Energy Storage System Safety

Wisconsin PUC Workshop

PRESENTED BY

D av i d R o s e wa t e r - 06 – 9 – 2021

Sandia National Laboratories is a multimission

laboratory managed and operated by National
Technology & Engineering Solutions of Sandia,
LLC, a wholly owned subsidiary of Honeywell

SAND2021-6548 PE
International Inc., for the U.S. Department of
Energy’s National Nuclear Security
Administration under contract DE-NA0003525.
2 Presenter Bio

❑Senior Technical Staff at Sandia National Labs

❑Lab Manager for Sandia’s Energy Storage Test Pad
(ESTP)
❑Over a decade of experience in battery
cell/module/system testing
❑BS, MS in Electrical Engineering from Montana Tech
❑PhD in Electrical and Computer Engineering from
UT Austin
❑Licensed Professional Engineer in New Mexico
❑Chair of IEEE P2686 Working group on Battery
Management Systems
3 Outline

❑References
❑Background
❑Part 1: Understanding battery fires and vent gas
❑Battery fire hazards
❑Cell to Cell propagation
❑Enclosed environments and gas buildup
❑Different mitigation design strategies
❑Special Topics
❑Part 2: Emergency Response Coordination
❑Hazard analysis report
❑Design objectives for firefighter safety
❑Firefighter Training
❑Special Topics
❑Questions?
❑Backup slides on safety codes and standards
4 References

McMicken Arizona Reports Firefighter Training Video / Resources

➢ http://www.firefighternation.com/wp- ➢ https://www.nfpa.org/News-and-
content/uploads/sites/10/2020/07/document.pdf Research/Resources/Emergency-Responders/High-
➢ https://docket.images.azcc.gov/E000007939.pdf risk-hazards/Energy-Storage-Systems
➢ https://ulfirefightersafety.org/docs/Four_Firefighters_Injured_In_Lit ES Codes and Standards Update
hium_Ion_Battery_ESS_Explosion_Arizona.pdf ➢ https://energy.sandia.gov/wp-
Full Scale Fire Test Reports content/uploads/2021/02/SC-Report-by-SDO-
➢https://www.nfpa.org/-/media/Files/News-and-Research/Fire- WINTER-2021_Final.pdf
statistics-and-reports/Hazardous- Hazard Analysis on Grid-scale Energy Storage
materials/RFFireHazardAssessmentLithiumIonBattery.ashx ➢ https://www.sandia.gov/ess-ssl/wp-
➢https://ulfirefightersafety.org/docs/UL9540AInstallationDemo_Report_ content/uploads/2020/09/Rosewater-APS.pdf
Final_4-12-21.pdf ➢ https://www.epri.com/research/products/000000003
➢https://www.nfpa.org/News-and-Research/Data-research-and- 002017136
tools/Suppression/Sprinkler-Protection-Guidance-for-Lithium-Ion- Journal Articles
Based-Energy-Storage-Systems ➢ https://www.osti.gov/pages/biblio/1257985
2021 Energy Storage Safety and Reliability Forum Presentation Archive ➢ https://iopscience.iop.org/article/10.1149/1945-
➢https://www.sandia.gov/ess-ssl/2021-essrf/ 7111/ab84fa

Note that this aggregation of reference information

is not an endorsement of their content or findings.
5 Background
Safety is critical to the widescale deployment of energy storage technologies.

Bloomberg Greentech Media The Korea Times

https://www.bloomberg.com/news/articles/20
19-04-23/explosions-are-threatening-lithium-
ion-s-edge-in-a-battery-race

There is a tendency to use

the availability heuristic https://www.greentechmedia.com/articles/
when considering risk. read/aps-and-fluence-investigating-
explosion-at-arizona-energy-storage-
To avoid this, consider how facility#gs.gpky5k
https://www.koreatimes.co.kr/www/tec
many batteries continue to h/2018/12/133_260560.html
operate without problems
every day.
6

Part 1 Understanding battery fires and vent gas

 7 Hazards – Thermal Runaway

“The process where self heating occurs faster than can be

dissipated resulting in vaporized electrolyte, fire, and or
explosions”

◦ Initial exothermic reactions leading to thermal runaway can begin at 80° - 120°C.
◦ Venting of electrolyte gasses
◦ Ignition of gasses (fire or explosive)
◦ Propagation within module
◦ External flame initiates preheating of additional cells/modules

Credit: This slide was prepared by Matt Paiss of PNNL (PNNL-SA-149573)

 8 Fire Tetrahedron

◦ Typically, all that is required for fires to occur

is O2, Fuel, and Heat.
◦ Some chemistries contain oxides that release
rapidly under high heat conditions.
◦ Li-ion fires can occur in low O2 atmospheres
◦ Flammable gasses will continue to be
produced.

Credit: This slide was prepared by Matt Paiss of PNNL (PNNL-SA-149573)

 9 Hazards – Flammable/Toxic Gases

Credit: This slide was prepared by Matt Paiss of PNNL (PNNL-SA-149573)

10 Mechanics of Thermal Runaway

How does thermal runaway start?

◦ Thermal, electrical, or mechanical “abuse”
◦ However, “abuse” thresholds are statistical
properties and can change with time, usage, and
environment

Thermal runaway measured via accelerated

rate calorimetry (ARC)
1. Self-heating onset temperature
2. Thermal runaway onset temperature
3. Total heat release (ΔT)

Figure Credit: Yuliya Preger https://custom.cvent.com/5B9EB96FC2FC4AC69710004DEF407285/files/f4910d8f6dec42fb8b312e3934da8826.pdf

11 Cell to Cell Propagation

Thermal runaway in one cell can drive nearby cells

into thermal runaway.
Whether or not thermal runaway cascades through
a module, and how quickly it does, depends on
abuse thresholds, heat capacity, heat
generation rates, and heat dissipation rates.

Figure Credit: John Hewson https://www.sandia.gov/ess-ssl/wp-content/uploads/2021/ESSRF/Hewson_John.pdf

12 Cascading failure modeling/testing

◦ Successful prediction over a range of reduced

SOC and metallic inserts.
◦ Collectively add heat capacity & increase time
delay for cell runaway.
◦ Prevent propagation for 30% increase in net heat
capacity.

Kurzawski, A., et al. (2020). "Predicting cell-to-cell failure propagation and limits of propagation in lithium-ion cell stacks." Proc. Combust. Instit. 38.
13 Enclosed environments and gas buildup

If thermal runaway occurs and,

If it propagates through a module,
How might the gasses produced be hazardous?

For one cell type, thermal runaway vent gasses were

found to be combustive above 6.3% concentration in air.
Combustive strength for this sample was in-between
methane and propane.
Source: K. Marr, V. Somadepalli, Q. Horn, Explosion hazards due to failure lithium-
ion batteries, Global Congress on Process Safety (2013).
 14 Deflagration Challenges

• System with clean agent

suppression system.
• TR continued after deployment
• Deflagration occurred @ 44min
• Event required direct water
application.

Credit: This slide was prepared by Matt Paiss of PNNL (PNNL-SA-149573)

 15 Deflagration Challenges

• System with 0.5 gpm sprinkler

system.
• Deflagration 30 min post water
application.
• UL demonstration of
deflagration risks highlights that
no matter what the suppression
system – deflagration still
occurs.

Credit: This slide was prepared by Matt Paiss of PNNL (PNNL-SA-149573)

16 Different mitigation design strategies

Some of the design options available

◦ Prevent thermal runaway (e.g. non-lithium-ion chemistry)
◦ Limit the size (energy) of any one module
◦ Don’t put the battery in an enclosure
◦ Enclosure deflagration venting

Andrew F. Blum and R. Thomas Long Jr. “Hazard Assessment

of Lithium Ion Battery Energy Storage Systems FINAL REPORT”
Fire Protection Research Foundation, 2016, Available:
https://www.nfpa.org/-/media/Files/News-and-
Research/Fire-statistics-and-reports/Hazardous-
materials/RFFireHazardAssessmentLithiumIonBattery.ashx

PNNL’s INTELLIVENT Technology:

Smoke detector activated ventilation
17 Special Topics: Early Detection From Published “Li-ion Tamer®” data

Li ion batteries release trace amounts of vent gas in the self-

heating phase prior to thermal runaway.
If these gasses are detected, then shutting down
charge/discharge may prevent thermal runaway.
The effectiveness of early detection depends on all the same
factors as for propagation: abuse thresholds, heat capacity,
heat generation rates, and heat dissipation rates.
Early detection can help serve as a part of an effective
wholistic propagation resistant design but is not in-and-of-
itself sufficient to ensure that a battery cell fire will not occur.

Image source: https://www.sandia.gov/ess-ssl/wp-

content/uploads/2021/ESSRF/18_Frank_Nick_S6.pdf
18 Special Topics: Battery Management During an Outage

Safe Storage temperature of lithium-ion batteries tends to be much wider than their operating
temperature. Hence it is rarely a safety issue for the temperature management systems to be
without power for a matter of a few days. This does not hold true in extreme climates where
temperatures in an enclosure can exceed safe storage temperatures.
However, a BMS collects battery fault data that could be very useful in a post-incident analysis. If a
power outage stops data acquisition, then it can be difficult to determine what went wrong. Hence,
it is recommended that a BMS includes some backup power (often in the range of 4-8 hours).
This is often accomplished with a small bank of VRLA batteries.
19 Special Topics: Access Control

Batteries are not hazardous to passing pedestrians, so access control is

less about the potential hazard they pose and more about the hazard
that people could pose to them.
◦ Batteries are expensive/valuable so it is conceivable that they could be stolen.
◦ Batteries are sensitive to mechanical abuse so a car accident could start a fire.
◦ In cybersecurity terms, a malicious actor with physical access to the battery
management system could cause significant damage or conceivable even a fire.
20 Take-Aways From Part 1

Design should prevent thermal runaway propagation at every

level of integration.

Design should mitigate vent gas combustion hazard.

21

Part 2 Emergency Response Coordination

22 Hazard analysis report

The objective of this research is to prevent fire and

explosions in lithium-ion based energy storage systems. This
work enables these systems to modernize US energy
infrastructure and make it more resilient and flexible (DOE
OE Core Mission).
The primary focus of our work is on lithium-ion battery
systems. We apply a hazard analysis method based on system’s
theoretic process analysis (STPA) to develop “design
objectives” for system safety. These design objectives, in all
or any subset, can be used by utilities “design requirements”
for issuing requests for proposals (RFPs) and for reviewing
responses as a part of their procurement process. The design
objectives can also serve as model standards for standard
development organizations (SDOs) to consider in the course
of their consensus-based work.
Similar Efforts:
◦ EPRI Guide to safety in energy storage system
◦ NFPA 855, Standard for the Installation of Stationary Energy
Storage Systems
◦ UL 9540 Ed 2, ANSI/CAN/UL Standard for Energy Storage The full report can be found at:
Systems and Equipment https://www.sandia.gov/ess-ssl/wp-
◦ FDNY: 2020 NYC Fire Code –Section 608 STATIONARY content/uploads/2020/09/Rosewater-APS.pdf
STORAGE BATTERY SYSTEMS
 23 Design objectives for firefighter safety

When firefighters arrive on the scene of a battery system fire, they initiate an
ongoing hazard assessment with priorities being life, property, then
environment.
The system must be designed such that firefighters can understand the current
state of the system without being exposed to a hazard.
When firefighter action is required, such as in the case of an uncontrolled
runaway reaction that may spread to nearby structures, options must be available
for actions that do not expose firefighters to the hazards inside the system.

From: https://www.sandia.gov/ess-ssl/wp-content/uploads/2020/09/Rosewater-APS.pdf
 24 Safety critical information availability to firefighters

Design objective 1.1: The system includes a durable, external display, accessible from a safe location, for
firefighters to access the following information: 1) what percentage of the cells in the system have
vented, 2) is the ventilation system working as expected, 3) what voltages are present in the system, 4)
what the temperature trending history is internally, 5) what actions have been taken by the automated
systems (e.g. fire suppression), and 6) the presence or absence of any gases in hazardous
concentrations (including smoke). Note that this design objective is dependent on local firefighters
having the training to interpret the information provided, as discussed in Section 2.3.6.
OR
Design objective 1.2: The system includes continuous monitoring by a designated individual . This
designee may be a trained and qualified utility employee or integrator employee. This designee must be
able to provide firefighters with the following information: 1) what percentage of the cells in the system
have vented, 2) is the ventilation system working as expected, 3) what voltages are present in the system, 4)
what the temperature trending history is internally, 5) what actions have been taken by the automated
systems (e.g. fire suppression), and 6) the presence or absence of any gases in hazardous concentrations
(including smoke). To meet this design requirement a designee must be available at all times to respond to
the site within a specified time period. The emergency telephone number must then be provided to the
fire department and posted visibly and durably on the outside of the enclosure. Note that this design
objective is dependent on local firefighters having the training to interpret the information provided, as
discussed in Section 2.3.6.

From: https://www.sandia.gov/ess-ssl/wp-content/uploads/2020/09/Rosewater-APS.pdf
 25 Example layout for an energy storage fire alarm control panel

From: https://www.sandia.gov/ess-ssl/wp-content/uploads/2020/09/Rosewater-APS.pdf
 26 Design for safe firefighter interaction

Design objective 1.3: The system includes one or more methods for firefighters to extinguish fires
and/or ventilate the environment inside the system without being exposed to fire or a
potentially explosive environment. At a minimum this includes a grid-disconnect (E-Stop) switch.

Figure source: https://share-ng.sandia.gov/ess/wp- Image source: https://www.fireengineering.com/fire-prevention-

content/uploads/2019/04/4a_Hoff_GSS-Safety-Overview-v15.6_MHOFF- protection/energy-storage-system-installations-fire-service-awareness/#gref
with-Notes.pdf

From: https://www.sandia.gov/ess-ssl/wp-content/uploads/2020/09/Rosewater-APS.pdf
 27 Firefighter Training Objectives

Training objective: This training will be classroom based and will focus on hazard
identification, risk assessment, and actions that can be taken in different example scenarios.
Firefighters will be trained to recognize the presence of lithium-ion batteries. They will be
trained to recognize high voltage hazards, and compounding factors (such as an enclosed
spaces). They will be trained to recognize that the smoke vented from batteries during a
fire could be combustible and should be allowed to ventilate before it is safe to
approach. Trainees will be instructed on how to interpret system state information provided
by the operations designee to perform on-site risk assessment. This information will include
system hazards such as 1) what percentage of the cells in the system may have vented, 2) is the
ventilation system working as expected, 3) what voltages are present in the system, 4) what the
temperature trending history is internally, and 5) what actions have been taken by the
automated systems (e.g. fire suppression). Guidance will be provided on how a visual
inspection may provide sufficient information to assess the hazard. Finally, best practices will
be provided on determining safe entry, methods for limiting the spread of a battery fire,
identifying when the best approach is to not put out the fire (letting hazardous stored
energy be dissipated safely), and determining when it is safe to leave an incident site.

From: https://www.sandia.gov/ess-ssl/wp-content/uploads/2020/09/Rosewater-APS.pdf
 28 Impact of a Lithium-Ion Fire to Adjacent Populations

In general, as we understand currently, the smoke from a lithium-ion battery fire is as toxic to human health as a
fire in a similar mass of common plastics. However, sometimes the safest/best firefighter response to a
propagating battery fire is to simply let the fire consume the active material, thereby dissipating the stored energy,
while protecting nearby structures. This may mean that a battery fire will produce smoke for a longer duration than
fires in plastics. As toxic impact to human health is based on both severity and duration of exposure a battery fire
in a neighborhood could have a greater impact than a fire in a comparable mass of plastics. Because of this, it is
recommended that siting policy consider the smoke produced during conflagration, whether generated by
thermal runaway or external fire, and the population potentially exposed to the smoke. This consideration
impacts duel-occupancy structures, locations where evacuation options are limited, and a location’s
proximity to vulnerable populations such as schools or elderly care facilities.

Source: P. Ribiere, S. Grugeon, M. Morcrette, S. Boyanov, S. Laruel-lea, G. Marlair, Investigation on the re-
induces hazards of li-ion battery cells by re calorimetry, Energy and Environmental Science 5 (2012) 5271{5280.

From: https://www.sandia.gov/ess-ssl/wp-content/uploads/2020/09/Rosewater-APS.pdf
 29 Impact of a Lithium-Ion Fire to Adjacent Structures

The risk that a battery fire spreads to nearby structures should also be considered. Rules
structuring the placement of large, oil-filled transformers can offer reasonable guidance on
how to structure the siting/offset requirements that could be applied to battery systems.
These rules should be based on the total volume of fuel and how much heat it produces during a
fire. For a given battery type these data can be found in the abuse test data for the cell, multiplied by
the number of cells per enclosure. Fire barriers can also be considered (similar to oil-filled
transformer siting requirements).
The risk of an explosion should also be considered in the design as well as location of the
system (see design objective 3.3 in Section 2.3.3).

From: https://www.sandia.gov/ess-ssl/wp-content/uploads/2020/09/Rosewater-APS.pdf
30 Keep in mind - the status quo is not risk neutral

Proximity to fossil power plants is associated with higher rates of asthma,

birth defects, and even early death [see references below]. PV and Battery
energy storage can provide similar services without harmful emissions.

Perera, Frederica P. “Multiple Threats to Child Health from Fossil Fuel

Combustion: Impacts of Air Pollution and Climate Change.” Environmental
health perspectives vol. 125,2 (2017): 141-148. doi:10.1289/EHP299

Schwartz, Joel. “Air pollution and children's health.” Pediatrics vol. 113,4
Suppl (2004): 1037-43. https://pubmed.ncbi.nlm.nih.gov/15060197/

https://noharm-uscanada.org/sites/default/files/documents-
files/828/Health_Effects_Coal_Use_Energy_Generation.pdf

https://www.catf.us/educational/coal-plant-pollution/
31 Take-Aways From Part 2

Battery systems should be designed to isolate hazards and readily

provide safety critical information in order to protect firefighters.

The permitting process should consider the localized impacts of

a possible fire (including smoke).

The status quo is not risk neutral.

32

Questions?

David Rosewater PhD PE

Sandia National Laboratories
[email protected]
O: +1 505 844 3722

Thank you to Dr. Imre Gyuk, DOE - OE

33

Backup Slides on Safety Codes and Standards

34 CODES AND STANDARDS UPDATE

Publication released quarterly

The following activities support that objective
and realization of the goal:
1. Review and assess C/S which affect the
design, installation, and operation of
energy storage systems (ESS)
2. Identify gaps in knowledge that require
research and analysis to provide data for
technical committee inputs
3. Identify areas in C/S that are potentially
in need of revision or enhancement and
can benefit from activities conducted
under research and development
4. Develop input for new or revisions to
existing C/S through individual
stakeholders, facilitated task forces, or
through laboratory staff supporting these
efforts

Available: https://www.sandia.gov/energystoragesafety-ssl/
 35 Codes & Standards Landscape

❑Building & Fire Codes are adopted typically at a state level

❑Product and installation Standards are referenced in the Codes.
❑Codes & Standards have evolved, but much slower than deployments
❑Typically updated every 3 years.

❑Energy density and system sizes increasing

❑Cabinets not mentioned in NFPA 855 or IFC

❑Challenges with fire protection

❑Product listing expensive and time consuming
❑Inconsistent Code adoption state-state

Credit: This slide was prepared by Matt Paiss of PNNL (PNNL-SA-149573)

 36 International Fire Code (IFC)

2021 International Fire / Residential Code

◦ IFC Section 1207 – IRC Section 327
◦ Changes from 2018
◦ Harmonizes with NFPA 855
◦ Requires listing to UL9540
◦ Scope ads O&M, retrofit, commissioning, decommissioning
◦ Exemption for telecom using Pb & NiCd @ < 60VDC
◦ Suppression system based on 9540a
◦ Explosion control: NFPA 68 or 69
◦ Post-Fire Mitigation Personnel
◦ Emergency Response Plan & Training

Credit: This slide was prepared by Matt Paiss of PNNL (PNNL-SA-149573)

 37 IFC Adoption Map

Credit: This slide was prepared by Matt Paiss of PNNL (PNNL-SA-149573)

 38 National Fire Protection Association (NFPA) 855

2020 NFPA 855 Standard for the Installation of ESS

◦ 1st Edition published
◦ Scope reserved for next cycle based on appeal by utilities for
exemptions.
◦ Covers
◦ Installation
◦ Commissioning
◦ O&M
◦ Emergency Response
◦ Decommissioning
◦ Referenced directly in 2021 NFPA1 Fire Code
Rooms, areas within buildings, and walk-in units require:
1. Automatic sprinkler w/ min density of 0.3gpm/ft2
2. or, based on UL9540a fire testing data

Credit: This slide was prepared by Matt Paiss of PNNL (PNNL-SA-149573)

 39 NFPA 855 Hazard Mitigation Analysis (HMA)

Hazard mitigation analysis (HMA) shall be provided for:

1. Battery technologies not specifically covered (Other)
2. Multiple battery technologies in a room with a potential for
adverse interactions
3. When allowed as a basis for increasing MAQs

Credit: This slide was prepared by Matt Paiss of PNNL (PNNL-SA-149573)

 40 HMA Continued

The HMA evaluates single fault failures

◦ Thermal runaway in a single battery array
◦ Failure of the energy management system
◦ Failure of ventilation system
◦ Voltage surges on the primary side
◦ Short circuits on the load side of the batteries
◦ Failure of the smoke or gas detection, fire suppression
The fire code official can approve deviations or new ESS technologies
based on the HMA.

Credit: This slide was prepared by Matt Paiss of PNNL (PNNL-SA-149573)

 41 Size, Separation & Minimum Acceptable Quantity (MAQ) Limits

Credit: This slide was prepared by Matt Paiss of PNNL (PNNL-SA-149573)

 42 Explosion Protection

Two options for meeting requirement:

1. NFPA 68 – Deflagration Venting.
Blow-out panels to protect structure from explosion based on max gas
production in cell tests.
2. NFPA 69 – Deflagration Prevention.
Exhaust system designed to keep below 25% of LEL in area.

Credit: This slide was prepared by Matt Paiss of PNNL (PNNL-SA-149573)

 43 Large Scale Fire Testing

The fire code official can approve the following installations based on large-
scale fire testing:
• Increased array (unit) size
• Reduced spacing to adjacent units and/or walls
• Increased MAQ in a fire area
Testing to be conducted by an approved test lab and show:
• A fire in one unit will not propagate to an adjacent unit
• A fire in one unit will be contained within the test room
• UL 9540A was developed to conduct these fire propagation tests

2021 IFC/NFPA1 specifies UL 9540A for this testing

Credit: This slide was prepared by Matt Paiss of PNNL (PNNL-SA-149573)

 44 UL 9540 Listing

• This is a SYSTEM listing, not for components.

• Includes a UL1973 listed battery & UL1741 listed inverter
• Construction & Performance
• Mechanical & Environmental Tests
• Communications Systems
• Functional Safety
• HVAC
• Includes requirements for UL9540a fire testing

Credit: This slide was prepared by Matt Paiss of PNNL (PNNL-SA-149573)

 45 UL 9540a Test Method

• Test method, NOT a listing or certification

• 4th edition has clarifications on module test
• Used to characterize gas characterizations
for chemistries that go into TR.
• Used to design fire protection and explosion
studies.

Evaluating/interpreting test
results can be challenging

Credit: This slide was prepared by Matt Paiss of PNNL (PNNL-SA-149573)

 46 UL 9540A Cell/Module Level Testing

Cell Testing : Module Testing :

◦ Determine if thermal runaway can be ◦ Evaluate thermal runaway propagation within
induced, a module,
◦ If so, document thermal runaway ◦ Develop data on heat release rate and vent
methodology, instrumentation, gas generation rate and composition,
◦ Document fire and deflagration hazards.
◦ Determine cell surface temp at venting
and thermal runaway,
◦ Measure gas generation and
composition.

Credit: This slide was prepared by Matt Paiss of PNNL (PNNL-SA-149573)

 47 UL 9540A Unit Level Testing

◦ Document thermal runaway progression within the unit,

◦ Document if flaming occurs outside the unit,
◦ Measure heat and gas generation rates,
◦ Measure surface temperatures and heat fluxes in target units,
◦ Measure surface temperatures and heat fluxes on walls.

Acceptable results * :
❑ No flaming outside the unit under test
❑ No explosion hazard observed (incl gases <25% LEL)
❑ Maximum temperatures on target units ≤ the vent
temperature in the cell level test, and maximum surface wall
temperature rise ≤ 97 °C (175 °F) above ambient.
* Deviations from above will require the test will be conducted
with a manufacturer recommended automatic sprinkler system
or other fire protection system present.

Credit: This slide was prepared by Matt Paiss of PNNL (PNNL-SA-149573)

 48

Credit: This slide was prepared by Matt Paiss of PNNL (PNNL-SA-149573)

49 Take-Aways From Part 3

The burden of proof is on the installation. They must positively show

that their chemistry/design/installation is safe. This is done through a
hierarchy of tests and analyses by third parties.

Standards continue to evolve as we learn more and as new design

strategies are developed.