Advanced Root Cause Analysis

TapRooT
WASAC
M. Tamer 1
 Root Cause Analysis (RCA)
TapRoot

Objectives:
By the end of this workshop, participants shall be able to:
 Classify accidents

 Explore different models in accidents causations

 Use different techniques in investigating based on categories

 Actively participate in investigating teams

 Use TapRoot technique in investigation accidents

 Effectively interview concerned individuals during RCA

 Write accident investigation report and recommendations

WASAC
M. Tamer 2
 PREPARATORY WORKSHOP
 Present an accident you are aware of (clear explanation)
 Consequences (injury, damage, losses, etc.)
 Investigating team
 Elements of investigation
 Investigation result (Root Cause Analysis)
 Recommendation for mitigation

WASAC
M. Tamer 3
 What Is An
Accident/Incid
ent?

An unplanned,
unexpected event that
interferes with or
interrupts normal activity
& potentially leads to
personal injury or loss
(equipment damage).
 Injur
y
- A traumatic wound or other
condition of the body caused by
external forces, including
stress or strain.
- The injury is identifiable to time
and place of occurrence and
member or function of the
body affected and is caused by
a specific event or
incident or series or events
or incidents within a single
day or work shift.
Occupational Non-traumatic
physiological harm or loss
Illness of capacity produced by:
 Systemic infection;
 Continued or repeated
stress or strain;
 Exposure to toxins,
poisons, fumes, etc.,
Or
 Other continued and
repeated exposures to
conditions of the work
environment over a long
period of time.
 A condition that does
not meet the definition
of an injury.
 Definition
of Safety

•“safety is
noted more in its
absence than its
presence.”
 Incident/Accident
FactAnalysis
Finding
•The investigation of incidents identifies
the specific root causes and causal
factors (contributing factors) for
incidents.
•There is less emphasis on identifying the
specific individuals responsible.
•Disciplinary actions are rare but likely if
there is a history of repeated
occurrences.
•There is usually a greater amount of
explanatory detail in the incident report.
•There is greater tendency in a fact-
finding organization to report a near
miss as well as minor incident events.
Incident
Investigation
Incident investigation is the
process of identifying the
underlying causes of incidents and
implementing steps to prevent
similar events from occurring.
 Objective of Incident
Investigation
The primary objective of incident investigation is to
prevent recurrence by applying recommendation from
incident result, our knowledge and experience.

WASAC
M. Tamer 10
 .

What is not reported cannot be investigated.

What is not investigated cannot be changed.

What is not changed cannot be improved

ACCIDENT CAUSATION MODELS

WASAC
M. Tamer 12
 Heinrich’s Theories

Most accidents are the

result of unsafe
behaviors, rather than Act of God – 2%
an unsafe
environment/unsafe Unsafe
conditions. Conditions
10%

Unsafe/At-Risk
Behaviors
88%

WASAC
M. Tamer 13
Sequence of Events Model Heinrich’s Theories

MISTAKES OF PEOPLE

WASAC
M. Tamer 14
 Multiple Causation Theory
Factors combined in random
fashion to cause accidents.

WASAC
M. Tamer 15
 The Barrier Analysis Process
Hard Barriers Soft Barriers
(Engineered) (Administrative)
Machine guards Procedures
PPE Training
Fall protection Supervision
Communication
Interlocks
Work planning
Electrical systems
Standards and
Safety valves regulations

Energy/ Energy/
Hazard Hazard

WASAC
M. Tamer 16
 Barrier Categories
1. Barriers that failed The barrier was in place and operational
`` at the time the accident, but it
failed to prevent the accident.

2. Barriers that were The barrier was available, but workers

not used chose not to use it

3. Barriers that did not The barrier did not exist at the time of the
exist accident

4. Wrong or ineffective The barrier exist but was ineffective at the

Barriers time of the accident

WASAC
M. Tamer 17
 • Effectiveness – how well it meets
General its intended purpose
Characteristic •
• Availability – assurance the barrier
s of Barriers will function when needed
• Assessment – how easy to
determine whether barrier will
work as intended
• Interpretation – extent to which
the barrier depends on
interpretation by humans to
achieve its purpose

WASAC
M. Tamer 18
 “Work-as-Done” Varies from “Work-as-Planned” at Employee Level

Work Performance Gap (ΔWg) “what” variation exists?

“why” the variation exists
 •Many people are
adversely affected
•Workers,
Enterprise,
Environment, Country
 Hazard–Barrier–Target
Theory Swiss
Cheese Theory
Latent/ Dormant Failure Model

This model views the accident to be the result of long-standing

deficiencies that are triggered by the active failures
WASAC
M. Tamer 21
 Factors Contributing to Organizational Drift

WASAC
M. Tamer 23
 Preventing System Accidents

WASAC
M. Tamer 24
 Safeguard/Barrier Analysis
Strongest 1.Remove/substantially reduce the hazard
2. Remove the target
3. Guard the target with effective safeguard
4. Improve human performance with good human
factors design.
5. Improve human performance with rules, proc.,
signs

6. Improve human performance with training

supervision, selecting smart people, etc.

Weakest

WASAC
25
M. Tamer
 Human Error
Human Error
Is not a cause of failure. It is the effect, or symptom, of deeper trouble. It is
not random, it is systematically connected to features of people’s tools,
tasks and operating environment.
It is not the conclusion of an investigation. It is the starting point!

Old View
Human error is a cause of accidents
To explain failure you must seek failure
You must find people’s inaccurate assessments, wrong decisions, bad
judgements

New View
Human error is a symptom of trouble deeper inside a system
To explain failure, do not try to find where people went wrong
Instead, find how people’s assessments and actions made
sense at the time,
given the circumstances that surroundedWASAC them.
. M. Tamer 26
 Organizational Factors
The Organizational Factors layer (slice) represents the defenses put
into place by top management. This level of system defenses might
include a company culture that puts safety first, and management
decisions that reinforce safety by providing well-trained employees
and well-designed equipment to do the job.
Resource Management Organizational Climate Organizational Process

 Inadequate  Inadequate  Inadequate

management of organizational established
human resources structure conditions of work
 Inadequate  Inadequate  Inadequate
management of organizational established
monetary resources policies procedures
 Inadequate design  Inadequate safety  Inadequate
and maintenance of culture oversight
facilities  Inappropriate
reward/punishments
WASAC
M. Tamer 27
 Unsafe Supervision
The second layer of defenses is the “supervision” slice. This refers to the
first-line supervisor and his safety-consciousness as well as organizational
factors as displayed by the operational decisions he makes.
Inadequate Planned Inappropriate Failed to Correct Supervisory
Supervision Operations a Known Problem Violations
•Failed to provide • Failed to provide • Failed to •
guidance correct data correct Authorized
•Failed to provide •Failed to provide document in unnecessary
operational adequate briefing time error hazard
principle •Failed to provide •Failed to •Failed to
• Failed to provide proper staffing identify an at- enforce rules
training •Failed to provide risk behavior and regulations
•Failed to track adequate operational •Failed to initiate •Authorized
qualifications procedure or plan corrective action unqualified
•Failed to track worker
performance

WASAC
M. Tamer 28
 Preconditions for Unsafe Acts
Certain substandard preconditions may foster a climate where incidents
can occur. Those preconditions may be related to human factors,
practices, or interface with work conditions or the environment.

Adverse or Inadequate Conditions of Operators

Adverse Mental States Adverse Physiological Physical/Mental
States Limitations
 Impaired  Insufficient reaction
 Focused attention physiological state time
 Distraction  Medical illness  Poor vision/hearing
 Mental fatigue  Physiological  Lack of knowledge
 Rush incapacitation  Incompatible
 Loss of situational  Physical physical capability
awareness fatigue
 Misplaced motivation

WASAC
M. Tamer 29
 Adverse or Inadequate Practices of Operators

Shift Resource Management Personal Readiness

 Impaired communications due to  Impaired due to
language difference medication
 Interpersonal conflict among  Inadequate rest
crew
 Failed to use all available
resources
 Misinterpretation of traffic calls
 Failed to conduct adequate
briefing
 Impaired communication/conflict
due to cultural difference
WASAC
M. Tamer 30
 Adverse or Inadequate Work Interface
Design Issues Maintenance Issues
 Ambiguous  Poorly maintained
instrumentation equipment
 Inadequate  Poorly maintained
layout or space workspace
 Substandard  Poorly maintained
illumination communications equipment
 Substandard
communications equipment
 Equipment substandard for
job

WASAC
M. Tamer 31
 Three Basic Causes of Accident

WASAC
M. Tamer 32
 1. Direct Cause
Three •The direct cause of an accident
Basic is the immediate events or
Causes conditions that caused the
accident.
•The direct cause should be stated
in one sentence
Ex.:
The direct cause of the accident was
the inadvertent activation of
electrical circuits that initiated the
release of CO2 in an occupied space.
 Direct & Indirect/Unrelated

2. Indirect Causes- Causal Factors –

Contributing Factors

•Contributing causes are events or

Three conditions that collectively with other
Basic causes increased the likelihood of an
Causes accident but that individually did not cause
the accident.

•Contributing causes may be

longstanding conditions or a series of prior
events that, alone, were not sufficient to
cause the accident, but were necessary for
it to occur.
 • Ex.

• Failure to implement safety

procedures in effect for the project
Three contributed to the accident.
Basic • Failure to erect barriers or post
warning signs contributed to
Causes the accident.
• Presence of frayed cable may lead
to electrical failure or/and fire
Three Basic
Causes
3. Root Causes
•Root causes are the causal factors that, if corrected,
would prevent recurrence of the same or similar accidents.
•Root causes may be derived from or encompass
several contributing causes.
•They are higher-order, fundamental causal factors that
address classes of deficiencies, rather than single problems or
faults.
 Three Basic
Causes
3. Root Causes
For example, root causes can include failures in management
systems to:
• Ensure that safety standards and requirements are known and
applied to work activities
• Ensure that staff are competent to perform their
responsibilities
• Ensure that resource use is balanced to meet critical mission
and safety goals
• Assessments performed by the Dept. HSE Auditor failed to
identify that some safety standards were not addressed
• by contractorofsafety
Implementation these management
requirements
WASAC
systems.
would have prevented
M. Tamtehr e 37
 Characteristics Failure Type Examples

Associated with familiar A simple, frequently-performed

tasks that require little physical action goes wrong:
conscious attention. • flash headlights instead of
operating windscreen
These ‘skill-based’ errors Slip wash/wipe function
occur if attention is (Commission)
• move a switch up rather
diverted, even momentarily. than down (wrong action
on right object)
Resulting action is not • take reading from wrong
intended: ‘not doing what instrument (right action on
you meant to do’. wrong object)
Common during
maintenance and repair • transpose digits during data
activities. input into a process
control interface

WASAC
M. Tamer 39
Characteristics Failure Type Examples
Short-term memory lapse; omit to
Lapse perform a required action:
(Omission)
• forget to indicate at a road
junction
• medical implement left in patient
after surgery
• miss crucial step, or lose place, in
a safety-critical procedure
• drive road tanker off before
delivery complete
(hose still
connected)

WASAC
M. Tamer 40
 Examples Typical Control Measures
A simple, frequently-performed physical action • Human-cantered design
goes wrong: (consistency e.g., up always
• flash headlights instead of operating means off; intuitive layout of
windscreen wash/wipe function controls and instrumentation;
• move a switch up rather than down level of automation etc.)
(wrong • checklists and reminders;
action on right object) procedures with ‘place
• take reading from wrong instrument (right markers’
action on wrong object) (tick off each step)
• transpose digits during data input into a
process control interface • removal of distractions and
interruptions
• sufficient time available to
Short-term memory lapse; omit to perform a complete task
required action: • warnings and alarms to help
• forget to indicate at a road junction detect errors
• medical implement left in patient after • often made by experienced,
surgery highly-trained, well-motivated
• miss crucial step, or lose place, in a staff: additional training not
safety- valid
critical procedure
• drive road tanker off before delivery
complete (hose still connected)

WASAC
M. Tamer 41
 Compliance/Noncompliance Technique

Useful when investigators suspect noncompliance to be a causal factor.

 Don’t Know: Questions focus on whether an individual was aware of or had
reason to be aware of certain procedures, policies, or requirements that
were not complied with.
 Can’t Comply: This category focuses on what the necessary resources are,
where they come from, what it takes to get them, and whether
personnel know what to do with the resources when they have
them.
 Won’t Comply: This line of inquiry focuses on conscious decisions to not
follow specific guidance or not perform to a certain standard.

WASAC
M. Tamer 42
 WASAC
M. Tamer 43
 CONDUCTI
NG THE
INVESTIGATI
ON

WASAC
M. Tamer 44
 Taking Control
of the Accident
Scene
Before arriving at the site, assure that
the scene and evidence are properly
secured, preserved, and documented
and that preliminary witness
information has been gathered.
At the accident scene, the
Chairperson should:
 Obtain briefings from all
persons involved in managing the
accident response.
 Obtain all information and
evidence gathered by the team.

WASAC
M. Tamer 45
Accident Investigation
Team
The Chairperson is responsible for ensuring
that all Investigation Team members work as
a team and share a common approach to the
investigation; brief them on:
 The scope of the investigation,
 The schedule and plan for completing
the investigation
 Information control and release protocols
 Recording and tracking incoming
and outgoing correspondence.
 The Investigation Team will clarify and
confirm priorities on handling the following
items:
 Handle serious injuries, a fatality, or
serious off-site effects,.
 Secure the area in preparation for
the investigation.
Accident
 Participate in an initial orientation
Investigat tour.
ion Team  Provide for the initial photographic or
videotape coverage of the directly affected
areas and the surrounding areas.
 Do not concentrate on the worst
damage and fail to photograph the fringe
of the undisturbed surroundings.
 Have a size reference in the picture, such
as a pencil.
Accident Investigation Team

The Chairperson should establish communication guidelines

and serve as an effective role model in terms of the following:
 Be clear and concise; minimize the tendency to think out
loud or tell “war stories.”
 Be direct and make your perspective clear.
 Use active listening techniques, such as focusing
attention on the speaker, paraphrasing, questioning, and
refraining from interrupting.
 Pay attention to non-verbal messages and attempt
to verbalize what you observe.
 Attempt to understand each speaker’s perspective.
Accident Investigation
Team
 Seek information and opinions
from others, especially the less
talkative members.
 Consider all ideas and arguments.
 Encourage diverse ideas and
opinions.
 Suggest ideas, approaches,
and compromises.
 Help keep discussions on track
when
they start to wander .
Guidelines in Managing Information
Collection Activities.

 Review and organize witness

statements, facts, and background
information provided by the witnesses or
other sources and distribute these to the
Board.
 Organize a Board walk‐through of the
accident scene, depicting events
according to the best understanding of
the accident chronology available at the
time.
 This can help the Board visualize
the events of the accident.
  Assign an administrative
coordinator to oversee the
organization, filing, and security of
collected facts and evidence.
 Develop draft of objectives
and areas to be covered in
topical
Guideline initial interviews and oversee
s in development of a standardized list
Managing of initial interview questions to
save interviewing time and
Informatio promote effective and efficient
n interviews.
Collection  If deemed appropriate, issue a
Activities site or public announcement
soliciting information concerning
the accident.
Guidelines in Managing Information
Collection Activities
 Ensure that witnesses are identified, and
interviews scheduled.
 Ensure that Board members preserve
and document all evidence from the
accident scene.
 Make sure all Board members enlist the aid
of technical experts when making decisions
about handling or altering physical evidence.
 Establish a protocol agreeable to the Board
for analyzing and testing physical evidence.
Guidelines in Managing Information
Collection Activities.
 Identify and initiate any
necessary physical tests to be
conducted on evidence.
 Assess and reassess the need
for documents, including medical
records, training records,
policies, and procedures, and
direct their collection.
 Emphasize to Board
members that to complete the
investigation on schedule, they
must prioritize.
Controlling the
Investigation
 Is the investigation on schedule?
 Is the investigation within scope?
 Are Board members, advisors, consultants, and support staff
focused and effective?
 Are additional resources needed?
 Are daily Board meetings still necessary and productive, or should
the interval between them be increased?

WASAC
M. Tamer 54
 Conducting
 Develop a timeline for the
the events leading up
to the incident, when Investigatio
this type of information n
is appropriate.
 Plan for
coordination and
communication with
other functions.
not
Seek to obtain
place blame.facts,
 Determining
Facts
Immediately following any accident,
much of the available information may
be conflicting and erroneous.
The principal challenge is to
distinguish between accurate and
erroneous information. This can be
accomplished by:
 Understanding the activity that
was being performed at the time of
the accident or event.
 Personally, conducting a walk-
through of the accident scene or,
work location.
 Determining
Physical Barriers – Ex. Facts
Facts related to physical barriers on the day of the accident are as
follows:
 There were no general barriers, warning lines, or signs to alert
personnel on top of the construction materials to the fall hazards
in the area.
 The platform was intended to catch falling tools or parts, but it
was also used as a work platform for personnel with 100 percent
fall protection.
 There were no static lines or designated (i.e., engineered)
anchor points for personnel to connect fall protection equipment
in the vicinity of the platform.
 Lighting in the area of the platform was measured at 2
foot‐candles. .
WASAC
M. Tamer 57
 Collecting
Evidence
Three key types of evidence are
collected during the investigation:
1. Human or testamentary
evidence includes witness
statements and observations;
2. Physical evidence is matter
related to the accident (e.g.,
equipment, parts, debris,
hardware, and other physical
items); and
3. Documentary evidence includes
paper and electronic
information, such as records,
reports, procedures, and
documentation.
 Collecting
The process of Evidence
pursuing evidentiary
material involves:
 Collecting human evidence
(locating and interviewing
witnesses);
 Collecting physical evidence
(identifying, documenting, inspecting,
and preserving relevant matter);
 Collecting documentary evidence;
 Examining organizational
concerns, management systems,
and line management oversight;
and
 Preserving and controlling
 Collect and Catalog Physical

Evidence
Equipment
 Tools
 Materials
 Hardware
 Operation facilities
 Pre- and post-accident positions of accident-related elements
 Scattered debris
 Patterns, parts, and properties of physical items associated with the
accident.

Identify Preserve Collect Examine Analyze

WASAC
M. Tamer 60
 Collect and
 Less obvious but potentially Catalog
important physical evidence includes Physical
fluids (liquids and gases). Evidence
 Many facilities use a multitude of
fluids, including chemicals, fuels,
hydraulic control or actuating fluids,
and lubricants. Analyzing such
evidence can reveal much about the
operability of equipment and other
potentially relevant conditions or
causal factors.
 In addition to pathogens, any
evidence may create a hazard for
persons handling it. This aspect of any
evidence should be considered and
addressed before handling it.
 Sketch and Map Physical

Evidence
Sketch and map the position of debris, equipment, tools,
and injured persons.
 Position maps convey a visual representation of the
scene immediately after an accident.
 Evidence may be inadvertently moved, removed, or
destroyed, especially if the accident scene can only be partially
secured. Therefore, sketching and mapping should be
conducted immediately after recording initial witness
statements.
 Precise scale plotting of the position of elements can
subsequently be examined to develop and test accident causal
theories.

WASAC
M. Tamer 62
 Photograph and Video Physical

Evidence
Photos or videos can identify, record, or preserve
physical evidence that cannot be effectively conveyed by
words or collected by any other means.
 Photographic coverage should be detailed and complete,
including standard references to help establish distance and
perspective.
 Video should cover the overall accident scene, as well as
specific locations or items of significance.

WASAC
M. Tamer 63
 Inspect Physical
Evidence
Following initial mapping and photographic recording, a systematic
inspection of physical evidence can begin. The inspection involves:
 Surveying the involved equipment, structures, etc., to
ascertain whether there is any indication that component parts
were missing or out of place before the accident;
 Noting the absence of any parts of guards, controls, or
operating indicators (instruments, position indicators, etc.)
among the damaged or remaining parts at the scene;
 Identifying as soon as possible, any equipment or parts that
must be cleaned prior to examination or testing and transferring
them to a laboratory or to the care of an expert experienced in
appropriate testing methodologies;

WASAC
M. Tamer 64
 Remove Physical
Evidence
Following the initial inspection of the scene,
investigators may need to remove items of physical
evidence. To ensure the integrity of evidence for later
examination, the extraction of parts must be controlled
and methodical.
 The process may involve simply picking up components
or pieces of damaged equipment, removing bolts and
fittings, cutting through major structures, or even
recovering evidence from beneath piles of debris.
 Before evidence is removed from the accident scene,
it should be carefully packaged and clearly identified.
 Investigator’s kit can provide general purpose
cardboard
tags or adhesive labels for this purpose.
 Remove Physical
Evidence
When preparing to remove physical
evidence, these guidelines should be
followed:
Extraction and removal or movement of
parts should not start until:
 witnesses have been interviewed,
since visual reference to the accident
site can stimulate one’s memory.
 position records (measurements
for maps, photographs and video)
have been made.
Remove Physical
Evidence
Be aware that the accident
site.
 Care during extraction
and preliminary
examination is necessary to
avoid defacing or distorting
impact marks and fracture
surfaces.
 The team lead and
investigators should concur
when the parts extraction
work can begin, in order to
assure.
 Collect and Catalog Documentary
Evidence
 Documentary evidence can provide important data (i.e.,
proof of “work-as-done”) and should be preserved and
secured as thoroughly as physical evidence.
 This information might be in the form of documents, photos,
video, or other electronic media, either at the site or in files at
other locations (this information should not be confused with
procedures and such).
 Some work/process/system records are retained only for
the workday or the week. Once an event has occurred, the
team must work quickly to collect and preserve these records
so they can be examined and considered in the analysis.
Electronic
Files
 Work orders, logbooks, training records
(certifications/qualifications), forms, time
sheets
 Problem evaluation reports
 Occurrence reports
 Nonconformance reports
 Closeout of corrective actions from
similar events
 Process metrics
 Previous lessons learned
 External reviews or assessments
 Internal assessments (management
and self assessments .
 Collecting Human
Evidence
Human evidence is often the most insightful and also the most
fragile. Witness recollection declines rapidly in the first 24 hours
following an accident or traumatic event.
 Therefore, witnesses should be located and
interviewed immediately and with high priority.
 As physical and documentary evidence is gathered and analyzed
throughout the investigation, this new information will often
prompt additional lines of questioning and the need for follow up
interviews with persons previously not interviewed.

WASAC
M. Tamer 70
Witness
es &
Intervie
ws
 Witness
 es
Principal witnesses and eyewitnesses
are identified and interviewed as soon as
possible.
 Principal witnesses are persons who
were actually involved in the accident.
 Eyewitnesses are persons who directly
observed the accident or the conditions
immediately preceding or following the
accident.
 General witnesses are those with
knowledge about the activities prior to or
immediately after the accident (the
previous shift supervisor or work
controller, for example).
Intervie
ws •A partial list of those who should
be interviewed includes:

• Those involved in accident

• Those who observed the incident
• Supervisors & managers
• People not directedly involved but
who have similar background
and experience
• Tech. experts, training personnel,
suppliers, manufacturers.
• etc.
 Management
Interviews
Management interviews are somewhere in between
Participant Interviews and Technical Expert interviews.
Managements participation in an incident usually isn't
direct.

But managers have more

involvement than most
technical experts. As W.
Edwards Deming said:

People work in the

system,
Mana
Management
Interviews
•For management interviews, try to elicit
how management sees their role in
developing the systems and enforcing the
policies, procedures, and rules of the system.
•You will often be collecting information
that will help you answer questions in the
Management System portion of the Root
Cause Tree®. These questions can often raise
sensitive topics about:
• What management knew about the work.
• How management planned to successfully
complete the work.
Management
Interviews
• How management allocated resources.
• What subtle signals management sent to employees.
• The adequacy of policies, rules, and procedures.
• How management enforced policies, rules, and procedures
• How involved management was in evaluating site
operations or services.
• How complaints or news about problems
reached
management's attention.
• How involved management was in problem
solving.
• How much importance management put on implementing
corrective actions.
 Management
Interviews
•To be prepared for this type of interview,
the interviewer will need to know the
sequence of events that led to the accident.
Therefore, management interviews are
usually done AFTER the interviews of
participants and the technical experts.
•The interviewer will also need a
complete understanding of the laws,
standards, rules, and policies that applied
to the work being performed, and their
enforcement.
•The interviewer should also be
familiar with any audit/observation
statistics that are applicable and
available.
Participants/
Witness
Interviews
One interviewer and one
interviewee
• Simplest to conduct
• Hold the interview in a
private location with no
distractions or perform
the interview at the
scene of the incident
 Participants/Witness
Interviews
Two interviewers and one interviewee
• May be used for controversial or
complex incidents.
• The interviewers should coordinate
their questioning to avoid
overwhelming the interviewee.
• Don’t gang up on the interviewee
or
ask rapid-fire questions.
• After the interview, the interviewers
should compare their
impressions of the
interviewee and reach a
consensus on their observations.
 14-Steps Cognitive Interviewing
Prevent interruptions, encourage more details, and stimulate the
memory

WASAC
M. Tamer 80
14-Steps
Cognitive
Interviewi
ng
 14-Steps Cognitive Interviewing

WASAC
M. Tamer 82
 14-Steps Cognitive Interviewing

After the Interview

WASAC
M. Tamer 83
 14-Steps
Cognitive
Interviewing
Create a Relaxed Atmosphere
□Conduct the interview in a
neutral location that was not
associated with the accident.
□Introduce yourself and shake
hands.
□Be polite, patient, and friendly.
□Treat witnesses with respect.
Cognitive
Interviewing
• Prepare the Witness
• Describe the investigation’s
purpose: to prevent accidents,
not to assign blame.
• Explain that witnesses may be
interviewed more than
once.
• Stress how important the facts
given during interviews are to
the overall investigative
process.
Cognitive
Interviewi
ng
Record Information
• Note crucial information
immediately in order
to ask meaningful
follow‐up questions.
Cognitive
Interviewing
Ask Questions
• Establish a line of questioning and
stay on track during the
interview.
• Ask the witness to describe the
accident in full before asking
a structured set of questions.
• Let witnesses tell things in their own
way; start the interview with a
statement such as "Would you please
tell me about...?"
• Ask several witnesses similar
questions to corroborate facts.
Cognitive
Interviewing
Ask Questions
• Aid the interviewee with reference
points; e.g., "How did the lighting
compare to the lighting in this room?"

• Keep an open mind; ask questions that

explore what has already been stated
by others in addition to probing for
missing information.
• Use visual aids, such as photos,
drawings,
maps, and graphs to assist witnesses.
 Cognitive Interviewing

Close the Interview

• Be an active listener, and give the witness feedback; restate
and rephrase key points.
• Ask open‐ended questions that generally require more than a
"yes" or "no" answer.
• Observe and note how replies are conveyed (voice inflections,
gestures, expressions, etc.). End on a positive note; thank
the witness for his/her time and effort.
• Allow the witness to read the interview transcript and comment if
necessary.
• Encourage the witness to contact the board with additional
information or concerns.
• Remind the witness that a follow‐up interview may be
 Cognitive
Interviewing
Uncooperative Witness.
•If confronted with a witness who refuses to
testify, they cannot be forced to testify.
• Emphasize that testimony is voluntary.
• Reemphasis purpose of the investigation
• Offer to reschedule the interview if there is anything
the witness is uncomfortable with such as time,
location, or lack of representation.
• Ask if the witness is willing to explain reason for
refusal to testify.
• Offer the witness contact information in case they
should change their mind. Then, close the interview,
noting possible state of mind issues.
Advantages and Disadvantages of
Individual vs. Group Interviews

WASAC
M. Tamer 91
 Layers of Incident
Causes
Accident investigation is like peeling an onion.
Beneath one layer of causes there are other layers.
The outer layers deal with the immediate causes
while the inner layers are concerned with the
underlying causes such as weakness in the
management system.

WASAC
M. Tamer 92
 Isolate Fact
From Fiction
 Establish the norms
 Use NORMS-based
analysis of information
•Not an interpretation
•Observable
•Reliable
•Measurable
•Specific
 If an item meets all five
of above, it is a fact.
 NORMS Of
Objectivity
Objective Subjective
Not an Interpretation Interpretations
Based on a factual description based on personal interpretations/biases.

Observable Non-observable
based on what is seen or heard. based on events not directly observed.

Reliable Unreliable
Two or more people independently agree on what they Two or more people don’t agree on what
observed
they observed.

Measurable Non-measurable
a number is used to describe behavior or a number isn’t used.
situation

Specific General
based on detail definitions of what based on non detailed descriptions.
happened.

WASAC
M. Tamer 94
INVESTIGATI
ON
TECHNIQUE
S
Investigation
Techniques Causal
Factor
Tree

Change
Others
Analysis

Investigation
Techniques
Event and Time
Cause Sequence
Analysis Model

Fault Tree
Analysis TapRooT
(FTA)
UNDERSTAND
ING
TapRoot
 What is a
Root
A RootCause?
Cause is the most
basic cause (or causes) that
can reasonably be identified
that management has control
to fix and, when fixed, will
prevent (or significantly
reduce the likelihood of) the
problem’s recurrence.

Focuses on the cause

Root
Cause
Analysis
Albert Einstein said:

“We can’t solve problems

by using the same kind of
thinking we used when
we created them.”
 What is a Root Cause Analysis?
RCA is the search for the best practices and/or the
missing knowledge that will keep a problem from
recurring.

Model of
“Accident
Causation”

WASAC
M. Tamer 100
Traditional
Accident/Incid
ent
Investigations
You can’t solve all human performance problems with discipline,
training, and procedures.
If you look at most industrial accident/incident investigations, you
find three standard corrective actions:
1. Discipline. Which starts with the common corrective action:
“Counsel the employee to be more careful when …”.
2. Training. This may be the most used (and misused) corrective
action of all.
3. Procedures. If you don’t have one, write one. If you already
have one, make it longer.
Often, people can’t see
effective corrective actions
even if they can find the root
causes.

Why? Because they have

performed the work the same
way for so long that they can’t
imagine another way to do it. Traditional
Accident/Incid
ent
Investigations
 WASAC
M. Tamer 103
7
Secret
• Your root cause analysis is only as good as the information
you collect.
s• Your knowledge (or lack of it) can get in the way of a good
root cause analysis.
• You have to understand what happened before you can
understand why it happened.
• Interviews are NOT about asking questions.
• You can’t solve all human performance problems with
discipline, training, and procedures.
• Often, people can’t see effective corrective actions
even if
they can find the root causes.
• All investigations do NOT need to be created equal (but
some investigation steps can’t be skipped).
TapRooT® 7-Step Process
 Before You
You need to get started
Startbefore an incident occurs.
Why? Because preparation saves time and prevents the
loss of irreplaceable evidence.
1. Write an investigation policy to:
 Standardize what get investigated
 Require evidence preservation
 Specify the notification of co. personnel and reg.
agencies
 Outline the investigation process
 Assign responsibilities
 Establish special requirements for inv.
 Before You
 Start
Establish the investigation review and approval process
 Outline the corrective action implementation process including
tracking
 Define requirements for corrective action verification and validation
 Specify investigation report requirements
 Specify investigation report retention requirements
 Outline requirements for auditing and improving the inv. process

WASAC
M. Tamer 107
Before You
Start
2. Prepare for evidence preservation.
Ensure that evidence is preserved by shift personnel and
first responders by having them use the SAy ESPN
Technique.
S: Safely arrive at the scene
A: Assess and take control of the scene (evaluate for
hazards, avoid creating additional injuries, establish
incident command roles and responsibilities)
E: Emergency services – are for the injured; protect the
env.
S: secure the scene
P: Preserve evidence
N: Notify appropriate company personnel and regulatory
Before You
Start
3.Develop investigation team requirements for certain
classes of problems.
4.Specify training requirements for the team and get
them trained.
5. Build an investigation kit
6.Establish contracts with expert consultants and analysis
laboratories for broken equipment and oil analysis, or others.
7.Establish consulting contracts with expert RC facilitators
or human performance investigation expert to help with
investigations that are beyond the skills of in-house
investigators.
8.Consult with your corporate council about
legal requirements.
What is a
Root Cause
Analysis?

How errors impact

accidents
An accident can happen
due to the impact of errors
on the Hazard, the
Safeguard, or the Target.
 What is a Root Cause
Stop the error Analysis?
that:
• Allows a Hazard to exist or to become too large
• Allow a safeguard to fail
• Fail to put a safeguard in place where one is needed
• Fail to keep a Target safely away from a hazard.
• Fail to prevent the incident from becoming worse after a
hazard contacts a Target

WASAC
M. Tamer 111
 How Root Causes Cause
Accidents
In taproot each error is called a “causal factor”.

Each causal factor has at least one Specific Root Cause and may have
one or more Generic Causes.
Most Specific Causes are related to some type of human error.

WASAC
M. Tamer 112
TapRoot
APPLICATION
 Plan the
Investigation

What? SnapCharT:
• A map of events &
related
conditions for
an incident
• Planning Tool
• Data Collection
Tool
• The basis for the
rest of your
analysis
 SnapCha
rT®
A SnapCharT® is the most fundamental tool for performing an
investigation. It will be used throughout the investigation.
In Step l, the Spring SnapCharT® is used to:
 Develop an initial picture of what happened
 Decide what information is readily available and what needs to
be collected
 Establish a list of potential witnesses to interview
 Highlight conflicts that exist in the preliminary information
 Plan the next steps in the investigation

WASAC
M. Tamer 115
 TapRooT® 7-Step
Process
How SnapCharT®s Help
In the TapRooT® System, the first tool an investigator uses is a
SnapCharT®.

The SnapCharT® is a visual depiction of the evidence. It

focuses
the investigator on “What happened?”
Any assumptions (not verified facts)
are easily identified by their dashed ?????
boxes.
The investigator then continues to look for more evidence
to verify the assumptions or show a different, but also
possible, sequence of events and conditions.
WASAC
M. Tamer 116
 SnapCha
rT®
To draw a Spring SnapCharT®,
sketch out what you know on a
sheet of paper or Whiteboard.
You may know the loss- the
reason you are investigating.
Write the worst thing that
happened in the circle on the
SnapCharT®.
 SnapChart
Events Event Events

assumptions

Events Incident Events

Incident

Condition
Condition

Condition
WASAC
M. Tamer 118
 Draw Events Next
To decide what goes into an Event, ask
“what action happened next?”
Events are conditions – Active Verbs, One action per box

Sequenced in Chronological order (Real Time)

Date/Time

Who does Who does What Who does

What? What? equipment does Incident what after the
what?
incident?

Dashed box or oval = assumption …… not yet proven fact.

Use job titles/functions instead of names

WASAC
M. Tamer 119
 How Events Flow on your SnapChart
a) Build START to FINISH

Incident

OR
b) Build FINISH to START

??? Incident

??? ??? Incident

OR
WASAC
c) Fill in the BLANKS
M. Tamer 120
 Now Add Conditions
What do I know about each Event? Clarifying facts/data
–
• Was there anything about this Event that • Positive or negative
was different than desired?
• Quantified if possible

Conditions use passive verbs,

“be” – is, was, has been, …..
WASAC
M. Tamer 121
 More About Conditions
Conditions
should NOT
Include
action
steps

Conditions may be actions NOT Conditions should be:

done • Factual
• Non-judgmental
Conditions tell about the Events: • Precise/Quantified

How/What/Where/Why/ To What Extent/ Under What Conditions

How actions or equipment response was different from desired?
WASAC
M. Tamer 122
 SnapCha
rT®

Doctor’s unauthorized
removal of the
Cataract from the left
eye,

WASAC
M. Tamer 123
 SnapCharT®
On a preliminary Chart, recognizing what you do NOT
know may be as important as what you do know.

In the fig. (Chart), we don’t

know what happened between
the signing of the consent from
the Doctor removing the
cataract in the left eye.
Therefore , we’ve placed an
empty dashed box on the chart
to denote the missing inf.

WASAC
M. Tamer 124
SnapCha
rT®
At this point, you are ready to review
the inf. on the chart.
• Does any part of the story not
make sense?
• Is there a conflicting information?
• What problems do you need to
investigate further?
• What inf. do you need from
interviews?
• What evidence do you need to
collect?
 Preliminary
SnapChart used to
plan the investigation

WASAC
M. Tamer 126
 SnapChart Guidelines Summary
Conditions:
Event?
Tell what we know about an Event
“Who does What?”
OR Some ex.
“What does What” • How/What/Where/Why/ To What Extent/
Under What Conditions
One Action per Event • What required actions were not done?
• How did equipment fail?
Include dates/times • What was different from design?
10/03/2023

If assumption/Unverified Facts:
ALL items on the Chart should be:
• Dashed box
• Factual
• Non-judgmental • Dashed Oval
• Precise/Quantified
WASAC
M. Tamer 127
 Exercise
At the end of their shift, two employees leave their work and
head to their cars in the parking lot, one employee step in a
pothole and sprain his ankle, the other employee notified the
security who arranged for the employee transportation to
ER, where he was treated and released.
During investigation, the followings where noted:

• Lights out for about 3 days • Pathole about 4” deep

• Parking lot lights burned out • Pathole reported 4 weeks
• Employee left after dark ago
• Pathole repair work order
• No work order for lights in
submitted
system
• No action taken
WASAC
M. Tamer 128
 Exercise

Events Event Events

Employee walks Another Employee transported

Employee steps Employee
to car Employee to ER, treated and
in pothole sprains ankle notifies security released

Employee walks Another Employee transported

Employee steps
to car Incident Employee to ER, treated and
in pothole
notifies security released

Investigate – get details & evidence

WASAC
M. Tamer 129
 Step 1: Plan Investigation – Get Started!
Draw Initial SnapChart
Events Incident Events

Incident Event Event

Event Event
Employee walks Employee Another Employee transported
Employee steps
to car sprains Employee to ER, treated and
in pothole notifies
ankle released
security

Use to develop Questions or Activities required to fill the gaps

People involved? Paper for Activities?

Interview employees Parking lot
maintenance
Equipment/Facilities requests/records
(plant) involved?
Go see pothole Recordings?
Check visibility Check security videos
WASAC
M. Tamer 130
 Step 2: Determine Sequence of Events

Event
Incident Event Event
Event
Employee walks Employee steps
Employee Another Employee transported
to car sprains Employee to ER, treated and
in pothole notifies
ankle released
security

Interview and get Information

• Pathole about 4” deep
• Lights out for about 3 days
• Pathole reported 4
• Parking lot lights burned out
weeks ago
• Employee left after dark
• Pathole repair work
• No work order for lights in order submitted
system • No action taken

WASAC
M. Tamer 131
 Add Conditions to Related Events

Incident Event Event

Event Event
Employee Another Employee transported
Employee Employee steps
Employee
in pothole sprains to ER, treated and
walks to car notifies
ankle released
security

No action
After dark taken These are proven facts &
questions, NOT opinion,
Conditions

reported 4 judgements, etc.

Lights out
weeks ago
Amplifying info about an Event,
What? & Why?
Hole 4”
3 days?
deep

No work
order work order
submitted submitted
WASAC
M. Tamer 132
 Step 3: Define Causal Factors
Review the information
on the SnapChart and
identify Causal Factors

Note: These are NOT

Root Causes!

WASAC
133
M. Tamer
 Causal Factors

‘Who did what wrong”?

or
What equipment failed or didn’t work as
intended?

WASAC
M. Tamer 134
Ask the following questions:
•What error allowed a Hazard or allowed it to grow too large?
•What error allowed a Safeguard to fail?
•What error allowed a safeguard to be missing?
•What error allowed a target to get too close to the Hazard?
• What error allowed an incident (or its consequences) to
become worse after Hazard contacted the Target?

WASAC
M. Tamer 135
 Four Steps Method

Step 1: Identify all problems on

the SnapChrart
Step 2: Group related conditions
Caus near the Event that they impact
al Step 3: List the “So What
resulted because of this
Facto problem?” method to arrange
rs each group logically.
Step 4: Identify the Causal
Factor for each group and mark
it with a triangle.
 Four Steps Method

Starting with the first problem in each group’s “So What?” chain,
follow the chain until you come to the first Event or the incident,
then stop.
The causal factor will be either that first Event or one or two “So
Whats?” back from the Event/Incident.

The Causal Factor will be:

Who did what
wrong”?
or
What equipment failed
or didn’t work as
M. Tamer
WASAC
137
 Step 1. Identify all the problems on the
snapCharT

Simple SnapChart of sprained ankle incident

with three problems identified

WASAC
M. Tamer 138
 Causal Factors
Step 2: Group related conditions near the Event that they impact.

IF SEVERAL CONDITIONS DO NOT

ARE RELATED TO EACH OVER-GROUP
OTHER, GROUP THEM
TOGETHER UNDER THE
EVENT THAT THEY
IMPACT. CONDITIONS
SHOULD INCLUDE
PROBLEMS AND ANY
INFORMATION
DESCRIBING THE
PROBLEMS.

WASAC
M. Tamer 139
 Over Grouping

Step 2: Group related conditions near the

Event that they impact

SnapChart with additional inf. and more

problems identified

WASAC
M. Tamer 140
Investigator revises the Events into parallel
paths to better reflect the logic of what
happened and moves the conditions around
near events
Revised SnapCharT® oWf ASSpArCained
M. Tamer 141
Ankle Incident with Conditions Moved in
SnapCharT® of
Sprained Ankle
Incident with
Grouped Problems WASAC
M. Tamer 142
Causal
Factors
Step 3: List the “So What resulted
because of this problem?” method
to arrange each group logically.
Select a problem on the
SnapChart and ask “So What”
resulted because of this problem?
Ex. Construction supervisor didn’t
plan or supervise deliveries.
Interview identify that supervisor
was knowledgeable about which
roads on that site were rated for
heavy loads. He thought the driver
knows too.
 Causal Factors
So What?
Dump truck driver delivers heavy loads to the const. site by driving
across the parking lot that was not rated for heavy loads.
So What resulted from the dump truck driver delivering heavy
loads to the const. site by driving across the parking lot that was
not rated for heavy loads?
Heavy loads on the parking lot caused the beneath the asphalt to
shift and created the pothole.
So What?
The existence of the pothole allowed the employee to step in the
pothole.
From details to the BIG picture
WASAC
M. Tamer 144
 ”So-What” Chains

WASAC
M. Tamer 145
 ”So-What” Chains

WASAC
M. Tamer 146
 ”So-What” Chains

WASAC
M. Tamer 147
 Four Steps Method
Step 4:
Identify the Causal Factor for each group and mark it with a
triangle. The causal factor will be either that first Event or one
or two “So Whats?” back from the Event/Incident.

The Causal Factor will

be:
“Who did what
wrong”, “, or “What
was done wrong ?
or
What equipment failed
WASAC
or didn’t work as
M. Tamer 148
 Human error, so we can analyze to find
RC.

Stepping in a pothole isn’t a causal Factor, but a

random act; it’s not a human error; he didn’t WASAC
do
M. Tamer 149
anything wrong.
 WASAC
M. Tamer 150
Find R.C
RC Tree

WASAC
M. Tamer 151
 Level 1 – Top of the Tree

WASAC
M. Tamer 152
 WASAC
M. Tamer 153
 Find Root Causes
This troubleshooting guide helps the investigator identify which of
the seven human performance related Basic Cause Categories to
investigate further.

The seven categories are:

1. Procedures
2. Training
3. Quality Control
4. Communications
5. Management System
6. Human Engineering
7. Work Direction
WASAC
M. Tamer 154
 WASAC
M. Tamer 155
 Equipment Difficulty

WASAC
M. Tamer 156
 Analyzing a Causal
Factor
When the investigator identified a Human Performance Difficulty,
they were guided to a set of 15 questions (part of the tree's
embedded intelligence) called the Human Performance
Troubleshooting Guide.
The first of the 15 questions of the guide is shown in Figure
below.

First of the 15 Questions in the Human Performance Troubleshooting

Guide
WASAC
M. Tamer 157
 Human Performance Difficulty

WASAC
M. Tamer 158
 Human Performance Difficulty

WASAC
M. Tamer 159
 Human Performance Difficulty

WASAC
M. Tamer 160
 Human Performance Difficulty

WASAC
M. Tamer 161
 Human Performance Difficulty

WASAC
M. Tamer 162
 Human Performance Difficulty

WASAC
M. Tamer 163
 Human Performance Difficulty

WASAC
M. Tamer 164
 WASAC
M. Tamer 165
 WASAC
M. Tamer 166
 WASAC
M. Tamer 167
 WASAC
M. Tamer 168
 WASAC
M. Tamer 169
 WASAC
M. Tamer 170
 WASAC
M. Tamer 171
 WASAC
M. Tamer 172
Strep 5- Generic Causes

WASAC
M. Tamer 173
 WASAC
M. Tamer 174
 Generic
Root Cause:
Causes
Equipment Difficulty – Storage

Corrective action would include:

Store the pump in a controlled environment

Generic Cause:
Our inventory control system does not specify proper storage
of pumps:

Corrective action would include:

Develop a policy for proper storage of pumps and other critical
equipment

WASAC
M. Tamer 175
 Example Analysis Using
TapRooT®
Initial Incident Description
During a normal night shift at a process plant, fish were killed
when a temporary (temp) water treatment unit overheated and
released hot, low pH water to one of the plant‘s outfalls.
An investigation that included a contractor representative
(contract personnel were operating the temporary water
treatment unit) was conducted using the TapRooT® System.
The investigation found a sequence of events shown on a
SnapCharT® in Figure next.

WASAC
M. Tamer 176
 Example Analysis Using
TapRooT®

WASAC
M. Tamer 177
Example Analysis Using
TapRooT®
Results of Additional Investigation
After considerable investigation including:
 interviews with all contract operators and
their supervisor,
 discussions with the temporary water
treatment unit vendor's engineers,
 interviews with plant personnel at the
process plant unit,
 interviews with procurement
personnel, and
 interviews with operations
management,
 Complete SnapCharT

WASAC
M. Tamer 179
 Analyzing a Causal
Factor
To analyze the causal factor - contract operator falls
asleep - the investigator started at the top of the
TapRooT® Root Cause Tree® and worked down the
tree trough a process of selection and elimination.

The investigator thus asks and answers questions to

identify the specific root causes for the causal factor.
 Analyzing a Causal
Factor

In this case, the causal factor (contract operator falls asleep)

was identified as a Human Performance Difficulty (one of
the four major problem categories at the top of the Root
Cause Tree®) and the other three categories were
eliminated.

WASAC
M. Tamer 181
 Analyzing a Causal
Factor
When the investigator identified a Human Performance Difficulty,
they were guided to a set of 15 questions (part of the tree's
embedded intelligence) called the Human Performance
Troubleshooting Guide.
The first of the 15 questions of the guide is shown in Figure
below.

First of the 15 Questions in the Human Performance Troubleshooting

Guide
WASAC
M. Tamer 182
 Analyzing a Causal
Factorby a "Yes" answer to the questions in the
Each category indicated
Human Performance Troubleshooting Guide was investigated
further to see if it could be eliminated or if one or more Near-
Root Causes and related Root Causes contributed to the problem
(and thereby helped "cause" the incident).
One of the seven Basic Cause Categories (Human Engineering) is
shown in the Figure.

WASAC
M. Tamer 183
 Analyzing a Causal
Factor
The completed analysis of one of these categories (Human
Engineering)

WASAC
M. Tamer 184
 Analyzing a Causal
Factor
When this causal factor was analyzed using the rest of the
applicable Basic Cause Categories (not shown here - Work
Direction, Procedures, Management System) the following root
causes and generic causes were identified:

1. Monitoring alertness needs improvement.

2. Shift scheduling needs improvement.
3. Selection of fatigued worker.
4.The "no sleeping on the job" policy needs to have a
practical way to make it so that people can comply with it.

WASAC
M. Tamer 185
Developing Corrective
Actions

WASAC
M. Tamer 186
Once the causes for all of the causal factors were
identified, the investigator used the Corrective Action
Helper® module of the TapRooT® Software to help develop
the corrective actions for each of the root causes.
This module of the software helps investigators:
1. Verify that they are addressing the real causes of the
incident.
2.Develop corrective actions to fix the specific cause
of the problem.
Corrective
Actions
Developing Corrective Actions
3.Develop corrective actions for
the generic (or systemic) cause (if
applicable) for the problem.
4.Develop additional implementing
actions needed to make the corrective
actions successful.
5.Find references to study the
problem in detail and learn more
about potential strategies to eliminate
the problem.
WASAC
M. Tamer 188
 Corrective Actions

The following is an example of the guidance provided by the

Corrective Action Helper® module of the TapRooT® Software for
one of the root causes (Monitoring Alertness Needs
Improvement) that was identified for the Fish Kill Incident:

Check:
You have decided that the problem was related to loss of
performance over time while monitoring. (The job was too
boring.)

WASAC
M. Tamer 189
 Corrective Actions
Ideas:
1.You should consider recommending the following
options: (Order does not indicate preference.)
a.Provide an alarm to alert the worker and relieve the
boredom of monitoring.
b.Provide an automated monitoring and response system to
replace human monitoring and response. NOTE: this will probably
leave the worker in supervisory control. You will need to consider
ways to keep the worker informed as to what the automation is
doing and to clearly indicate why it is doing it.
You should also consider ways to keep the workers involved in the
process so that they maintain their situational awareness and
maintain their manual control proficiency.
WASAC
M. Tamer 190
 Corrective Actions

c. Rotate the person monitoring more frequently. (Experiment to

find out how long they can monitor reliably and then rotate
people so that they only monitor for less than that time.)

d. Redesign the job to provide other tasks that don't compete

with the monitoring task to keep the person alert and involved.
(For example, playing the radio while driving.) Do not provide
tasks that compete for the same resource. (For example, reading
a book while driving.)

WASAC
M. Tamer 191
 Corrective
Actions
e. Provide false signals to keep the worker involved. However, you
should also consider that people may ignore real signals if they
become accustomed to receiving only false signals.

f. Consult the workers to see if they have ideas that would make
the task more interesting without conflicting with the monitoring
requirements.

WASAC
M. Tamer 192
 2. Fatigue can also combine with
monitoring alertness problems.
Consider training supervisors to
understand that fatigued personnel
should not be assigned to tasks that
require a high degree of monitoring
alertness.
Correcti
ve
3. Also, consider testing individuals for
Actio their alertness before assigning them to
ns a monitoring task.

4. Once changes have been approved,

consider training the workers about the
changes and their intended impact.
 Corrective
Actions
Ideas for Generic Problems:
If monitoring alertness is a generic problem, consider
recommending a review of the jobs to redesign them and add
more active tasks.
Corrective Again, the Causal Factors were:
Actions
1. Fire hose ruptures
2.Contract Operator
Asleep (can't see readings
change)
3.Automatic shut-
off jumpered
4.Contract operator can't
hear alarm due to noise
 Corrective Actions
After reading all the Corrective Action Helper® Modules for all the
root causes that were discovered and after considering the
seriousness of each, the potential for future problems, and the
systemic (generic) nature of each cause, the following corrective
actions for all causal factors/root causes were developed.

1.Replace the old fire hose with a new, tested fire hose.
(Causal Factor 1)
2.Develop policy on testing and use of equipment in
temporary situations. (CF 1)

WASAC
M. Tamer 196
Corrective
Actions
2.Remove the jumpers and place the automatic trip feature back
in service. (CF 2, 3, & 4)

3.Update automatic trip feature with new module to

prevent spurious failures. (CF 2 & 3)

4.Negotiate contract revision so that contractor must notify

and get approval from the facility prior to disabling any alarm or
automatic safety feature. (CF3)

5.Move diesel driven compressor away from temporary water

treatment unit so that the alarm on the unit can be heard. (Cause 2
& 4)
Note that all causal factors
are addressed but some root
Correcti
causes were not corrected. ve
For example, the sleeping
policy and shiftwork
Actio
practices of the contractor ns
were not addressed as part
of this incident.
Investigators decided that a
single incident at a
temporary water treatment
unit was not enough
evidence to make sweeping
changes to contractor
policies and negotiated
contracts.

WASAC
M. Tamer 198
 Corrective Actions
However, these root causes will be trended in the facility's
database and if these types of problems repeat - even during
proactive audits, additional corrective action may be justified.

The same approach was agreed upon about the placement of

temporary equipment (diesel too close to water treatment unit).

WASAC
M. Tamer 199
Also, the corrective actions were
reviewed to ensure they were specific,
measurable, that someone was Correcti
accountable (no responsible people
were listed here), reasonable, timely (no ve
due dates were listed here and no
interim corrective actions were
Actions
provided for long term projects),
effective, and reviewed for
unanticipated consequences.

Before final approval by management of

the corrective actions, the corrective
actions could also be prioritized by use
of a risk ranking matrix or other
prioritization tool based on the risk
being addressed, the benefit and cost of
a particular corrective action.
 Corrective Actions
Finally, all lessons learned that could be applicable to other
company sites would be referred to the corporate lessons learned
clearing house.

Also, as time passed and data was accumulated, data from the
root causes would be reviewed to detect potential areas for
generic improvements and also reviewed to detect negative
trends or verify that improvement has occurred.

WASAC
M. Tamer 201
 Comparison of Results

A real incident similar to the Fish Kill

incident was reported in an industry
trade magazine. The magazine
reported that the contract operator
Correcti had been fired because they had
violated the company's no sleeping
ve policy.

Actions Compare the "fire the contract

operator” corrective action and its
effectiveness with the corrective
actions presented before.
 Corrective Actions
Comparison of Results

Firing the contract operator:

1. Is easy.
2. Provides an example to others that they need to be alert.
3. Is consistent with the company policy.
4.Seems effective in that no other operators are found
sleeping for several weeks after the contract operator is fired.

WASAC
M. Tamer 203
 Corrective Actions
However, what factors were missed and left uncorrected that
could contribute to future incidents?

1. No actions were taken to improve the equipment reliability

(either the reliability of the fire hose or of the automatic shutoff
and alarm).

WASAC
M. Tamer 204
 Corrective Actions
2. No effective corrective actions were taken to improve
monitoring alertness. At best, only a temporary improvement in
alertness was achieved.
In fact, the results of spot audits could be non-representative
because operators may be "covering” for each other to ensure
that no one else gets fired.

 The moving of the diesel so that the operator hears the alarm
and the fixing of the auto shut off feature makes the sleeping
problem doubtful.

WASAC
M. Tamer 205
 Corrective Actions
3. After a contract operator is fired, other operators will view
future investigations with suspicion and will be less likely to be
fully cooperative.

For example,
 Would an operator admit that they had nodded off?
 Would another operator "tell" on a fellow operator if he found
the other operator sleeping? or would they just "handle it on-
shift" and not tell anyone?
 Would covering up mistakes get in the way of effective learning
from mistakes?

WASAC
M. Tamer 206
 Corrective Actions
Even though:
- Root cause analysis using TapRooT® and developing corrective
actions is more difficult than blaming those involved, and -
TapRooT® suggests more thorough and potentially more difficult
to implement corrective actions than the easy "fire the
contractor" answer,

If the problem really needs to be solved to improve process safety,

industrial safety, quality, or productivity, then good root cause
analysis and implementing effective corrective actions can be
worthwhile.

WASAC
M. Tamer 207
 Defining Causal Factors for an
Incident with an
Equipment Failure
When an incident involves an equipment failure, the investigator
could recognize that the equipment related groups might be
“over-grouped” or “under-grouped” because the investigator
doesn’t have the inf. to understand why the equip. failure
happened. More inf. is needed.

How Equifactor can help define Causal Factors.

Ex. A unit shuts down because an essential pump fails. One of the
Event is:
Pumps failed due to failed bearing

WASAC
M. Tamer 208
 WASAC
M. Tamer 209
Defining Causal Factors for
an Incident with an
Equipment Failure

Starting with the first step of the Four Step method, we will find
Causal Factors:
STEP 1: Identify all problems on the SnapChart.
Using Equifactor Troubleshouting Table leads to the discovery of
additional Events and conditions to add to the SnapChart.
Ex. the first symptom will be excessive vibration from a failed
bearing, but failed bearing is only a symptom.
 Defining Causal Factors for an
Incident with an
Equipment
Using the centrifugal pump Failure
troubleshooting table under the
“vibration and noise” symptom, we see that excessive vibration
can have many causes including:
Suction problem – pump is cavitating
Suction problem – insufficient immersion of suction pipe or bell
Hydraulic system – total system head higher than design head of
pump
Mechanical system – unbalance pump
Mechanical system - misalignment
Mechanical system – casing distorted
from excessive pipe strain

WASAC
M. Tamer 211
 Defining Causal Factors for an
Incident with an
Equipment
Mechanical system – inadequate Failure
grouting of base
Mechanical system – bent shaft
Mechanical system – obstruction in lines or pump
housing
Mechanical system – mechanical defects – worn, rusted, defective
bearings
Mechanical system – unbalance driver
Mechanical system – motor troubles

WASAC
M. Tamer 212
 Defining Causal Factors for an
Incident with an
Equipment
Some of these are equipment Failure
oriented, some are design oriented,
some are installation oriented, and others are related to the
operation of the pump.
Before you can develop a Causal Factor, additional investigation is
needed to understand the Events that lead to the pump’s failure
by analyzing potential causes and process of elimination.
Based on expert knowledge you can eliminate non-potential
causes. Maintenance can disassemble the equipment while
collecting inf. to systematically eliminate items from the list until
the cause of failure is identified.

WASAC
M. Tamer 213
 Defining Causal Factors for an
Incident with an
Example: Equipment Failure
Investigator discovers that the pump had been run with a clogged
suction strainer until vibration and overheating had caused the
bearings to fail.
Where did the debris come from?
The debris looks like it came from a previous repair of a check
valve upstream from the strainer. Thus the Equifactor
Troubleshooting Tables helped the investigator add information to
the SnapChart including the new Events (Fig. below.24

WASAC
M. Tamer 214
 WASAC
M. Tamer 215
 Defining Causal Factors for an
Incident with an
Equipment
 Check valve replacement in progress Failure
 Debris collects in pipe during work
 Suction strainer clogs with debris
 And the new Conditions:
 Cleanliness control steps in work order not followed
 Didn’t remove debris from pipe after work was complete
 Didn’t detect clogged strainer
 No remote indication, alarm, or regular checks on strainer
differential pressure
 Didn’t detect hot, vibrating pump until after bearing
failure
 Supervisor did not inspect piping prior to close-out
 WASAC
No close-out inspection requirement
M. Tamer 216
 Defining Causal Factors for an
Incident with an
Equipment
Thus, new problems that were found outFailure
using the Equifactor are:
 Debris collects in pipe during work
 Cleanliness control steps in work order not followed
 Debris was not removed after work was complete
 Suction strainer clogs with debris
 Operations didn’t detect clogged suction strainer
 Operations didn’t detect hot, vibrating pump until after bearing
failure
 Supervisor did not inspect piping prior to close-out
 No close-out inspection requirement
 No remote indication, alarm, or regular checks on strainer
differential pressure
WASAC
M. Tamer 217
 STEP 2: Group related Conditions
Defini near the event that they impact.
ng Causal Three problem groups for the Pump
Fails Incident are shown in fig. next.
Factors for Organize the inf. in a cause chain
an using the “So What?” method – the
Incident third step in the Four Step Method for
with Defining Causal Factors.
Failure
a
Organize the inf. in a cause chain
n using the “So What?” method – the
Equipme third step in the Four Step Method for
nt Defining Causal Factors.
 WASAC
M. Tamer 219
Defining Causal
Factors for an
Incident with an
Equipment
Failure
STEP 3: Use the “So What?”
method to arrange each group
logically
Visually display the “So What?”
logic directly on the SnapChart in
fig. below
 WASAC
M. Tamer 221
Defining Causal Factors for an Incident with
an Equipment Failure
STEP 4: Identify the Causal Factor for each group and mark it
with a triangle.
Start with 1st. problem in each group’s “So What?” chain,
following the chain until you come to the first Event or the
Incident, then stop.
The Causal Factor will be either the Event or one or two “So
What?” back from the Event/Incident. The Causal Factor will be:
 Who did what wrong or what was done wrong?
 What equipment failed or did not work as intended?
We find four “So What?”Causal Factors as shown in fig. below

WASAC
M. Tamer 222
 WASAC
M. Tamer 223
Prepare A
Report
•Be objective!
•State facts.
•Assign
cause(s),
not blame.
•If referring
to an
individuals
actions,
don’t use
names in
the
recommen
dation.
•Good: WASAC
M. Tamer
All 224
 The Incident
Report
The incident report is designed to communicate the
investigation results to a wide audience.
The goal of the investigation is to prevent a similar
incident.
An exceptional investigation report willfully
explains the technical elements and issues
associated with the incident.
It will describe the management systems that
should have prevented the event and will detail the
system root causes associated with human errors
and other deficiencies involved in the incident.
 Report
Requirements
The report shall, at a minimum, include:
•Date and time of the incident.
•Date and time that the investigation began.
•A list of the investigation team members including members' job
titles.
•A description of the incident including a detailed chronological
sequence of events.
•An Emergency Responder Report of the tactical operations if
appropriate and useful.
•The factors that contributed to the incident.
•Recommendations resulting from the investigation.
Report
Requirements

WASAC
M. Tamer 227
 Executive
Summary
The purpose of the executive summary is to convey to the reader a
reasonable understanding of the accident, its causes, and the
actions necessary to prevent recurrence.
Typical executive summaries are one to five pages, depending on
the complexity of the accident.
The executive summary should include a brief account of:
 Essential facts pertaining to the occurrence and major
consequences (what happened)
 Conclusions that identify the causal factors, including
organizational, management systems, and line management
oversight deficiencies, that allowed the accident to
happen (why it happened) .

WASAC
M. Tamer 228
 Executive
Summary
The executive summary should not include a laundry list
of all the facts, conclusions, and recommendations.
Rather, to be effective, it should summarize the
important facts; causal factors; conclusions; and
recommendations.
In other words, if this was the only part of the report
that was read, what are the three or four most
important things you want the reader to come away
with? 23
 Executive
Introduction
Summary
A fatality was investigated in which a construction
subcontractor fell from a temporary platform in the [Facility]
at the [Site]. In conducting its investigation, the Accident
Investigation Team used TapRoot technique.
The Team inspected and videotaped the accident site,
reviewed events surrounding the accident, conducted
extensive interviews and document reviews

Accident Description The accident occurred at approximately [Time] on [Date] at

the [Facility], when a construction worker, employed by
[Subcontractor], fell from a temporary platform.
The worker was transported by ----

Direct And Root Causes The direct cause of the accident was the fall from an
unprotected platform.
The contributing causes of the accident were: (1) ---
Conclusions And summarized in Table 1
Recommendations

WASAC
M. Tamer 230
 Developing Recommendations
Ineffective recommendations may only serve to
transfer the hazard or even create a new hazard that
was not present before the initial incident.

A tie should exist between the facility’s incident

investigation management system and their
management of change (MOC) program.

WASAC
M. Tamer 231
 Accident Investigation Startup
Activities List
Name of Designated Lead
Description of Activity
HQ Site Other

WASAC
M. Tamer 232
GOOD
LUCK

WASAC
M. Tamer 233
 WASAC
M. Tamer 234
 WASAC
M. Tamer 235
 WASAC
M. Tamer 236
GOOD
LUCK