett Place
City Hall, Room 347 City Hall, Room 347
San Francisco, California 94102 San Francisco, California 94102

Applied Technology Council

Ayse Hortacsu Justin Moresco
Applied Technology Council Applied Technology Council
201 Redwood Shores Parkway, Suite 240 201 Redwood Shores Parkway, Suite 240
Redwood City, California 94065 Redwood City, California 94065

Project Technical Committee

John D. Hooper (Task Leader) Gregory Deierlein
Magnusson Klemencic Associates Department of Civil and Environmental
1301 Fifth Avenue, Suite 3200 Engineering
Seattle, Washington 98101 Blume Earthquake Engineering Center
Stanford University
David Bonowitz Stanford, California 94305
605A Baker Street
San Francisco, California 94117 Shahriar Vahdani
Applied GeoDynamics, Inc.
1205 Contra Costa Drive
El Cerrito, California 94530

Project Working Groups

Carlos Molina Hutt Anne McLeod Hulsey
Department of Civil Engineering Department of Civil and Environmental
The University of British Columbia Engineering
Vancouver, British Columbia Stanford University
V6T 1Z4 Canada Stanford, California 94305

Preetish Kakoty Alireza Eksir Monfared

Department of Civil Engineering Department of Civil Engineering
The University of British Columbia The University of British Columbia
Vancouver, British Columbia Vancouver, British Columbia
V6T 1Z4 Canada V6T 1Z4 Canada

ATC-119-1 Project Participants P-1

Max Rattie Wen-Yi Yen
Magnusson Klemencic Associates Department of Civil and Environmental
1301 Fifth Avenue, Suite 3200 Engineering
Seattle, Washington 98101 Stanford University
Stanford, California 94305

Task Review Panel

Mark X. Haley (reviewer for Part 2) William Walton (reviewer for Part 2)
Haley & Aldrich, Inc. GEI Consultants
465 Medford Street, Suite 2200 120 West Madison
Boston, Massachusetts 02129 Suite 1305
Chicago, Illinois 60602

City and County of San Francisco Tall Buildings Executive Panel

Naomi Kelly (Chair) Tom Hui
City Administrator Director
Office of the City Administrator Department of Building Inspection
1 Dr. Carlton B. Goodlett Place 1660 Mission Street
City Hall, Room 362 San Francisco, California 94102
San Francisco, California 94102
Brian Strong
Mary Ellen Carroll Chief Resilience Officer and Director
Executive Director Office Resilience and Capital Planning
Department of Emergency Management 1 Dr. Carlton B. Goodlett Place
1011 Turk Street City Hall, Room 347
City Hall, Room 362 San Francisco, California 94102
San Francisco, California 94102

Kathryn How
Chief Engineer, Assistant General Manager of
Infrastructure
Public Utilities Commission
525 Golden Gate Avenue
San Francisco, California 94102

City and County of San Francisco Tall Buildings Stakeholders

Board of Supervisors Private Engineers and Architects
District 3 DBI Code Advisory Committee
District 6 Heller Manus Architects
Building Inspection Commission
Developers HOK
San Francisco Apartment Owners Maffei Structural Engineering
Association BXP
Tishman Speyer SEAONC AB 82/83 Code Advisory
Boston Properties Committee
Fennie+Mehl Architects

P-2 Project Participants ATC-119-1

Business City and County of San Francisco Staff
San Francisco Chamber of Commerce Department of Building Inspection
Real Estate Division
Real Estate Department of Emergency Management
Real Estate Advisory Services Planning Department
Fire Department
Community and Non-profit Organizations SFPUC
SPUR Public Works
BOMA Office of Resilience and Capital Planning
The East Cut Community Benefit District

ATC-119-1 Project Participants P-3