Professional Documents
Culture Documents
A Joint Publication of the American Institute of Chemical Engineers and John Wiley &
Sons, Inc.
The rights of CCPS to be identified as the author of the editorial material in this work
have been asserted in accordance with law.
Registered Office
John Wiley & Sons, Inc., 111 River Street, Hoboken, NJ 07030, USA
Editorial Office
111 River Street, Hoboken, NJ 07030, USA
For details of our global editorial offices, customer services, and more information
about Wiley products visit us at www.wiley.com.
Wiley also publishes its books in a variety of electronic formats and by print-on-demand.
Some content that appears in standard print versions of this book may not be available
in other formats.
10 9 8 7 6 5 4 3 2 1
This book is one in a series of process safety guidelines and concept books
published by the Center for Chemical Process Safety (CCPS). Refer to
www.wiley.com/go/ccps for full list of titles in this series.
It is sincerely hoped that the information presented in this document will lead to
a better safety record for the entire industry; however, neither the American
Institute of Chemical Engineers, its consultants, CCPS Technical Steering
Committee and Subcommittee members, their employers, their employers'
officers and directors, nor Greenstreet Berman, Ltd., and its employees and
subcontractors warrant or represent, expressly or by implication, the correctness
or accuracy of the content of the information presented in this document. As
between (1) American Institute of Chemical Engineers, its consultants, CCPS
Technical Steering Committee and Subcommittee members, their employers, their
employers' officers and directors, and Greenstreet Berman, Ltd., and its
employees and subcontractors, and (2) the user of this document, the user accepts
any legal liability or responsibility whatsoever for the consequence of its use or
misuse.
Human Factors Handbook for Process Plant Operations
is dedicated to
Jack L. McCavit
Jack is passionate about process safety, especially in the areas of culture and
human factors. His work, both in his career at Celanese, and after his retirement,
has concentrated on educating workers and industry leaders on the importance
of process safety, the payback of sustaining a great program, and most
importantly, the impact of not making process safety a top priority. Jack had first-
hand experience with the latter when he witnessed a butane vapor cloud explosion
at the Celanese site in Pampa, Texas, in 1987, resulting in three fatalities and
dozens of injuries. Based on his significant and relevant expertise, Jack was
selected as the technical manager for the prominent Baker Panel investigation of
the BP Texas City Explosion in 2005.
Jack is a CCPS Fellow, an AIChE Fellow, and is rumored to be the fifth most famous
Texan in history. He was the committee chair for the CCPS flagship book,
Guidelines for Risk Based Process Safety, and a driving force behind CCPS’s Vision
20/20.
Part 3: Equipment............................................................................................ 91
9 Human Factors in equipment design ..................................................... 93
9.1 Learning objectives of this Chapter ............................................ 93
9.2 Definitions ...................................................................................... 93
9.3 Major accident example ............................................................... 94
Table of Contents xi
18.3 Why do we fail to capture, challenge, and correct errors? ... 217
18.4 Coaching people to recognize risk of making errors ............. 218
18.5 Error Management Training ...................................................... 220
18.6 Enabling challenge of task performance ................................. 224
18.7 Key learning points from this Chapter..................................... 231
19 Communicating information and instructions ................................... 233
19.1 Learning objectives of this Chapter ......................................... 233
19.2 Incident example ........................................................................ 233
19.3 Causes of poor communication ............................................... 234
19.4 Human Factors of communications ......................................... 235
19.5 Avoiding communication overload .......................................... 237
19.6 Human Factors in shift handover ............................................. 241
19.7 Key learning points from this Chapter..................................... 245
APPENDICES
Index................................................................................................................ 437
List of Figures
Accident: An event that can cause (or has caused) significant harm to workers, the
environment, property, and the surrounding community.
Anthropometrics: The science of measuring the size and proportions of the
human body (called anthropometry), especially as applied to the design of
furniture and machines.
Behavioral marker: Non-technical behaviors that can be observed and described.
They refer to a prescribed set of behaviors and are indicative of specific types of
non-technical skills performance (e.g., effective decision-making in emergencies)
within a work environment.
Cognitive overload: A mental state where an individual is unable to process all
the information provided by the system.
Cognitive underload: A mental state when an individual is under-stimulated due
to insufficient workload. This mental state leads to lack of attention.
Competency Assessment: System which allows measuring and documenting
personnel competency. The goal of competency assessment is to identify
problems with employee performance, and to correct these issues before they
affect performance.
Competency: Set of skills and knowledge which enables a person to perform tasks
efficiently, reliably and safely to a defined standard.
Competency Gap: Difference between the current competency level and the
required competency level of an employee.
Competency Management: Method of categorizing and tracking the
development of individual employee competency, allowing an organization to
track progress, and identify future training needs.
Fatigue: Fatigue is a decline in physical and/or mental performance.
Hold Points: Point where change cannot happen until there has been verification
that the prerequisites have been achieved.
Human Error: Intended or unintended human action or inaction that produces an
unintended result. This includes, but is not limited to, actions by designers,
operators, planners/schedulers, maintainers, engineers or managers that may
contribute to or result in accidents [1].
Human Factors: Discipline concerned with designing machines, operations, and
work environments so they match human capabilities, limitations, and needs. This
includes any technical work (engineering, procedure writing, worker training,
worker selection, operations, maintenance, etc.) related to the human interface in
human-machine systems [1].
Human Performance: Measure of an individual’s ability to execute a task
effectively.
Incident: Event, or series of events, resulting in one or more undesirable
consequences, such as harm to people, damage to the environment, or
asset/business losses.
xxiv Human Factors Handbook
Job aid: Specific information or material intended to help workers execute a task
more effectively.
Learning: Acquisition of knowledge or skills through study, experience, or being
taught.
Major accident: Major accident means an occurrence such as a major emission,
fire, or explosion resulting from uncontrolled developments in the course of the
operation of any establishment, and leading to serious danger to human health or
the environment (whether immediate or delayed) inside or outside the
establishment, and involving one or more dangerous substances [2].
Mistake: A decision or judgement that is misguided.
Non-technical skills: The cognitive, social, and personal resource skills that
complement technical skills and contribute to safe and efficient task execution [3].
Performance Influencing Factors (PIFs): Characteristics of the job, the individual
and the organization that influence human performance [4].
Performance standards: Description of how the job is a description of what
(actions/tasks) needs to be taken/executed, how the job must be done
(behaviors/methods) and outcomes/results that will define satisfactory or
acceptable performance.
Psychological safety: The outcome of an open workplace culture where people
are willing to express an opinion, or admit mistakes or unsafe behaviors, without
fear of being embarrassed, rejected, or punished.
Root cause: Fundamental, underlying, system-related reason why an incident
occurred that identifies a correctable failure(s) in management systems. There is
typically more than one root cause for every process safety incident.
Rota: A period of work taken in rotation with other workers (an abbreviation of
rotation).
Rotation: A period of work taken in rotation with other workers.
Shift working (shifts): Work which takes place on a schedule outside traditional
day work hours. It can involve evening or night shifts, early morning shifts, and
rotating shifts.
Training: “Practical instruction in job and task requirements and methods.
Training may be provided in a classroom or at the workplace, and its objectives are
to enable workers to meet some minimum initial performance standards
(minimum required competency level), maintain their proficiency, or to qualify
them for promotion to a more demanding position” [5].
Vigilance decrement: Decline in “the ability to sustain attention and remain alert
to a particular stimulus over a prolonged period of time” [6].
Acronyms
Acronym Meaning
ANP Agência Nacional do Petróleo (Brazil Petroleum Regulator)
BP British Petroleum
CCPS Center for Chemical Process Safety
CK Checklist
CSB Chemical Safety Board
CRM Crew Resource Management
DCS Distributed Control System
DFC Diagnostic Flow Charts
DIF Difficulty, Importance and Frequency Analysis
DOE Department of Energy
DT Decision Tree or Diagnostic Tree
EEMUA Engineering Equipment and Materials Users Association
FCCU Fluidized catalytic cracking unit
GC Grab Card
GUI Graphical User Interface
HIRA Hazard Identification and Risk Analysis
ICAO International Civil Aviation Organization
IChemE Institute of Chemical Engineers
IOGP International Association of Oil and Gas Producers
ISO International Standards Institute
ISOM Isomerization
LEL Lower Explosive Level
LFL Lower Flammability Level
LOPA Layers of Protection Analysis
MDMT Minimum design metal temperature
MEB Material and Energy Balance
MOC Management of Change
NATO North Atlantic Treaty Organization
xxvi Human Factors Handbook
Acronym Meaning
OIM Offshore Installation Manager
OSHA Occupational Safety and Health Agency
PFD Process Flow Diagram
P&ID Piping and Instrumentation Diagrams
PSB Plant Status Boards
PSI Process Safety Information
PSV Pressure Safety Valve
PTW Permit to Work
RBPS Risk Based Process Safety
SCTA Safety Critical Task Analysis
SH Shift Handover
SOP Standard Operating Procedure
SRK Skills, Rule and Knowledge
STAR Stop Think Act and Review
QRA Quantitative Risk Analysis
WI Work Instruction
UK United Kingdom
U.S. United States
Acknowledgements
The American Institute of Chemical Engineers (AIChE) and the Center for Chemical
Process Safety (CCPS) express their gratitude to all the members of the Human
Factors Handbook for Plant Operations Project Team and their member
companies for their generous efforts and technical contributions. The committee
structure for this concept book differs from other CCPS books in that this was a
project done in collaboration with the Energy Institute (EI) and the generous efforts
and technical contributions of the EI Technical Partner and Technical Company
members is also gratefully acknowledged.
The writers from the Human Factors consultancy Greenstreet Berman Ltd are
also acknowledged, especially the principal writers Michael Wright and Dr. Ludmila
Musalova, with additional inputs from David Pennie, Rebecca Canham and
Ninoslava Shah.
Gabriela Dutra (ex Braskem), Sahika Korkmaz (ex Chevron) and Josué Eduardo
Maia França (Petrobras) also contributed to certain stages of the project.
Before publication, all CCPS and EI books are subjected to a thorough peer
review process. CCPS and EI gratefully acknowledge the thoughtful comments and
suggestions of the peer reviewers. Their work enhanced the accuracy and clarity
of this concept book. The peer reviewers have provided many constructive
comments and suggestions. They were not asked to endorse this book and were
not shown the final manuscript before its release.
Peer Reviewers
Linda Bellamy White Queen BV
Michelle Brown FMC
Denise Chastain-Knight Exida
Palani Chidambaram DSS
Ed Corbett UK Health and Safety Executive
David Cummings DuPont
Rhona Flin Aberdeen University
Jerry Forest Celanese
Jeff Fox CCPS Emeritus, ex Dow
Osvaldo Fuente Dow
SP Garg GAIL
Zsuzsanna Gynes The Institution of Chemical Engineers
John Herber CCPS Emeritus
Alison Knight 3M
Susan Lee Marathon
Maria Chiara Leva TU Dublin
Keith Mayer Kraton Polymers
Rob Miles Hu-Tech
Chelsea Miller Chevron
Raphael Moura ANP
Cathy Pincus ExxonMobil
Tim Thompson Braskem
Elliot Wolf Chemours
Neal Yeomans Advansix
The affiliations of writers, project team members and peer reviewers were
correct at the time of publication.
Foreword
Humans are resourceful, resilient, innovative, smart creatures. They can also be
error-prone – forgetting to complete a step in a sequence, misunderstanding
instructions, making mistakes in task execution. Disentangling these strengths and
limitations, determining how and why human performance can be both resilient
and fragile is the science of human factors.
The military and aviation sectors were the first to appreciate that the design of
equipment and task environments had to take into account the psychological,
anatomical and physiological capabilities of the human operators. The influential
role of the organizational culture and its component systems on both managers
and workers also became apparent. As the hybrid blend of engineers,
psychologists, designers and other human factors specialists began to coalesce in
the late 1940s, professional human factors and ergonomics societies were formed,
helping to systematize an established body of evidence relating to human factors
science, with a range of accepted methods for investigation and intervention. But
it has taken some time for the value of this approach for the management of
workplace operations to be recognized across industrial sectors.
More than two decades later, at the time of this book’s publication, awareness
and understanding of the factors influencing human performance in the process
industries has become more active. This volume, one of a series directed by the
Center for Chemical Process Safety, reflects the increased activity in the process
industries. It provides an essential handbook for people on the frontline of plant
operations, helping them apply good human factors principles and knowledge
with practical techniques.
Practical tools and techniques are provided for each topic area with guidance
for application and more experienced practitioners will discover new ideas for
their portfolio of Human Factors methods.
Rhona Flin
Professor of Industrial Psychology
Aberdeen Business School
Robert Gordon University
Human Factors Handbook For Process Plant Operations: Improving Process Safety and System
Performance CCPS.
© 2022 CCPS. Published 2022 The American Institute of Chemical Engineers.
1 Introduction
Human Factors also provides a set of principles and concepts that can be used
to guide day-to-day decisions. The decisions focus on how best to support
successful human performance. This approach helps people to understand tasks
from the perspective of the person doing the work and provides ideas on how to
support people to perform better. It advocates an orientation (a way of thinking)
towards making improvements that support human performance and the
prevention of error. It recognizes people’s capabilities and commitment, and it
aims to maximize people’s roles in safe and productive operations, and to build
their ability to cope mentally and emotionally with stressful and demanding tasks,
i.e., psychological resilience.
Human Factors covers a very wide range of topics including, training, work
planning, and fatigue. Many of these topics come under existing management
systems, such as the operation of rotating shift schedule systems, and training
systems. Human Factors provides knowledge, tools, and insights that can be
integrated into an organization’s existing systems of work and operational
management, safety assessments, incident investigations, and day-to-day
operational decision-making. In this book, the terms ‘incident’ and ‘accident’ will be
used interchangeably.
The handbook:
This handbook can be read in conjunction with other CCPS guidance on safety
culture and process safety management, including:
Some of the elements within “Guidelines for Risk Based Process Safety” are
relevant to this handbook. Therefore, they have been referenced at various points
throughout the handbook as additional information where this would be helpful
to the reader.
• Frontline supervisors.
• Designers.
• Operations and maintenance managers.
• Plant superintendents.
• Process engineers.
• Project managers.
• Construction managers.
• Process safety and health and safety personnel with the role of coaching
higher-level managers on Human Factors aspects.
The handbook is intended for people who understand process operations and
have some process safety management experience.
The explanation of some topics has been intentionally simplified and phrased in
normal everyday language, rather than in scientific terms. This has been done in
order to make the document more accessible, readable and more usable in the
practical domain, and also with the aim of making it more understandable for an
international audience.
1. Introduction 7
For example, the term ‘mistake’ is used in this book to refer to both mistakes
and other kinds of error, even though human factors specialists commonly
understand the term ‘mistake’ to mean a specific kind of error that is to do with
judgement and decision-making, as distinct from other kinds of error such as ‘slips
and lapses’. The term 'mistake' is used generally in the book, but where specific
types of error are being discussed then the specific appropriate terms are used
where that aids clarity.
In those cases where obvious signs of poor Human Factors were found,
stakeholder confidence in the company was greatly reduced and employee morale
was destroyed.
The United States Chemical Safety and Hazard Investigation Board (CSB)
investigation of the Texas City accident cited that previous accidents have shown
that Human Factors plays a role in industrial accidents [14]. The Texas City event
includes several examples of Human Factors. People had worked without rest for
many weeks or worked excessively long days. In some cases, it was known that
process instrumentation was unreliable or that critical information such as Piping
and Instrumentation Diagrams were out of date, and that training on new control
systems had not been provided.
Human Factors is more than common sense. People may make mistakes for many
reasons. Many factors influence how people perform. Process operations can be
complex and involve many difficult tasks. Technology is constantly changing.
8 Human Factors Handbook
In a dynamic process environment, with many complex tasks and safety critical
operations in flux, a high level of human performance needs to be achieved
systematically. Process safety does not depend on a single person’s view of what
is “common sense”. Recognized and implemented good practice and guidance is
necessary to achieve a high standard of human performance.
An overview of the handbook is given in Figure 1-2. The handbook is loosely split
into eight parts. Each Chapter can be read by itself. It is, however, useful to read
Chapters 2 to 4 first.
The other Chapters can be grouped into Parts around a set of core topics, such
as Part 2 job aids, Part 3 equipment design, Part 4 competence, Part 5 task support,
Part 6 non-technical skills and Part 7 covering working with contractors and
managing change.
Non-technical skills
Procedures and job aids Operational competence Task support
20. Situation awareness and
5. Human performance 10. Human performance and 15. Fatigue and staffing levels
agile thinking
and job aids operational competency 16. Task planning and error
21. Fostering situation
6. Selecting a type of job 11. Determining operational assessment
awareness and agile
aid competency requirements 17. Error management in task
thinking
7. Developing content of 12. Identifying learning planning, preparation, and
22. Human Factors in
a job aid requirements control
emergencies
8. Format and design of 13. Operational competency 18. Capturing, challenging,
job aids development and correcting operational
14. Operational competency error
assessment 19. Communicating
information and
instructions
9. Human Factors in
Equipment design
This chapter provides some Human Factors principles that will help people to
reduce the likelihood of errors and mistakes. By the end of this chapter, the reader
should be able to:
The “Miracle on the Hudson” happened on January 15, 2009, when a bird strike
occurred shortly after US Airways flight 1549 took off from New York’s LaGuardia
airport [15]. The Airbus struck a flock of Canada geese while on the climb from the
airport. The Captain, Chesley Sullenberger, and First Officer Jeff Skiles decided to
ditch (emergency water landing) the aircraft in the Hudson River, saving all on
board. This famous event was portrayed in a 2016 film (Sully) starring Tom Hanks
as Chesley “Sully” Sullenberger. The successful ditching of an unpowered
passenger airline onto the Hudson River, within six minutes of the bird strike, when
both engines had failed, is an example of skilled and knowledgeable human
performance.
Following the bird strike, a very short period of time was available for the pilots
to determine what had happened, enact a Mayday, determine they could not
return to the airport, decide they had to glide around and find an alternative
landing site (the Hudson River), and identify a new course. They achieved this and
ditched on the Hudson River, after which the 150 passengers and five crew were
rescued by nearby boats and ferries.
The normal procedure for dual engine failure was to attempt to return to the
airport. This turn back to the airport was not possible at the plane’s low altitude. It
was also not possible to complete a “dual engine failure” checklist due to the
limited time available prior to ditching. Simulator training did not cover ditching.
12 Human Factors Handbook
The Captain and First Officer needed to make decisions quickly based on their
knowledge and judgment.
2.2.2 How did they perform successfully?
Despite the popular title of “Miracle on the Hudson”, the successful unpowered
ditching did not happen by luck.
First, airline pilots receive a high level of training in piloting, annual simulator-
based training in handling emergencies, and training in Crew Resource
Management (CRM). CRM provides training in understanding human performance,
interpersonal skills, communications, leadership and decision-making. This
includes maintaining situation awareness and making decisions in high stress
emergencies. The National Transportation Safety Board investigation report
(NTSB, 2010) [15] stated:
“The captain credited the US Airways CRM training for providing him and the
first officer with the skills and tools that they needed to build a team quickly and
open lines of communication, share common goals, and work together.” (p61)
Second, the plane’s “fly by wire” design meant that after the pilot changed
course, the computers adjusted the flight control to maintain plane stability. This
allowed the crew to focus on emergency decision-making. The system design
reduced the crew workload.
Third, having two pilots also allowed them to multi-task as a team and to check
each other’s judgments and actions.
2. Human performance and error 13
The “Miracle on the Hudson” shows that with training and experience, people
are able to carry out complex tasks reliably and accurately. With education on how
a system works, people can use their knowledge and experience to quickly come
up with ways to handle new situations. Training in decision-making helps people
make better judgments and decisions, and to act quickly. While the application of
a Human Factors approach can prevent many errors and mistakes, it should
greatly improve human performance and reduce the potential for unrecoverable
errors.
In 2005, a major explosion occurred at the BP refinery in Texas City, United States
of America. A summary of the accident is given in B.1 (page 383) and the ‘The B.P.
U.S. Refineries Independent Safety Review Panel’ provides a very detailed report
of this accident [16].
The CSB [14] 2007 investigation report stated people were “set up to fail”. Some
points from the CSB investigation which highlight Human Factors issues include:
There were many factors influencing the operational decisions and actions.
Deficiencies in each of these factors combined to exacerbate operational
problems.
“You cannot change the human condition, but you can change the conditions in
which people work.” Professor James Reason (Chapter 7, page 96) [17].
A Human Factors principle is that it is vital to ask how and why errors occur.
This includes asking:
Environment-
related
• Temperature,
humidity,
ventilation
• Noise
Person-related • Lighting Equipment-
• Attitudes & • Space related
behaviors
• Layout
• Training &
• Fit for purpose
experience
• Accessibility
• Capabilities
• Complexity
• Relationships
Performance
influencing factors
Information-
related
Job-related
• Clarity
• Fatigue
• Information
overload • Organizational
Task-related stressors
• Accuracy
• Situational stressors • Workload
• Completeness
• Availability • Distractions
• Multi-tasking
• Complexity
• Time available
• Task frequency/
duration
• Workload
2. Human performance and error 17
This Chapter provides some key principles for identifying options to support
human performance. The later Chapters provide advice on these options.
In order to support human performance and reduce the potential for errors
and mistakes, it is important to understand the nature of the tasks, the type of
human performance required, and the causes of possible error and mistakes.
The Skill, Rule and Knowledge-based model (SRK) is a commonly used way of
thinking about how people perform. The idea is that people perform differently
according to the type of task they are doing and their familiarity with the task. The
SRK performance model is shown in Figure 3-1.
This model has been used since the 1980’s to help identify ways to support
people in performing process operations tasks. It was originally proposed by
Professor Jens Rasmussen [18] and further developed by other researchers such
as Professor James Reason [19].
Rule based tasks tend to require following a procedure, assessing the situation
(situation awareness), decision-making, and experience related to carrying out the
procedure. An example is identifying a fault in a car engine that is explained in a
manual.
20 Human Factors Handbook
The words “rule” and “procedure” are not meant to be used literally. A written
procedure may not exist. These words are used to mean the right set of actions
are known in advance of performing a task and it has been determined when these
actions should be performed.
Skill-based performance means people can reliably and quickly perform tasks
with a low level of conscious mental effort, as the actions are so well practiced they
do not require much thought. Steering a car is a typical example. This may also
include being able to gather and understand information very quickly, such as
reading process instrumentation.
• Quickly carrying out control actions, without using much thought, such as
operating the controls, would use a skill-based performance.
Different tasks will require different levels of each type of performance ability.
Table 3-1 provides definitions and examples.
Different types of errors or mistakes are linked to the three types of human
performance discussed in 3.2.
Lapses of attention occur when a person This may happen when someone
loses their place in a series of actions (for incorrectly interprets what is
example, steps in a procedure), skipping a happening or selects a wrong
step due to a distraction or a gap in course of action.
memory. The intention is correct and the
knowledge may be right, but one or more
steps are missed.
Table 3-1: SRK types of human performance
Type of human
Definition Examples
performance
Knowledge- When a task is new or complicated, a person will need to • Diagnosing a process upset
based task pay a lot of attention to what they are doing and decide • Working out how to perform a rare
performance what to do based on general knowledge and experience. maintenance task
When a person can perform a task to a high level of • Pressing an accelerator pedal to maintain
Skill-based
accuracy and reliability with a low level of attention, the steady speed of a chemical road tanker
performance
performance is based on skill. • Applying the right amount of torque to a bolt
Figure 3-2: Human performance modes, errors and mistakes
Type of human
Type of error or mistake Example
performance
Skill based
Further The Esso Longford gas plant accident. Report of the Longford
reading Royal Commission. June 1999 [20]
3. Options for supporting human performance 25
A knowledge-based task, such as diagnosing a rare process upset, may fail due to
a lack of knowledge of the process or a lack of familiarity with the rare event. The
likelihood of this failure may be greater if training focused on how to carry out the
procedure (procedural instruction) and did not provide detailed knowledge of
what is happening during the process (the underpinning knowledge of the
process).
It is also possible the people with the knowledge may not be available when
they are needed. For example, if senior engineers only work on day shifts, and
night shifts rely on operators; or if specialist engineers work from a central shared
facility or from a remote location.
Ensuring process safety information, such as P&IDs, are accurate and available
will help people understand what is happening and make correct decisions.
Too much information may make it harder for someone to understand an event
and to use their knowledge, especially if it includes unimportant or irrelevant
information.
Designing control information and alarms such that their intents, priorities, and
relationships are clear can set up the operator for success.
26 Human Factors Handbook
Time is needed to process information, consider the information and put together
an opinion. If the time to perform the task is short, this may not allow someone to
develop an opinion, especially if it is a complicated matter.
Some options for helping people include, having more people available to
assess the information, simplifying the task such as by prioritizing shut down
actions, adjusting alarm set points to increase operator response time, or
simplifying the information.
As with all tasks, tiredness (fatigue) and situation or organizational stress can make
it harder for a person to think or to remember what they know. Being interrupted
or distracted from a task, can also take attention away from decision-making.
Shift systems and staffing levels should be designed to avoid fatigue. Having
sufficient staff to enable the management of overtime for shift workers reduces
the likelihood that fatigue will affect performance. Task design and workspace
design should minimize distractions.
If the task is infrequent or must be performed quickly, a person may rely too much
on their knowledge to make decisions or judgments or they may quickly improvise
plans of action, instead of recalling instructions and procedures. A person may
incorrectly assume that the task can be done in the same way as a similar task, or
they may not know what the correct task steps are.
28 Human Factors Handbook
Even when an individual has the right knowledge, skills, and experience to do a
task properly errors can still occur. Skill-based errors tend to occur during highly
routine activities, when attention is diverted from a task, either by thoughts or
external factors. Table 3-4 provides an example of a skill-based error, in this case,
a lapse. The example comes from the same accident in section 3.3.4, the Formosa
Plastics Vinyl Chloride Monomer Explosion in 2004 [21].
3. Options for supporting human performance 29
During skill-based tasks, people pay less attention to what they are doing, because
they are skilled and can perform the task without focused thought. As such,
attention can lapse even without distractions or interruptions, causing a person to
leave out or skip a step, or to make a slip without noticing. If they have not had
sufficient time or experience to practice an infrequent task, they may not be able
to perform it reliably and accurately.
Even an experienced and capable person may suffer a lapse due to:
• If the task is long, it is easy for a person to forget their place in the task
and/or forget the sub-steps to be carried out, especially where there are
no job aids. A person has to accurately remember everything.
• If two tasks are very similar or use similar equipment, a person may
carry out the wrong actions accidentally and without knowing.
• If a task is repetitive, it is easy to miss a step.
• If a task is not carried out very often, it is harder to be accurate and
reliable, as it is not as well practiced.
30 Human Factors Handbook
The design of equipment and the work environment can also create lapses. For
example:
• If two similar controls are close to one another, the wrong control may
be used.
• In a noisy environment, or if heavy respiratory protection must be used,
verbal communication may be misheard.
• Information displays can be confusing and hard to read.
The types of support for knowledge-based and rule-based tasks are similar. These
are shown in Figure 3-3 and summarized next.
3.4.2.1 Knowledge
Information, Diagnostic,
Education in
schematics, communication &
process, system,
decision-making decision-making
faults & hazards
aids & procedures skills
Teamwork, shared
Task & team situation
Workload &
design awareness, co-
fatigue
(distractions & ordination, clear
management
interruptions) roles &
responsibilities
Up-to-date procedures and job aids can show the See Chapters 5,
circumstances and conditions where a sequence of actions 1, 7, and 8 for
should be used – it will also outline what these actions are. A more
logical step-by-step guide or list of clear instructions can help information on
with understanding and carrying out these actions, especially job aids.
when a person has had previous training and experience.
These should be designed to be practical and meaningful to operators as noted in
the CCPS “Guidelines for Risk Based Process Safety” as per the Operating
Procedures element [5].
Training and operational experience can help people See Chapters 10, 11,
12, 13 and 14 for more
to remember and use their process and procedural
information on training
knowledge. This is part of the ‘Training and and performance
Performance Assurance’ element of the CCPS assessment.
“Guidelines for Risk Based Process Safety” [5].
32 Human Factors Handbook
To be able to use their knowledge and procedures, a person should have real time
information on the process operations. Accurate and complete information should
be available to help operators correctly assess a situation and choose an
appropriate procedure or action. This can include
See Chapters 5 - 9 for
information from process indicators, labelling, visual more information on
inspection of processes, schematics such as Piping and job aids.
Instrumentation diagrams, equipment labels, and
decision-making aids such as diagnostic flow charts.
This is part of the ‘Process Knowledge Management” element of the CCPS
“Guidelines for Risk Based Process Safety” [5].
3.4.2.6 Teamwork
Skill
development, Procedures,
Task planning &
instruction & instructions &
checking
operational memory aids
experience
Task,
Workload & environment & Controls &
fatigue team design instrumentation
management (distractions & design
interruptions)
Procedures, instructions and job aids can all help people See Chapter 22
remember the correct steps and check their progress (which for more
steps they have done) when working on long and complex information on
tasks. Double-checking and peer checking task completion task verification.
can help people spot and correct slips and lapses.
See Chapter 16
Attention can decline when carrying out skill-based tasks
for more
and people can be distracted. Tasks and working information on
environments should be designed to help minimize task planning.
interruptions, which helps to minimize distractions.
Controls and equipment can be designed to help avoid See Chapter 9 for
skill-based slips, such as placing the most frequently used more information
controls closer to the operator and ensuring information on equipment
displays are readable. design.
34 Human Factors Handbook
4.2 Attention
• Minimize distractions.
• Minimize late information.
• Training to recognize deviations and drift.
• Developing cognitive skills to be aware of personal tendencies and drift.
• Training staff to focus on relevant information (Chapter 13).
• Avoiding alarm overload by alarm prioritization.
• Developing psychological skills (Chapters 21 and 22).
4.3 Vigilance
Task design and easy to read instrumentation (at eye level and See Chapter 16
center of visual field) helps to ensure high levels of vigilance. for more
This includes designing tasks so that operators are not information on
required to remain vigilant for long periods without a break. task planning.
4.4 Memory
Carrying out long, occasional or complex tasks can be helped by using procedures
and job aids, and by clear task organization, as discussed in Chapters 5 - 8. These
methods can reduce reliance on memory to remember the right procedure and
the sequence of actions to be taken, therefore reducing the risk of mistakes.
Cognition is how a person uses their mind to gain knowledge and understanding.
Cognitive capacity is the amount of information, decisions and judgments a person
can hold and process in their mind at any one moment.
Chapter 15 discusses how to make sure that the right number of people are
available for a task. It also explores “real time” systems for managing fatigue, i.e.,
helping to spot tired people at work. Changes in staffing levels, shift systems and
workloads should be managed, as covered in Chapter 15.
4. Supporting human capabilities 39
Although cognitive heuristics are useful and important ways to make decisions,
these “rules of thumb” can also cause mistakes. By taking a short cut when making
a judgment, some information will be missed, which means the right option may
not be considered. This may be even more likely if a person feels they are under
time pressure or are feeling fatigued, demotivated or bored.
However, they can have a negative impact on judgment, thinking and decision-
making. In a safety critical setting, it is vital to recognize the potential for these
mental shortcuts becoming inaccurate bias and to correct this.
40 Human Factors Handbook
Confirmation bias
Authority bias
• After a person identifies a possible
• An operator has too much
cause of a process upset, he/she then
confidence in the opinion of those
looks for information to support their
in authority, while not trusting their
opinion, while ignoring information
own feelings.
that suggests a different cause.
• The person goes with their original
• The person goes with a senior
engineer’s decision despite feeling
opinion, despite some information
that the judgment is wrong.
suggesting they are wrong.
People can develop and effectively apply their cognitive skills by use of
techniques such as “20 second scans”. People are trained and directed to pause
before starting a task and to scan the work site to identify anything unusual or
unexpected. They should then think whether it is safe to start work or do they
need to seek help or change their plan of action. This pause reduces the
possibility of “task focus” and reduces the possibility that a wish to “get the job
done” causes someone to not see unexpected hazards.
Chapters 16 and 17 cover how to identify tasks that may be prone to error and
how to pre-empt these errors. Chapter 18 looks at ways to help people develop
non-technical skills and detect errors and mistakes made by other people.
Job aids include things like instructions, checklists, procedures and information. By
the end of this chapter, the reader should be able to understand the:
Read this Chapter with reference to CCPS “Guidelines for Risk Based Process Safety”
[24]. This Chapter and Chapters 1, 7 and 8 build on the “Operating procedures” and
“Process Safety Knowledge” CCPS elements by providing additional insights and
advice on the Human Factors of procedures and job aids.
The CCPS 1996 book “Guidelines for Writing Effective Operating and
Maintenance Procedures” [25] also provides sample formats and checklists, and
advice on job aids.
Like many events, the Bayer explosion had multiple causes. Some causes were
associated with new or non-operating equipment. However, other causes were
clearly related to Human Factors.
• Bayer had upgraded the computer control system for the unit, installing
a new Distributed Control System (DCS).
• The DCS had slower response times.
• DCS displays were hard to navigate, as operators had to switch between
screens to complete tasks. In addition, only one process variable could
be changed at a time.
46 Human Factors Handbook
Job aids provide people with the necessary information and knowledge to perform
tasks. They help people perform tasks in a “rule based” mode of human
performance rather than having to rely on their general knowledge. Procedures
and instructions can help to ensure all members of a team have a consistent and
shared view of how to perform a task. This supports effective teamwork and
adoption of a safe way of performing tasks.
5. Human performance and job aids 47
Job aids also minimize the potential for error. This is because:
Be aware, that the dissemination of too many job aids and written procedures,
can create risks including:
More information can be found in the CCPS Guidelines for Writing Effective
Operating and Maintenance Procedure [25].
Section 8.5 shows how pictorial information can help. This is part of “Good
Human Factors Guidance” in Figure 5-1. Figure 5-1 advises a match of task
demands to a suitable type of job aid. In some cases, a job aid may not be needed,
such as for low risk tasks.
Valid content
The instructions and information in a job aid are important. Some ways to
decide what information is needed include Hazard Identification and Risk Analysis
(HIRA), task analysis, and task walk-through. It is important to not only include the
end user, but to involve them in the authoring and writing process to ensure they
are feasible, useful, and can be understood. If job aid is inaccurate or poorly
reflects task, personnel are less likely to use procedure.
Accepted by users
The approach to developing job aids should ensure that users accept that:
• They provide the correct and best way of performing the task.
• It is necessary to follow the procedure to safely operate and maintain
process operations.
This requires engagement with those people who are expected to use the job
50 Human Factors Handbook
Supervisors, managers, and other staff should routinely verify in the field that
job aids and procedures are practicable and can be used as intended. Infrequent
and inconsistent verification can lead to large discrepancies between what
supervisors and managers think is occurring within the operation and what is
actually being practiced by the workers in the field.
Up to date
All job aids should be kept up to date with changes in processes, equipment,
risk analyses, and legislation or regulations to ensure an efficient and safe
sequence of actions. All outdated job aids must be removed from the work place
to avoid confusion. This is a requirement of the CCPS “Guidelines for Risk Based
Process Safety” Management of Change element [5].
Safety culture
The underlying culture of the organization should promote and reward the use
of job aids and procedures. The CCPS guide “Process Safety Leadership from the
Boardroom to the Frontline” provides advice on organizational safety culture.
Figure 5-1 provides an overview of how to achieve the attributes cited in section
5.4.1. The approach aims to ensure that the task is properly understood and that
the instructions and advice in job aids are practical as well as correct. This requires
a combination of analysis, engagement with users and validation in developing a
job aid including the following.
Task
characterization
Select
type of job
Update &
maintain
job aid
52 Human Factors Handbook
6.2.1 Overview
Procedures with many tasks are time consuming to write and maintain. HIRA may
be used to prioritize the higher risk tasks and identify lower risk tasks for which a
job aid may not be required. In order for job aids to be accepted as necessary it is
important to produce them only when they are really needed.
However, as noted in 5.4, it is important to remember that low risk tasks may
not require any form of job aid to be used every time a task is performed. Also,
people should be trained for tasks that must be performed very quickly, such as
emergency response, especially if task completion time frames prohibit reading
through procedures.
For example:
Less frequent, more complex and critical tasks may benefit more from
step-by-step instructions and checklists. An example is process start-up.
Such tasks may be prone to errors (slips and lapses), especially if they
have many steps or take a long time. If the task is complicated and
involves judgment and decision-making, then SOPs and job aids can
support “rule-based” performance.
If the task is infrequent and complex, then it may be helpful to use
decision-making aids, such as diagnostic flow charts. These can give
operators the knowledge they need to decide what actions to take,
especially in abnormal or unique operational situations, such as process
upsets.
6.2.2.1 Overview
Figure 6-1 provides a flow chart to help judge whether a step-by-step guide or a
job aid may be more useful for a task. The best type of procedure or job aid
depends on:
For example:
The abbreviations for examples of job aids are given in Figure 6-1.
The chart is for guidance only. It can be amended to match company specific
policy and practice.
Figure 6-1: Selecting a type of job aid for operational use
Key:
CK = Checklist.
GC = Grab card.
DFC = Diagnostic flow chart
DT = Decision tree
Info = Information (e.g.,
chemical safety datasheet)
Log = Operational log
M = Manual
PTW = Permit to work
SH = Shift Handover
SOP = Standard Operating
Procedure
WI = Work Instruction
6. Selecting a type of job aid 57
The safety criticality of a task can be assessed using knowledge of the task-related
hazards. The results of Hazard Identification and Risk Analysis (HIRA) can be used
to rate task risk. A common HIRA approach is to use a qualitative risk matrix to rate
the risk from very low to very high. This risk matrix approach can be used to rate
the risk of a task. If a HIRA has already been completed for a process, the results
can be used directly. These risk ratings may be applied to the Task Criticality in the
flow chart previously shown in Figure 6-1.
The example matrix in Figure 6-1 uses three risk ratings – high, medium and
low. HIRA may use a risk matrix, as in Figure 6-2. Figure 6-2 also gives a potential
alignment of HIRA ratings to high, medium and low in Figure 6-1, with red cells
being high safety criticality, yellow being medium and green being low.
Figure 6-2: Using HIRA risk matrix results to assess task safety criticality
High
Moderate Medium
Likelihood
Low
Consequence
Task/ error Task Error Error Failure class MAH consequence Existing controls / Training & Actions
identifier guideword description of error recovery competency required
1.1 Liaise with the Wrong info Wrong info Mistake No MAH Trips
control room passed on
to and carry
out risk
assessment
1.2 Ensure test Operation Use Lapse MAH fail to detect Yearly calibration
measuring omitted incorrectly flame out
equipment calibrated
calibrated equipment
The engineered protection does not involve or require human action in order
to operate. It is tested and kept in operational condition so that it operates
automatically when it is needed.
Figure 6-4 shows an example, taken from INEOS ChlorVinyls Ltd, of how to use
consequence and barrier analysis to rate task safety criticality. A task where failure
has a low consequence (rated here as less than 4, where 4 includes fatality) or has
one engineered protection (e.g., a high level trip in a storage tank), would be rated
as “Criticality level 4, and would not be classified as safety critical. A high hazard
task (4 or more) without engineered nor procedural protection, is classified as
criticality level 1 or 2 and requires further assessment and management. This type
of analysis helps to screen tasks and select higher risk tasks for further
consideration.
These methods are relatively simple, can help identify safety critical tasks, and
can be referred to as qualitative analyses. One simple method that can identify
safety critical tasks and provide additional insights is Bow Tie analysis [31]. The
Bow Tie method graphically maps out failures that could initiate an accident and
the “barriers” against these failures and their consequences. This is explained in
the CCPS Bow Tie analysis guide [31].
Guidance is provided Table 6-1 and Table 6-2 for rating the remaining factors.
Some low complexity tasks may be performed frequently, such as depressurizing
oil storage tanks every day. However, sometimes the circumstances may change.
For example, a change in wind direction and speed may require special
precautions, such as turning off ignition sources downwind of the tanks. This could
be a low frequency task and higher complexity.
6. Selecting a type of job aid 61
The time available (see Table 6-3) refers to either the time scheduled for a task,
such as in a work instruction or schedule, or the time in which the task must be
performed before an accident occurs. This is sometimes called “process safety
time”. This is defined by the International Electro-technical Commission (IEC 61508
Edition 2.0) [34] as the “Period of time between a failure occurring in the process
or the process control system and the occurrence of the hazardous event if the
safety function is not performed”.
62 Human Factors Handbook
Job aids can be thought of as a reference, for use in training for example, and as a
“real time” task aid, for use in everyday operations. Figure 6-5 gives a mapping of
types of job aid, to types of operational human performance. The types of human
performance have been expanded. For example, knowledge-based tasks have
been split into Problem Solving and Diagnosis. This expanded list of types of tasks
provides a more precise way of matching the type of job aid to the type of human
performance.
Process safety information, Process flow diagrams, Mass balance charts and
Piping and Instrumentation Diagrams may all be used as a form of Diagnostic Tool
and Task planning aid. They are included as ‘Info’ in Figure 6-5.
Table 6-4 defines each type of job aid and notes their role in supporting human
performance.
Figure 6-5: Mapping of type of job aid to type of task performance
Key:
CK = Checklist.
GC= Grab card.
DFC = Diagnostic flow chart
DT = Diagnostic tree
Info = Process safety information
Log = Log books etc.
M = Manual
PSB = Process Status Board
PTW = Permit to work
SH = Shift Handover
SOP = Standard Operating Procedure
WI = Work Instruction
Table 6-4: Definition of types of operational job aids
Role in supporting human
Job aid Definition Task types
performance
A document that provides information
(e.g., how a piece of equipment works),
Process operational, systems
Manual (M) and instructions (e.g., how to operate a
and safety knowledge
piece of equipment). May contain
pictures with notes.
Documents explaining the physical,
Process safety chemical, and toxicological information Process operational, systems
information (PSI) related to the chemicals, process, and and safety knowledge
equipment. Knowledge-based operational
A diagram that shows the material flow planning.
from one piece of equipment to the Development of procedures and
Process operational, systems
other in a process. It usually provides training.
and safety knowledge
Process flow information about pressure, temp., Fault diagnosis.
Support process upset and fault
diagrams (PFD) composition, and flow rate of various Decision-making
diagnosis and decision-making
streams; heat duties of exchangers;
in rule-based tasks
and other information to help
understand the process.
A schematic showing properties
Process operational, systems
Material and (phase, temp., pressure, etc.) and
and safety knowledge
Energy Balance material inputs to and outputs from
Assists with rule-based process
(MEB) each stage of a process, including raw
operational decisions
materials, waste and by products.
Table 6-4 continued
One of the major benefits of electronic checklists is that they can keep track of
skipped items and can track items that are not marked as complete. Table 6-5
notes the pros and cons of handheld devices.
68 Human Factors Handbook
Can enable automated calculations to Glare and bright light conditions may
be performed reduce visibility.
• The need for a job aid depends on task safety criticality, frequency,
complexity and time available to complete the task.
• Many types of job aids are available.
• The best type of job aid depends on the type of task performance.
Human Factors Handbook For Process Plant Operations: Improving Process Safety and System
Performance CCPS.
© 2022 CCPS. Published 2022 The American Institute of Chemical Engineers.
This Chapter builds on the CCPS guide “Guidance for Writing Effective
Operating and Maintenance Procedures” [25]. In particular, this Chapter cites the
use of task analysis, the output from Hazard Identification and Risk Analysis (HIRA)
and task walk-throughs to help produce job aids. The Energy Institute provides a
detail guide on how to perform task analysis [29].
Task analysis includes identifying the task steps, describing the task actions, and
assessing the judgments and decisions needed to perform a task. The outputs
include:
The development of task analysis can be time consuming, but it can be used
for other purposes. For example:
• To define a minimum staffing level to safely and accurately carry out the
task.
• To identify opportunities to clarify the task instructions.
• To identify opportunities to improve the work environment (e.g.,
labelling, lighting, access).
It can also be used to help identify potential errors. This will help to show when
job aids are needed (tasks more prone to error) and where to add warnings and
cautions into procedures and job aids.
An example of one type of task analysis is shown in Figure 7-1. The diagram is
reproduced as a table in Table 7-1. The table shows which staff member does each
task – this also provides a minimum staffing level.
Storage
1 Prepare for Communication Communication
N/A Control
offloading task task
Room
2.1 Connect
earth to road
Offloading
tanker Action Checking task N/A
bay
grounding
point
2.2 Connect
hose from Offloading
Action Checking task N/A
vapor return to bay
tanker
2.3 Connect
hose from road Offloading
Action Checking task N/A
tanker to filling bay
point
4 Disconnect Offloading
Action N/A N/A
road tanker bay
Communication Offloading
5 Leave site Action Checking task
task bay
72 Human Factors Handbook
The output from Hazard Identification and Risk Analysis (HIRA) can be used to help
produce job aids. HIRA is explained in the CCPS “Guidelines for Risk Based Process
Safety”.
Engineers and safety specialists are given the job of writing or updating a
procedure. It is important that they include people who will use the job aid and will
carry out the work. These people have experience in the operational and
maintenance tasks and can provide useful advice.
In a walkthrough of an
Representatives of operational and maintenance emergency response
teams should therefore be engaged during job aid procedure, personnel
and procedure development. This will help to ensure realized that the
that they are based on a realistic and accurate view supervisor had to be in
of “how work is done”, and not based on how work is two places at the same
envisioned. time!
7. Developing content of a job aid 73
Task walk-through
It can help to use a questionnaire to assist with the production of the job aid or
procedure. This should include questions, with prompts or suggestions to help
encourage information sharing, and to capture details about the task. It should
also be used to record responses during the walk-through, to capture detailed
information about the task steps. An example process is provided in Figure 7-3.
The Human Performance Oil and Gas (HPOG) group also provide a Walk
Through Talk Through template and guide [35]. This is a free resource that also
covers capturing task steps, potential errors and ideas on error prevention.
It is important to validate job aids on their first use and on an ongoing basis.
An operational validation includes a review with operational and maintenance
7. Developing content of a job aid 75
teams. This review should aim to identify unforeseen issues with the practicality,
accuracy, and fitness for purpose of job aids. A technical validation may involve
safety specialist or process engineer checking the job aid.
The CCPS guide “Guidance for Writing Effective Operating and Maintenance
procedures” [25] provides further guidance on procedure approval.
When job aids are updated there should be a formal process to identify and
remove all out-of-date procedures to avoid the inadvertent use of old procedures.
This can be helped by color-coding versions or using watermarks with the date of
issue. All job aids should be updated by applying a formal process with final sign
off and authorization. The update process must be efficient and able to produce
timely updates. Long delays in the update cycle will create frustration with the
process and discourage operators from inputting them.
76 Human Factors Handbook
Table 8-1 provides advice on good practice for structuring procedures. The
structure should be intuitive and easy to follow. Table 8-2 provides a checklist for
the layout of job aids, including task instructions. Table 18-6 in Chapter 18 provides
information on task verification.
• Authors
• Safety and operating rules, limits and conditions
• Process summary
• Hazards
• Scope
• Minimum staffing and competences
• Roles and responsibilities
• Task pre-requisites
• Operating instructions
• Associated documents
78 Human Factors Handbook
2. Make headings stand out from the surrounding text to help people
identify information.
3. Use spacing, images, and/or blank or white areas to reduce clutter, and
to make it easier to identify and recognize information.
8. Clearly indicate who should perform the task step if the task involves
more than one person.
Prerequisites
This procedure must be completed by:
EHS requirements
8.3 Navigation
For larger job aids, such as manuals and long procedures, it is important to support
navigation to help the reader find the information they want. Ways to support
navigation include:
Headings
Headings should be large and stand out from the surrounding text. This will
help people to identify information. They can also be color-coded in terms of
information, or to indicate the level of the heading.
Color-coding
To help the reader find and identify relevant information color can be used to
code and group text. For example, sub-sections can be color-coded as per Figure
8-4. As some people are color blind, it is important also use other means of
distinguishing information, for example, by the use of icons (see section 8.6). Some
examples are also shown in Figure 8-4.
8. Format and design of job aids 83
It is also possible to utilize tone variance with color variance, especially to cater
for color blindness. For example, a saturated red and a soft green.
Hyperlinks
Hyperlinks can be used to help people find reference information. These are
not functional where people use printed copies and instead complete references
should be used.
When using hyperlinks, so that people can navigate back to their original page,
it is important to consider either:
Table 8-3 provides good practice guidance on the language used in procedures
and job aids.
• "Change the valve” rather than “The valve shall be changed by the
operator”.
The words “shall” and “should” may be misinterpreted, as they might be read
to mean that the valve shall already be changed. These terms do not convey
present tense.
11. Start each sentence with a verb (doing word) e.g., “close valve HV-001” rather
than “the valve should be closed”.
It is much clearer for the reader to understand directions if verbs are used at
the start of a sentence. For example, words such as: stop, open, check,
ensure, avoid, move, press, rotate, and lift.
Make reference to the relevant Isolation Preparation Make reference to the Isolation Preparation
Checklist. Checklist IP15 Issue 2.1.
Warnings and mandatory safety requirements are The warning is clear. Mandatory safety
hidden within the text. requirements are explicit.
Ensure that the valves HFA1 and HFA2 are closed. Close valves HFA1 and HFA2.
8. Format and design of job aids 87
Diagrams, images, pictures, and illustrations (such as Figure 8-6) are very effective
tools in communicating information to the user in a simple and usable way.
Drawings and diagrams clearly explain what needs to be done. Control system
interface screen prints help to explain and demonstrate how to use the functions
of a system. They also show where relevant information is displayed. Table 8-4
summarizes some uses of pictorial information.
8.6 Icons
Icons and symbols can help to give meaning to a sentence or paragraph. For
example, they can indicate the level of importance very quickly. They also help to
break up dense text, which can be unappealing and tiring to read.
It is important that they are used consistently and only with a clear link with the
text. Some examples are shown in Figure 8-7.
Icons should be used only where necessary. The excessive use of icons may
create clutter.
8. Format and design of job aids 89
Hold point: Stop work after inserting the second blind into
the manifold. Ask the team leader to verify the blinding.
• The structure and layout of a job aid can make it easier to navigate and
understand the information provided.
• The use of language tips can make written instructions much easier to
understand.
• Navigation through a job aid can be made easier by using color-coding,
hyperlinks and headings.
• Flow charts, decision trees and pictures can show complex information
in an easy to understand way.
• Icons and symbols can help indicate important meanings quickly.
Human Factors Handbook For Process Plant Operations: Improving Process Safety and System
Performance CCPS.
© 2022 CCPS. Published 2022 The American Institute of Chemical Engineers.
Part 3: Equipment
Human Factors Handbook For Process Plant Operations: Improving Process Safety and System
Performance CCPS.
© 2022 CCPS. Published 2022 The American Institute of Chemical Engineers.
9.2 Definitions
For the sake of this handbook, equipment refers to things that people use in their
jobs including:
• Tools.
• Instrumentation e.g., hard-wired display panels.
• Communication devices e.g., phones.
• Physical controls e.g., hand wheels or levers.
• Process equipment e.g., field switches.
• Computer systems e.g., display screens or input devices.
The user is a person who actively uses the equipment. This could also include the
maintainer, or more occasional users such as a supervisor or a person conducting
calibration or quality checks.
94 Human Factors Handbook
It is important to identify and consider all possible users during the design
process.
The explosion at the Buncefield fuel storage facility (2005) in Hemel Hempstead
(North of London) was one of the biggest in peacetime Europe. It measured 2.4 on
the Richter scale and was audible in Belgium, France, and the Netherlands. The fire
engulfed over 20 fuel tanks, and the resulting smoke plume was visible from over
60 miles (97 kilometers) away. The devastation was enormous with many nearby
properties damaged. Remarkably, there were no fatalities, although 40 people
were injured.
During a gasoline filling operation to tank 912, several safety controls failed
that should have prevented the tank being overfilled. Eventually large quantities
of petrol overflowed from the top of one of the storage tanks. A vapor cloud
formed, which ignited and caused a massive explosion and a fire that lasted five
days.
Figure 9-1: The Buncefield fuel storage facility before and after
(reproduced from HSE [38])
The accident report is available from the UK Health and Safety Executive [38].
• Only one visual display screen was available. This meant that the status
of only one tank could be viewed at a time, with information on other
tanks stacked behind them. On the night of the accident, the display for
tank 912 was at or near the back of the stack of tank displays.
9. Human Factors in equipment design 95
Getting the design right is very important. Well-designed equipment will help
people perform to the best of their ability and will reduce the likelihood of errors.
Sometimes problems with equipment are not fully appreciated because time
has not been taken to understand the tasks that the equipment is going to be used
for, how the user needs the equipment to perform and how they are likely to need
to interact with it. When user needs are not used to inform equipment design then
people have to ‘make do’ with poor design. In these circumstances, instead of
equipment being designed to suit the human, humans must adapt to poorly
designed equipment.
Problems with poor design can be complex and expensive to resolve. Often
workarounds are required as it can be prohibitive / impractical to re-engineer
equipment. These workarounds can sometimes be unsafe, inefficient and increase
the opportunity for error or equipment misuse.
The context in which the iPhone was going to be used had not been properly
considered. If the task demands were understood at the time of purchase, i.e., that
the device would be used to read lengthy complex documents, outside, sometimes
in poor weather when wearing PPE / gloves, then a more appropriate device could
have been identified.
The iPhone is well known for its intuitive and well thought-out design.
However, its small screen size means that it is not suited to showing large
documents, and it can be difficult to see in bright sunlight. It is also difficult
to use wearing gloves and in poor weather.
The origin of the discipline of Human Factors is partly rooted in tackling these
difficult design issues. In the mid-1940s, the United States Air Force identified a
problem with certain aircraft crashing on landing. Previously, when they could not
find an obvious reason for losing an aircraft such as a mechanical failure, the
response was to either retrain existing staff, to recruit new staff, or to change
96 Human Factors Handbook
recruitment procedures. Instead, the response this time was to bring in a specialist
psychologist to consider what was causing the crashes. It was identified that the
incidents were due to the positioning of the landing gear and wing flaps.
Designing equipment to fit the user is a key principle of Human Factors. It is far
more effective to design equipment well than training a person to use inadequate
equipment.
Even the most diligent people can make errors and fall for error traps.
9. Human Factors in equipment design 97
Error traps
Error traps are common. One of
the most familiar is to pull on a
door handle when the door
opens away and needs to be
pushed. The stereotype of a
handle is to pull on it. A door push
plate is pushed. By its design, a
thing indicates it use: this is
termed affordance.
The steps for considering Human Factors in equipment design are as follow:
1. Begin to take note and recognize what poor design is. Get a measure of
how this can impact human performance.
2. When the opportunity arises (e.g., when purchasing or designing
customized equipment) apply a “user-centered” design approach.
3. Intervene to address existing poor design if replacing problem
equipment immediately is difficult.
The United States Chemical Safety Board found that poor Human Factors design
of equipment was a factor in the ExxonMobil Baton Rouge Refinery Isobutane
Release and Fire in 2016 which resulted in four serious injuries to workers and
injured two others [39]. The incident occurred during minor maintenance on a
flammable isobutane line which failed, releasing isobutane into the unit which
ignited.
A 30 year old gear box needed to be removed. The bolts on the old gear box
were different to other newer gear boxes on the site. Most gear boxes were
newer. New gear boxes were removed using 4 vertical bolts on the top-cap of
the valve body. The old gear box could be removed using two horizontal bolts.
The top vertical bolts in the old gear secured the pressure retaining cap. The
operators removed the top vertical bolts. This breached the principle of “making
it easy for people to do things right and hard for them to do things wrong”.
Poor design can often lead to errors. Table 9-1 provides a summary of some of the
most common examples, and how these can impact on performance. Good
practice advice is given in sections 9.7.
Table 9-1: Examples of poor design for hard-wired interfaces – physical panels
Poor design Impact on performance
Hard-wired interface
Poor labelling e.g., it is:
• Not meaningful
(unfamiliar
abbreviations).
Poor labelling can mean important information
• Difficult to read is not identified or is misinterpreted.
(damaged, scratched)
• Poorly located (poorly
lit).
Figure 9-4 shows an image of two control screens for a set of filters and vessels.
Design issues with the panel include:
• The screens are mirrored images. The filters 1 to 4 are shown on the left
of top screen and filters 5 to 9 are on the right side of the top screen. On
the lower screen, filters 1 to 4 are shown on the right and filters 5 to 9
are shown on the left. This creates a potential for the operator to
confuse the filters between the two displays.
• The screens have no “mimic” of the connections between the filters and
vessels.
• The lower screen is very cluttered.
• The PSI indications are equidistant between the vessels at the bottom of
the lower screen, creating a potential to confuse which vessel the PSI
indicator relates to.
• Some of the indicated values obscure the vessel abbreviated names.
• The “FLOW In Out” labeling is not meaningful.
• Red text is used on the upper screen for RESET despite this not being a
warning.
• The gallons are shown in the upper screen as seven digit numbers such
as 3808998 which is harder to read than 3,808,998.
102 Human Factors Handbook
[40]
9. Human Factors in equipment design 103
A key principle of Human Factors is that the user remains central to the design
process. This ensures that equipment is fit for purpose and works in the way the
user needs it to. The ISO Standard 9241-210:2010 can be applied [41]. Methods to
help do this include:
Prototypes
In addition to ensuring that the user is central to the design process, there are
several simple design principles to follow. Many Human Factors standards are
available, such as ISO 26800:2011 Ergonomics — General approach, principles and
concepts [42]; ISO 6385:2016 Ergonomics principles in the design of work systems
[43]; ISO 11064-1:2000(en) Ergonomic design of control centers — Part 1: Principles
for the design of control centers [44]; and EEMUA Publication 201 Control rooms:
a guide to their specification, design, commissioning and operation [45].
Good equipment design should consider the physical world and how it minimizes
error and supports human performance in terms of:
For example:
A key process safety measure in process industries is for control room staff to
know what is happening out on the plant, and what processes are currently in
operation. This means having a representation in their minds (or a “mental model”)
of the plant function, so that they can start and stop processes as needed and
intervene if something goes wrong. The “correctness” of the mental model will
influence the correctness and appropriateness of the operator’s actions.
One of the problems at the Buncefield site (see Section 9.3) was that the control
room staff did not have a clear mental model of which tanks were being filled or
of their status. If they had, they might have realized that tank 912 was being filled
to an unsafe level, and they could have stopped the filling process.
A key aid to helping provide an accurate mental model are “mimics” that display
the plant status. These mimics should be designed so that they give correct
information that is easy to interpret. When the operator makes changes or inputs
to the system, they should be provided with accurate and timely feedback in
response to their actions. A key aspect of supporting this mental model is, as far
as possible, to “map” to the arrangements in the real world. This is called “natural
mapping”.
This is called “knowledge in the world” rather than “knowledge in the head”.
This means it is not necessary to know or remember any additional information to
understand how an object should be operated.
108 Human Factors Handbook
Figure 9-6: Examples of good and poor natural mapping for a stove
The three main levels of natural mapping between controls and the object
being controlled are:
1. Controls are on the item to be controlled e.g., a water faucet where the
control and water outlet are part of the same unit.
3. Controls are arranged in the same spatial configuration (or layout) as the
objects to be controlled, as shown in Figure 9-6.
The controls to open and close the valves, and to turn the pumps on and off
are co-located. The blue arrow shows the reservoir is filling, and the gauge
associated with the valves and reservoirs indicate flow and fill rate.
9. Human Factors in equipment design 109
9.7.3 Alarms
The use of, and response to, alarms plays a critical role in plant and process safety.
Despite this, control rooms often operate with less than optimum alarm
management, for example:
• Alarms are not identified – alarms are missed because they are either
not seen or not heard, or because they are ignored due to previous
nuisance alarms or “alarm flooding”.
• Alarms are incorrectly prioritized – alarms do not indicate the priority,
or the priority is difficult to determine e.g., if audible alarms of different
priority sound the same.
• Alarms are not informative – alarms do not come with supporting
information that provides additional information about what the
problem is.
110 Human Factors Handbook
EMMUA 191 sets out a very simple performance metric to help assess alarm
system performance that shows how this can impact on operator ability to
respond. These are as follows:
The EMMUA 191 guidance goes on to suggest that if these benchmarks were
achieved, operators would find alarm systems more manageable. A summary of
the key principles of good alarm design is shown in Figure 9-8.
9. Human Factors in equipment design 111
It is important to remember that mitigating poor design may not be simple nor
effective. Introducing measures to compensate for poor equipment design such
as delivering additional training, implementing additional checking steps or writing
new procedures can be a difficult undertaking. It can also shift the emphasis of risk
management towards administrative or organizational controls. This can be a
significant burden to maintain and increases the opportunity for human error, or
shifts the error to a different element of the system (i.e., failure of administrative
controls).
112 Human Factors Handbook
Alert
Alarms should focus operator attention on the
important or urgent issues
Inform users
Alarms should provide information to help users
understand the issue, and then decide upon
appropriate actions
Indicate importance
Alarms should indicate the priority of the alarm
and the timeframe for response
Be easy to interpret
Alarms should not require uses to obtain
additional information from other sources befoe
they understand how to respond
9. Human Factors in equipment design 113
The CCPS “Guidelines for Risk Based Process Safety” [24] cites “Process
Safety Competency” and “Training and Performance Assurance” as
elements.
This Chapter and Chapters 11, 12 and 13 build on the CCPS guidelines
by providing additional insights and advice on the Human Factors of
competency, learning, and Competency Management. These Chapters
focus on operational competency.
Competency is also defined as the ability to perform work activities reliably and
consistently, to the required standards. Competency can be measured against
these standards. A term such as “Suitably Qualified and Experienced Person
(SQEP)” can be used to indicate that a person is competent in their role/in the tasks
they are conducting. This includes routine and non-routine tasks; abnormal and
upset; first line emergency response; safety-critical maintenance, inspection and
testing activities.
118 Human Factors Handbook
Competency includes not only the application of technical skills and knowledge,
but non-technical skills such as communication; this is important for supervisory
roles.
This phase includes assessment of the gap between individual and team
competency, and the competency that individuals need to develop. It is
then necessary to identify the learning required to bridge the gaps,
including the type of learning most suitable to develop and demonstrate
the competency. This includes providing a description of learning
objectives, to aid development of the on-the-job learning and training
programs in Phase 3.
More detail on Phases 1, 2, 3, and 4 are provided in Chapters 11, 12, 13 and
14, in order.
120 Human Factors Handbook
Warwickshire Oil Storage Ltd followed a process to identify high-risk tasks, and
to define competency standards. Good practice aspects included:
The Esso Longford gas explosion in Australia in 1998 was an industrial accident
with severe consequences [20]. It is summarized in B.3 (page 387).
“Though the existence of a link between this failure and the occurrence of
the accident is hard to evaluate, appropriate management of change risk
assessment may have exposed important and relevant weaknesses in the
level of operator knowledge, in training programs, in communication
systems, in operating procedures and in other aspects of Esso’s
management system.” [20]
The operators and supervisor present on the plant on the day of the accident
were highly experienced individuals, yet no-one recognized the hazards associated
with the plant conditions.
• Training programs: the programs did not include training with respect
to hazards associated with the loss of lean oil flow, hazards associated
with uncontrolled flows, critical operating temperatures of vessels, or
circumstances where brittle structure may occur.
• Plant Operating Procedures Manual: the Operating Procedures did
not contain any reference to loss of lean oil flow, or procedures on how
to deal with such events. Some information referring to the “Loss of
Lean Oil Circulation” could be found in the Red Book (1975 Operating
Instructions for Absorption Oil System), which was located in the training
room, and was not part of the Operating Procedures.
• Lack of understanding: the operators did not fully understand the
dangers of cold metal embrittlement, due to flaws in their training and
assessment.
• Knowledge assessment failures: the assessment did not test for real
understanding, because operators could give the correct answer to
questions without understanding the meaning of their answers – they
were learning by memorization. For example, operators knew the
correct answer to a question on the action of a valve was to “prevent
thermal damage” but did not know what was meant by “thermal
damage”.
• Reassessment of knowledge: this was conducted superficially on the
basis on operators’ self-confirmation of knowledge, rather than
examination of their understanding. Operators were asked if they
understood the matter. If they said “yes”, they were “ticked off” as
competent.
Operator competency, and actions during the day of the accident, played a
significant role in the outcome of the event. Three main factors impacted
competence.
The first factor was the number of on-site staff. Numbers had been reduced
due to gradual reduction of staff (over the period 1993 to 1998). The number of
supervisors and associated staff was reduced from 25 to 17 and the number of
maintenance staff was reduced from 67 to 58. This staff reduction led to
124 Human Factors Handbook
The second factor was the reduced expertise on plant, due to the centralization
of engineers. All Longford engineers were relocated to Melbourne. Melbourne is
140 miles (230 kilometers) away from Longford, so any “technical” help was not
around the corner. Engineers lost awareness and detailed knowledge of the plant
activities; and operators were less able to consult engineers when required and
lacked the additional on-site support and expertise to maintain safe operations.
The final factor was the organization’s Competency Management Systems. The
training programs, knowledge assessment, and operating procedures failed to
equip operators with the required knowledge and skills to cope with emergency
situations.
• How can the system help set the individual up for success?
• How can the system absorb mistakes that the individual can make?
Activities in the process industry are often carried out in difficult conditions.
Hazardous environments, complex processes, and production pressures demand
higher levels of competency.
Figure 11-1 links the level of safety criticality to the level of training and
competency assurance. Each task can be rated against:
• Task criticality.
• Task complexity.
• Task frequency.
• Time available to complete the task.
Learning needs and their requirements range from “Very High” to “Very Low”.
For example:
Key:
CK = Checklist.
GC = Grab card.
DFC = Diagnostic flow chart
DT = Decision tree
Info = Information (e.g., chemical safety
datasheet)
Log = Operational log
M = Manual
PTW = Permit to work
SH = Shift Handover
SOP = Standard Operating Procedure
WI = Work Instruction
130 Human Factors Handbook
Job and task analysis explore the required job or task competency in detail and
provide inputs for defining performance standards.
“Task analysis for training design is a process of analyzing the kind of skills and
knowledge that you expect the learners to know how to perform” [52, p. 3].
Safety Critical Task Analysis can be used to make a distinction between which
type of knowledge should be provided during training. The Safety Critical Task
Analysis can be applied as follows:
Various forms of job and task analysis exist. Some examples of task analysis
methods include:
For illustration purpose, this Chapter focuses on tabular form analysis, as this
form of analysis allows for clear identification of required competency. The tabular
task analysis can identify required skills, procedural competency, and knowledge
per task and sub-task. It can also support decisions with regards to training
(memory-based information versus resource access and application).
The task analysis used in Chapter 6 is extended here for the purposes of
identification of required competency. In Appendix D, Table D-1, Table D-2 and
Table D-3 provide different levels (1, 2, and 3) and type of tasks (Skill-based, Rule-
based and Knowledge-based).
replace written procedures with on-the-job training that uses detailed verbal
instructions and tasks demonstration.
• Performance Standards.
• Key competency list consisting of:
o Skills (e.g., torqueing a bolt).
o Rules and procedure competency (e.g., able to enact
procedures, able to diagnose faults).
o Knowledge (e.g., understanding a chemical reaction).
It is also possible that a generic standard may already exist as per the example
in Table 11-1. A generic standard can be used instead of a customized standard. A
11. Determining operational competency requirements 133
check is required of whether the generic standard identifies and includes all of the
competency required for a specific process operation. If it does not, then the
specific competency should be identified, and added in.
The competency standards should take into account the allocation of roles
among a team. For example, a supervisor may be required to have the ability to
develop a safe operating procedure, while an operator is required to be able to
understand and apply the procedure.
Competency – Job Level Requirements – For Petrochemical Staff of Cairn India Limited (CIL) “Reservoirs”
Level 1: Awareness Level 2: Basic Application Level 3: Skilled Application Level 4: Mastery
• Describe • Be able to characterize • Interpret reservoir maps and • Participate in integrated
lithostratigraphic reservoirs based on integrate them with seismic reservoir studies and
(strata or rock layers) interpretations of interpretations to predict recommend new strategies for
sections in cores and sedimentological and structural reservoir size, geometry, and optimizing reservoir models.
outcrops, based on models. trends. • Know the current state of the
sedimentary rock type • Build and interpret isopach and • Use regional geology art in reservoir
classification. net-to-gross sand maps. concepts and integrate them characterization and apply
• Use log • Characterize reservoir quality with seismic data to predict new techniques to improve
interpretations to based on sedimentological and reservoir trends. results.
generate petrophysical rock properties. • Create conceptual geological • Develop and foster networks
lithostratigraphic models from diverse data outside of Cairn, to access
descriptions and
• Be able to interpret diagenetic
processes in sediments, and sources, in preparation for specialists experienced in
identify facies. constructing geocellular various types of reservoirs.
construct paragenetic
• Determine reservoir sequences using textural models. • Visualize potential application
rock quality using information from a variety of of the adequate reservoir
core-sampling sources. geological models, to enhance
techniques.
• Be able to integrate diagenetic reservoir production.
• Recognize basic events with structural history. • Lead mentors and coaches.
techniques for
processing and
• Adopt a Quality
Assurance/Quality Control role
validating reservoir
wherever necessary and
data.
appropriate.
136 Human Factors Handbook
Competency Gap Analysis and Training Needs Analysis are important stages in
identifying learning objectives, and training needs. Normally, they would be carried
out together, as information from the Competency Gap Analysis feeds directly into
the Training Needs Analysis.
Completion of the Competency Gap Analysis and See Chapter 14 for more
Training Needs Analysis template includes information on
information about: competency assessments.
Process control room supervisors should be able to successfully manage a simulated emergency response in
Competency
three tests (out of a possible 10 scenarios), & display appropriate skills such as task delegation, & effective
standards
communication.
Competency Skills
Knowledge of emergency response procedures
current Delegation: Awareness: Level 1
in various situation: Basic application: Level 2
level Communication: Basic application: Level 2
Skills
Competency Knowledge of emergency response procedures
Delegation: Skilled application: Level 3
required in various situations: Skilled application: Level 3
Communication: Skilled application: Level 3
Improve delegation skills: Move from Level 1 to Level 2 to Level
Learning Increase knowledge of emergency response
3
needs procedure: Move from Level 2 to Level 3
Improve communication skills: Move from Level 2 to Level 3
Improve knowledge of emergency procedures by Improve non-technical skills used during emergencies by
attending training, & completing required attending training, & completing required assessments,
assessments, within next four weeks. within next six weeks for Level 2, & within next 12 weeks for
Learning Level 3.
Effectively manage emergency response in a
objectives
series of simulated exercises, by using correct Demonstrate effective & efficient communication &
emergency procedures suitable for each delegation skills during emergency response simulation
scenario. exercises.
“Bridging
Classroom-based training; Walk-through Classroom training on Non -Technical Skills; Simulation
the gap”
procedures; Simulation training/case studies training
training
Learning Direct questioning; Open questions; A “show me
Quiz; Case studies; Observation
evaluation how” observation
12. Identifying learning requirements 141
• “Bridging the gap” is the training type that will help a person move from
one level of competency to another (e.g., on-the-job training, shadowing,
classroom learning, simulator, online learning, or a professional
qualification), with on-the-job learning providing most effective learning
opportunities.
• “Learning evaluation” is about assessing an individual’s competency after
the training has taken place.
o Assessments can take various forms (e.g., multiple-choice quiz,
“show me how” observations, or verbal explanation of the
process).
o As assessment would be carried out:
• Immediately after the training, to see what has been
learned.
• Some-time later, so that the person has a chance to use
(and show use of) new knowledge or skills.
Setting realistic objectives for the learning opportunity, as noted in Table 12-1,
is important when designing learning programs. The learning objectives should:
• Clearly state the purpose (e.g., how they fit with new procedures).
• Describe the expected outcomes that an individual should achieve at the
end of the learning opportunity.
o For example, move from Level 1 to Level 2 (from awareness to
basic application) competency, such as being able to follow
emergency procedures with instructions from others versus
being able to follow emergency procedures with minimal
instruction from others.
• Consider elements required for the learning e.g., pre-requirements for
attendance/admission, materials, resources.
• Ensure the learning objectives are SMART:
o Specific (clear)
o Measurable
o Attainable (achievable)
o Relevant (appropriate for the needs)
o Time-bound (to be done within a stated timeframe)
• An example of a SMART objective is shown in Table 12-1.
Learning objectives set expectations about developmental needs. They also act
as a measuring point to check progress or development of competency and
performance. Assessment of competency focuses on whether or not the stated
objectives were met.
142 Human Factors Handbook
• Two types of analysis are usually carried out to identify training needs
requirements:
o Competency Gap Analysis –identifies which competency need to
be developed.
o Training Needs Analysis –identifies what learning opportunities
are appropriate to develop the competency identified in
Competency Gap Analysis.
• Competency Gap Analysis and Training Needs Analysis also allow
learning objectives to be set.
• Learning objectives:
o Set the expectations of a person’s competency.
o Show the development (or required development) of an
individual’s competency.
o Allow progress of development (changes in competency) to be
effectively measured.
Human Factors Handbook For Process Plant Operations: Improving Process Safety and System
Performance CCPS.
© 2022 CCPS. Published 2022 The American Institute of Chemical Engineers.
Learning
Description
method
Psychomotor tasks may require more practice time, for example, on-the-job
learning. This is because many of these tasks require to be carried out
automatically, without much thought required about how the job should be done.
5% Lecture
10% Reading
20% Audio-visual
Passive teaching
methods 30% Demonstration
Specific recommendations/conditions
Type of learning
determining type of learning
Event-based learning -
introduces events or “trigger • Useful for problems with a particular
situations” within training subset of tasks, and the tasks can be
exercises that provide simulated.
opportunities to observe skills.
A common theme from accident reviews shows that it is often assumed that
because individuals have been provided with specific training, they should be
competent in conducting their job. This is often not the case.
Various methods can be used to aid competency assessment. The chosen method
should be suitable for the assessment of the competency in question. For example:
Peer review
Suitable for rule-
• Assessors (individuals or
Feedback by • Subjective due to colleagues providing
line manager,
based and skill- • Superior knowledge of
relationship between feedback) should be
based task execution
supervisor, or assessor and individual trained in providing
competency
colleague feedback
156 Human Factors Handbook
14.4 Reassessment
Individual and group competency should be maintained over time and reassessed
to prevent skill fade. For example, refresher training is important for safety critical
roles and infrequent tasks. The reassessment requires use of methods that are
suitable for assessing competency and human performance, as shown in Table
14-1.
14.5.1 Overview
• Ineffective training:
o The training did not specifically target the required set of
competencies.
o Learning objectives were not clearly defined.
o The method of learning was not sufficient to develop the
required competency.
• Skill fade linked to infrequent tasks.
o Even with acquired learning, if there is not enough opportunity
to put the new competency into practice, an individual’s ability
to do the task to the required standards decreases over time.
• Excessive workload (e.g., due to understaffing) affects personal and
team performance.
• 'Drift' over time. Even individuals who are experienced and
knowledgeable start to 'do things in their own way'- over time. This
behavior can drift into being unsafe and need to be recognized when
carrying out reviews.
• Poor assessment practices, such as:
o Poorly defined competency assessment matrices.
o Assessors lacking training and expertise in assessment.
• Personal reasons and circumstances, such as:
o Poor health.
o Challenging family circumstances.
• Work-related conditions including, such as:
o Poor quality of tools used for the task.
o Poor team cohesion.
o Difficult or poor relationships with co-workers.
• Performance gaps should be dealt with promptly. The aim would be to:
o Demonstrate commitment to a high-reliability culture.
o the issue from becoming more serious over time, and prevent
it having negative safety-related outcomes (e.g., leading to an
incident).
• Welfare and confidentiality should be maintained by:
o Providing people with relevant support. For example, having
one-on-one discussions to investigate issues which may impact
employee performance.
o Ensuring confidentiality of performance reviews and follow up
discussions.
160 Human Factors Handbook
• Unique set of competencies required for each job role, safety critical
task(s), and sub-task(s).
• Equipment, tools, and machinery used within each task, and skills
required to use these tools effectively.
• Date, duration, and type of learning opportunity provided for each
competency.
• Proof of evidence demonstrating competency level. This may include a
statement of compliance, or sign-off sheets.
• Historical record of all previous training, and any changes to skills sets.
The 2005 Texas City refinery explosion summarized in B.1 (page 383) occurred
during the start-up of the isomerization (ISOM) unit, following maintenance [14].
One contributing factor was fatigue.
In addition to working 12-hour shifts, they spent time commuting to and from
work, and assisting at home. There was limited opportunity to take rest breaks
when on shift.
These operators made mistakes on the day of the accident. For example, the
Day Board Operator did not recognize that feed was entering the unit but not
being removed, causing it to overfill. When the tower experienced pressure spikes,
the operators tried to reduce the pressure without exploring what was causing the
pressure spikes. They were focused on the symptom of the problem rather than
its cause. Awareness, vigilance, monitoring, and decision-making are all tasks that
can be affected by sleep deprivation and/or fatigue.
High levels of fatigue can cause people to uncontrollably fall asleep or have
“micro naps”. People will not be aware that their performance is affected by fatigue
and may think incorrectly that they can “power through” or use stimulants such as
caffeine, to combat fatigue. This is not true.
Sleep allows the brain to recharge and remove toxic waste by-products which
accumulate when awake. Sleeping helps to “clear” and reset the brain. A reduction
or disruption of the sleep cycle prevents the brain from maintaining their normal
function. Sleep is important for optimal cognition and judgement.
Lack of sleep and inadequate rest breaks will affect more complex tasks and
tasks that require judgment and decision-making more so than simpler tasks.
However, tasks that place very low levels of demand on people, such as monitoring
a process, are also vulnerable to fatigue.
15. Fatigue and staffing levels 165
Acute and cumulative sleep deficits create fatigue. Acute deficits might be
caused by working one night, while cumulative deficits might be caused by working
seven nights without a rest day.
It is also possible that medical conditions can cause fatigue, such as narcolepsy
which causes daytime drowsiness and “attacks” of sleep. The management of
medical conditions is beyond the scope of this book, but an option is to test people
for conditions such as narcolepsy, especially if they show signs of daytime
drowsiness.
A set of examples of fatigue risk are shown in Figure 15-1 to Figure 15-3. These use
the United Kingdom’s Health and Safety Executive’s Fatigue and Risk Index [60] .
This is a free online tool that predicts fatigue and risk according to working hours,
rest breaks, and task demands. Task demands include both physical and mental
demands.
The Health and Safety Executive’s Fatigue and Risk Index has been used here
to provide examples that illustrate the impact of long hours. As with all analysis
methods, the results are approximate. It is also best suited for assessing rotating
(day- night) shift patterns rather than permanent day or night working.
The fatigue score ranges from zero to 100. A score of 50 indicates a 50% chance
that people will struggle to stay awake. The Health and Safety Executive’s Fatigue
166 Human Factors Handbook
and Risk Index does not prescribe acceptable levels of fatigue risk. It advocates
that fatigue and risk scores should be reduced as a low as reasonably practicable.
The Fatigue and Risk Index also gives a risk score that indicates the potential
for an accident/incident to occur.
In all examples it was assumed that the person takes 40-minutes to travel to
work and has a “moderately” demanding role that required attention “most of the
time”.
A long day
Figure 15-1 shows an example of how fatigue “jumps” when working a 16-hour
day. The fatigue score rises slowly from day one to day four. Then the fatigue score
jumps on day five. It then stays high for days six and seven.
Figure 15-1: Example of rapid rise in fatigue scores from a 16-hour day
The lack of rest breaks causes fatigue scores to be increased three times.
The fatigue level at the end of the seven days without rest breaks would be
roughly a one in three chance of struggling to stay awake at work.
15. Fatigue and staffing levels 167
Figure 15-3 shows a seven-day rotation starting with four-day shifts (07:00-
19:00) and ending with three-night shifts (19:00-07:00). It assumes rest breaks are
taken and includes a 24-hour break between the final day shift and the first night
shift.
The fatigue level jumps upon starting night shifts, with a nearly 50% chance of
struggling to stay awake on the final night shift.
The fatigue score can be reduced by about one third by having more frequent
and longer breaks. For example, breaks could be of half an hour and every two
hours at night instead of fifteen minutes and every two and a half hours in the
daytime. In addition, risk can be reduced by performing work that is less
demanding at night.
168 Human Factors Handbook
Further guidance on fatigue risk management is also available from the Energy
Institute [62].
The policy on maximum working hours and rest breaks should take account of
the physical and mental demands of tasks. More demanding tasks require more
rest. In addition, it should control people volunteering for over time. The policy on
the maximum hours worked should also limit permitted voluntary over time and
avoid a small number of workers taking on excessive hours.
15. Fatigue and staffing levels 169
Maximum hours worked per day, per week, and per month
Minimum rest periods within a shift, and minimum sleeping time between
shifts
Shift design requirements, such as forward rotating and avoiding early starts
Action to be taken if people are working excessive hours each day or too many
days without a break
Figure 15-5 gives guidance on shift design and working hours for safety critical
roles. This includes allowing people to nap (sleeping for about 20 to 30 minutes) at
work, especially when working nights (e.g., at around 02.00 am), during periods of
inactivity if safe. Longer naps may induce sleep inertia, such as feeling “groggy”.
170 Human Factors Handbook
Principle Guidance
Regular rest breaks Rest breaks help people recover during a shift.
People may be screened for signs of fatigue at the start of shifts and monitored
throughout a shift. Tasks should allow for rest breaks throughout the shift. Fatigue
detection technology can monitor eye closures and head posture.
As noted in section 15.2.3, people’s sleep/wake cycle means that complex tasks
may best be scheduled for the start of the day and the start of the work block,
when people are most alert and rested. Work planning should avoid scheduling
complex tasks for night shifts, after lunch (13.00 to 15.00), or during early starts.
There are differences in people’s fatigue risk. Those at greatest risk can include
older people, especially over 50 years of age, people with a challenging home sleep
environment (e.g., with young children), and those with long commutes. The
allocation of work and the scheduling of tasks may take account of individual
needs. For example, the tasks requiring the highest level of concentration should
be allocated to individuals at lower risk from fatigue.
174 Human Factors Handbook
Working hours can be monitored. For example, the actual hours worked can be
checked by a team leader at the end of a week. If excessive hours are being
worked, the reasons for this can be explored and action taken. For example, if it is
due to other staff being absent due to illness, a request may be made for staff to
be redeployed from elsewhere. It is especially important to monitor the hours
worked by people when:
Facilities should maintain sufficient staff to cover absences and people leaving
an organization. If this is not the case, real time monitoring of hours worked
becomes even more important.
It is common that a shift or activity will have a defined number of staff, such as
a maintenance team of eight. It is good practice to map workload against staff
numbers and set out what tasks can and cannot be done by an understaffed team
e.g., no intrusive maintenance if team is depleted by 2 staff. The actual staffing
level can be monitored on a daily and weekly basis, against the defined minimum
level. Understaffing should be reported to responsible management. There should
be pre-planned contingencies, such as calling on additional staff in the event of
(for example) staff absence due to illness.
15. Fatigue and staffing levels 175
It is also possible that a person’s workload is too high due to the design of the task.
An example is where the time taken to start up a process is so lengthy, that it stops
people from taking a rest break for over four hours, or the work is high intensity
and tiring.
Figure 15-8 shows the options for managing workloads. Typical examples of
these include:
The number of people required for a task can be determined from past
experience of performing that task, such as replacing a pump. Past experience can
be used to decide how many people will be needed the next time this activity is
performed.
15. Fatigue and staffing levels 177
The example refers to locked blinds, as part of HEC (Hazardous Energy Control).
In some countries it is more common to use blind tags. The U.S. OSHA equivalent
of HEC is LOTO (Lock Out- Tag Out).
180 Human Factors Handbook
Four locked blinds under hazardous energy control (HEC) were removed from
the transfer line under the coke drums.
The blinds should have been left in place for a confined space entry isolation
to the heater.
Three of the blinds were found hanging from the cables with the locks and
tags attached. A cable had been cut to remove the fourth blind.
The product could have leaked through the valves, entering the tubes inside
the heater.
The permitted scope was too broad. It covered two jobs and 11 different
blinds, which were generically referred to as “blinds”.
Lack of communication.
Lack of clarity around removing locked blinds. A workaround allowed the same
crew to remove locked blinds when a hydro blind was leaking.
The onboarding should be updated to ensure new staff know that locked
blinds should not be removed.
In addition, motivated staff can be very “task focused” and intent on completing
the task and solving the problems. This can create a risk of losing awareness of the
situation, improvising unsafe ways of completing a task, and overlooking
unexpected events or conditions that require a change in their actions. When
people are task focused, they can miss “weak signals” around them that the
situation is unsafe or is changing.
182 Human Factors Handbook
Error assessment involves foreseeing the potential for human error in a specific
task, and the conditions that may cause failure. The person(s) planning tasks
(such as team leaders, supervisors, and senior engineers) are error managers.
They should be evaluating for potential errors and preventing them by good
task planning.
This type of assessment may be carried out as part of Job Safety Analysis,
development of Permits to Work and task specific work instructions. It requires the
people undertaking the error assessment to have some understanding of human
failure (18.4), and a belief that humans will fail if negative PIFs are present.
The task planner “walks through” the task in the field (not just on paper) and
the conditions in which the task is to be performed and identifies potential failures.
The task planner should refer to any available process hazard analyzes and
previous post job reviews.
The task planner should ask their team and other specialists to help with the
assessment. Ideally, the “walk-through” should be done by the team who will be
performing the task. This allows verification of task sequence, tools, staffing
requirements, timing and so forth. The team approach may identify some
improvements in tools or methods not used before.
16. Task planning and error assessment 183
The task plans are communicated as part of “Tool Box” talks, Tail Gate briefing,
start of production shift briefings and other forms of operational briefings.
Typical error-likely situations are shown in Figure 16-1. The presence of these
conditions may make the task high risk.
Communicating
Shift handover with
Starting up a modified between physically
defective equipment
process for the first separated individuals
or plant in abnormal
time or teams during a
state
long duration task
More information can be found in Table 16-2, which also offers some tactics for
managing these three types of situations. It can be used as a checklist to identify
error-likely situations.
Chapters 2 to 4 discussed types of human error and their causes – slips, lapses,
and mistakes. Task-specific conditions can contribute to potential errors and
mistakes, such as unclear instructions, task complexity, and inadequate task
experience.
and control
17.2 Overview
Figure 17-1 provides an overview of Human Factors for task planning, preparation
and control. The results of assessing the impact and likelihood of error from
Chapter 16.4 should feed into task planning and preparation.
Human factors task planning is necessary for the operational continuity of the
progress of the plan between and among shifts. “Discipline” is required in
performing the familiar phases of permitting, energy isolation, zero energy state
verification, maintenance work and returning equipment to service.
190 Human Factors Handbook
Team briefings
and committed, and mistakenly believe they can achieve the unlikely or
impossible.
A wish to avoid conflict within a team or a strong team leader may prevent
individuals from challenging schedules. This is sometimes termed “group think”
(see Chapter 19.5). Similarly, a common fallacy (or mistaken belief) is to work
backwards from a deadline to determine the time available to complete a task,
rather than working forwards from an estimate of the time needed to complete
the task. Another fallacy is that increasing resources or setting a challenging
schedule will reduce the time taken to complete a task. The task completion time
may be determined by, for example, the rate of depressurization of a vessel and
whether or not unforeseen equipment faults and defects are discovered.
An open team culture that invites, values, and accepts questions and
alternative opinions is vital. When questions are asked about decision and plans,
they should be explored and understood. People need to be able to trust that their
questions and opinions will be welcomed by their colleagues. This can be helped
by the use of open and neutral language that depersonalizes common goals. Refer
to Figure 17-2 for examples.
Leaders and meeting facilitators, before ending a meeting, can use a simple
technique to ensure each individual has shared any remaining concerns they may
have. Using their first name, the meeting leader simply asks each individual if they
have any additional concerns or comments to share with the group - e.g., 'Jim, do
you have any additional concerns or comments? This approach is can elicit
concerns more effectively than asking the group as a whole 'Does anyone have any
additional concerns?' The reason for the effectiveness of this approach is simple:
many people simply will not speak up unless asked directly.
If someone is direct and clear in asking questions and stating opinions, this
increases the likelihood of being heard, especially if the team is operating under
time pressure.
In the following examples, the top sentence in the “Do not say” box fails to
recognize the problem. The second sentence may be interpreted as a complaint
and lead to a discussion about whether people are being treated fairly, rather than
trying to check how much time is needed to perform the task.
In the “Do say” box, the problem is clearly stated. The absence of accusation
avoids confusion regarding the required response to the statement.
Fails to state
Ambiguous the problem
Accusatory
Some more tactics for reducing the potential for overly optimistic scheduling
are noted in Table 17-1.
17. Error management in task planning, preparation and control 193
Compare the experience of teams who performed this task in the past with the
current team. A less experienced team may take longer.
Consider how long it has been since this task was last completed. If it has been a
long time, people may need to spend time to remind themselves of how to
perform the task.
Add an assumed uncertainty, such as +/- 20%, to task completion times.
In the event of a task sequence being interrupted, standard operating practice
can require that people return to the start of the task sequence and verify that
each step has been correctly performed. This time should be factored into the
schedule.
Consider the sequencing of tasks presented in the procedures, any required
checks/interruptions, and how that may impact schedule.
Have an independent third party facilitate planning sessions for larger scale
works, such as plant turnarounds.
Encourage an open culture where team members can challenge plans without
fear of rebuke or ridicule.
Encourage a culture where team members can trust that leaders and colleagues
will respond neutrally or positively.
Strongly promote the importance of safety over other business objectives and
have this as a shared imperative.
Share examples of when upper management chose safety over production
within the company. Share specific examples of when safety was selected over
production at the floor level, and this decision rewarded by upper management.
Explicitly run “Decision Reviews” where team members are required to identify
and state potential problems and delays; and offer alternative plans.
Have team leaders explicitly say that people should “speak up” and not self-
censor if they think plans are unrealistic or if plans have not taken potential
problems into consideration.
194 Human Factors Handbook
Some tactics for minimizing distraction and interruptions are provided in Figure
17-4. These can be put in place through a number of methods, including:
This may be applied especially during peak workload periods, such as starting
up a process, devising Permits to Work, shift handover, or responding to a process
upset.
17. Error management in task planning, preparation and control 197
Noise reduction
Task design and planning
Wear hearing protection, use
Do not require people to multitask or
temporary sound barriers, and use
switch from one task to another.
noise muffling equipment.
Schedule low priority tasks in low
Use hand signals and written
workload periods.
communication.
Communication
Low priority alerts Limit communication within the team
to critical points. Use brief and
Minimize low priority alerts and
formal communication to minimize
alarms.
duration of interruption and level of
attention required.
198 Human Factors Handbook
Job aids can be used to reduce the impact of distractions, as discussed in Chapter
5 to 8.
• Checklists can record when a task has been completed, such as with
personnel signoff on each check point. This helps avoid false memory of
task completion after being interrupted. This is sometimes called “place
keeping”.
• Logs can help record information and reduce reliance on memory.
• Hold Points or Stop Points can check that a task sequence has been
completed before proceeding further.
For example:
Everyone has a limited attention span. The timeframe that people can maintain
attention and concentration varies among people and between tasks. Some of the
factors influencing attention spans are illustrated in Figure 17-5.
Attention and concentration can fail after about 15 to 20 minutes or even faster
in situations of low task demands, especially if:
These activities are long, do not require significant action and require
monitoring of an unchanging situation.
One of the risks of losing attention is that a person may not realize that their
attention has lapsed. This may reduce the ability to recognize a loss of attention,
and therefore reduce the likelihood that they will take corrective action.
Motivated people performing engaging tasks can maintain attention for longer,
possibly a few hours, especially for diverse tasks. However, a long high demand
task may actually exceed peoples’ ability to maintain concentration.
200 Human Factors Handbook
Shorter Longer
Attention span
The potential attention span should be considered. Where there is a potential for
loss of attention during a task, some tactics for maintaining attention are noted in
Table 17-3.
Many of these tactics aim to either enable people to take a break from a task
before they lose attention or increase their stimulation levels by factors such as
task or environmental enrichment.
For low demand, task requirements may be created to keep people engaged.
For example:
The scheduling of additional tasks needs to ensure that they do not distract
from the primary task. For example, the secondary task may be short or performed
with the primary task still within the visual field of the primary task, such as
completing a log at the same workstation.
greater sleepiness during the daytime. Tasks requiring high levels of attention may
best be scheduled for periods of alertness, such as 08:00 to noon, and not
scheduled for periods of sleepiness, such as during night shifts.
Task enrichment
Task sharing
Redesign the task to increase the
Switch tasks between people
level of stimulation
Ambient environment
Maintain temperature (e.g., around Lighting
65 oF to 72oF/18 oC to 22oC) and Higher levels of ambient lighting
humidity
Shift design
Task scheduling
Adopt good shift design to
Schedule tasks requiring attention
minimize fatigue
to higher energy times of the
See Chapter 15 for more
circadian rhythms
information on shift design
Alert or alarms
Automate high attention tasks
Use to reduce demands on
Via control systems
monitoring
202 Human Factors Handbook
Control of work packages are frequently used multiple times a day. This creates
a potential for people to perceive control of work packages to be too detailed or
unnecessarily prescriptive. If the control of work package is perceived as being too
detailed, this may reduce its acceptance. If the control of work package repeats
generic safety requirements this may also cause people to think they are not
needed.
17. Error management in task planning, preparation and control 203
Good Human Factors practice includes ensuring that control of work packages
are used by competent people as part of safety management. It also includes:
• Having different levels of work control and task verification, with simpler
work packages for lower risk or less complex tasks, and more detailed
work packages for higher risk or more complex tasks.
• Using task specific assessment, including error assessment, to develop
the task plans.
• Indicating Hold or Stop Points, Checks, and key safety actions.
• Using quantitative (numbers-based) criteria for accepting test results
and/or stopping an activity.
• Avoiding creating or using generic safety requirements and procedures –
requirements and procedures must be specific to the task.
• Avoiding unnecessary detail that people will already know.
• Raising awareness that Permit to Work should be one of the safety
management systems that must be used properly on all occasions.
• Having a common Permit to Work system across a site and across
functions, to minimize error from factors such as inconsistent layout,
content, and icons etc.
• Ensuring that people who produce and use control of work packages are
competent.
Tool Box Talks, Tailgate Meetings (a team briefing at the rear of a vehicle) and other
forms of team briefings are a standard part of process operations.
• Isolating the wrong part of the system or working on the wrong system,
such as due to poor or no labelling.
• Improvising an unsafe isolation, such as having inadequate or
inaccessible blind points, or failing to note blind points on process
diagrams.
• Failing to test the system state, such as not testing:
o Pressure levels.
o Residual content.
o Purging effectiveness.
While isolation is a frequent and highly practiced task, it can sometimes present
increased risk of error. For example:
As shown by the incident in Table 17-4, experience alone does not ensure
correct performance. Even an experienced operator makes mistakes. It is
important for everyone to check for zero energy – no exceptions.
17. Error management in task planning, preparation and control 207
A pig trap was not depressurized before attempting to remove it from a pipe.
When the trap door was opened, a sudden release of high-pressure gas caused
the door to be blown 30 feet (10 meters) across the deck, through two handrails
and overboard.
A worker was injured (the Injured Person - IP) and treated for facial lacerations.
Check the pressure indicator and ensure ‘zero energy’ before opening a pig trap
door. Breaking containment is a high-risk activity.
When opening a pig trap door, ensure that staff members are positioned
outside of the direct ‘line of fire’ in order to prevent fatality.
• Higher risk isolation should be locked off with secure key control and (as
applicable) removal of actuating devices.
• Lower risk isolations should be locked off with the team controlling the
keys.
If an existing system does not meet the good practices in Table 17-5 and it is
impossible to improve it, then problem locations should be identified, documented
and communicated in team briefings and instructions.
Interlocks and automatic trips are important and frequently used methods of error
management. They are methods of stopping errors and mistakes from causing an
incident. For example, they can reduce the potential for someone to open a
drainage outlet on the wrong vessel (that is full of product).
Systems and equipment must be locked out before working on them. In some
cases, people may not realize that a system should not be operated or is in an
unsafe state. For example:
• Confusing two similar storage tanks (one full and one empty). Operating
on the full one, while thinking that the interlock can be safely defeated
because the tank is empty.
• Incorrectly thinking the system has been isolated and purged, due to a
communication error or starting work prematurely.
These mistakes may lead to someone attempting to work on the wrong system,
without realizing their mistake. They then, in good faith, defeat the interlock to
allow them to complete their work.
Past incidents show that, in some cases, it can become routine to bypass or
defeat interlocks. Some of the reasons for doing this may include:
Robust methods of system and equipment interlock, tags, and lockout are
essential. Guidance on interlocks and their management is included in the
International Standard EN ISO 14119:2013 [70]. This notes, for example, that there
should be clear procedures describing the conditions where bypasses are
acceptable, any approvals required, and time limits for bypasses. The use of
bypasses should be infrequent and for a limited duration of time.
Some Human Factors guidance is given in Table 17-7. These ideas aim to
minimize the conditions for incorrectly defeating interlocks, specifically to:
Table 17-7: Human Factors good practice for interlocks and trips
Human Factors good practice for interlocks and trips
1. The removal of an interlock should only be allowed as part of a Permit to
Work or equivalent management control procedure.
2. The activation of trips should be clearly stated or enunciated and
recorded in system logs.
3. Compliance with interlock rules and procedures should be a mandatory
rule.
4. Faulty or unreliable equipment that causes frequent process interruptions
and “workarounds” should be fixed or replaced.
13. Interlock key access systems should be designed so that it is obvious that
keys are in use.
17. The testing of an interlock after its maintenance, to verify its effectiveness,
should be a mandatory and permit controlled activity.
18. The routine defeat of interlocks and trips should be recognized as a sign of
declining safety standards and acted upon immediately.
• Realistic tasks plans should create the conditions for successful task
performance and anticipate high risk situations.
• Many common elements of process safety management are also key
aspects of error management, including Permits to Work (work
instructions), team briefings, interlocks, and isolation procedures.
• These elements of process safety management need be managed to be
highly reliable, resistant to human error, and meet good Human Factors
practice.
Human Factors Handbook For Process Plant Operations: Improving Process Safety and System
Performance CCPS.
© 2022 CCPS. Published 2022 The American Institute of Chemical Engineers.
error
Proactive detection and correction of errors prevents accidents and leads to faster
recovery. Limitations to individual cognitive capability can be enhanced by
effective teamwork, including checking and error reporting. Teams that detect,
report, and learn from errors are high performing, resilient teams.
The causes of the draining pump error probably lay within three main Human
Factors. These are as follows:
Limited When individuals have completed the same task successfully on repeated
self- occasions, they become complacent, believing everything will be fine, as
scrutiny it typically has been.
The success of error coaching depends upon the attitude of the recipient. A
defensive response can shut down the error reduction effort.
18. Capturing, challenging and correcting operational error 219
An example of error training from the process industry [75] is shown in Table
18-2.
8. The core concept of the course was “creative mistrust”, where the emphasis
was on constantly being aware, continually improving safety performance,
and anticipating potential issues. Trainees were taught how to think
through tasks and identify where errors can occur.
9. The program received excellent feedback from employees. The feedback
was based upon surveyed data and anecdotal evidence.
220 Human Factors Handbook
Many errors could be prevented if individuals were more aware of their own and
others’ actions. Table 18-3 shows examples of observable behaviors that are likely
to lead to errors. Knowledge of these behaviors could help workers to prompt
themselves and each other that something is amiss, to regain focus and situation
awareness.
Error-prone situations, coupled with time pressure and task overload, and
error inducing factors such as fixation, stress, trepidation, or confusion can lead to
unsafe behaviors. This can result in slips and lapses as shown in Figure 18-4. This
is an example of a simplified error taxonomy used for training operational
personnel. Educating or training people about factors contributing to errors,
makes them more alert to errors and consequently more likely to:
High risk or
complex tasks
Error condition
SLIPS LAPSES
222 Human Factors Handbook
It can also help people develop mental resilience for dealing with the
consequences of their errors in operational settings. Mental resilience means
being able to recover quickly from difficult situations. Practicing recovering from
error can help develop self-confidence, and thereby help people maintain focus
after making an error.
Training on error management targets three core cognitive skills groups [76]:
• Information management.
• Planning and mental simulation.
• Monitoring and evaluation.
Information
management
Information gathering and search
Plan formulation
Systematic decision-making
Self-monitoring
Monitoring and
Systematic scans and checks
evaluation
Divergence detection
224 Human Factors Handbook
Training on error prevention also includes building resilience. That is, the ability to
recover quickly from difficulties, otherwise known as “mental resilience”. Resilience
training is based on the idea that operators can be helped to bounce back from
difficulty or change. They can also be helped to learn to cope better with the
demands around them. This is achieved by a slow, step-by-step introduction to
complex, high hazard, life-saving, or threatening scenarios.
Trainees start at lower level or less stressful scenarios. For example, they
complete a well-rehearsed evacuation procedure within 10 minutes. They then
move to high-level or more complex tasks. For example, they have to complete an
evacuation procedure in new or unexpected circumstances within two to three
minutes. The training aims to develop individuals’ resilience and their ability to
cope with highly stressful situations, by slowly increasing the challenge and by
allowing them to recover from error in a safe environment.
Psychological safety
Humans are prone to errors. Using error detection tactics may reduce the
potential for error and safety non-compliance occurrence. Examples of these
techniques are noted in Table 18-4.
As noted in the table, the error detection techniques rely on teamwork and
verification by others. It is more likely that someone will spot aspects that have
been previously missed, if one or more other individuals review the task. It also
ensures that the cognitive limitations of one individual can be mitigated or “offset”
by that of another team member. This helps to prevent errors due to impaired
situation awareness, memory lapses, and incorrect perception of the situation.
These techniques do not replace self-checking, which should be conducted in
parallel to the peer or team-checking techniques noted next.
• Feeling able to report one’s own errors and concerns, without fear of
repercussions.
If an error occurs, using neutral language such as “What happened” and “How did
that happen” can help people speak about events without fear.
Figure 18-7: Challenging skills
230 Human Factors Handbook
It is not possible to prevent all errors. Some tips on how to recover from error in
an operational setting are shown in Table 18-5.
A heat exchanger had become very cold, after loss of lean oil flow, and was
leaking. The intention was to recommence the flow of warm lean oil. A production
coordinator realized the flow rate of warm lean oil into the cold heat exchanger
needed to be reduced to prevent brittle fracture. Brittle fracture occurs when
metal is exposed to a temperature below its minimum design temperature and is
then pressurized without warming above its minimum design temperature.
This was a very complex event with a series of equipment failures, errors, and
mistakes. It escalated over about five hours. There was a high volume of
communication between dispersed personnel by radio. They were communicating
safety critical information. As the event escalated, actions should have been
performed quickly.
People in process plants often communicate with one another from different
locations, such as from a control room to a remote part of the site. People may be
working in an area with high levels of noise from equipment and may be wearing
hearing protection.
Words and sentences can be partly obscured by “radio noise” or weak signals.
The recipient may mishear what is being said or incorrectly “fill in” the missing
words.
“B” and “P”. This means that if, for example, valves are referred to a valve B123 and
valve P321, the B and P may be misheard.
Unclear articulation, ambiguity, and using words with double meanings can
each contribute to miscommunication.
19.4.1 Overview
Safety critical task analysis can identify which tasks are safety critical, and which
tasks involve communication of safety critical information (See Chapter 1).
Common examples of safety critical communications include:
One good practice is to use the NATO/ICAO [81] phonetic alphabet and number
pronunciation, such as Delta and Bravo for D and B, and NINER for 9. These types
of communication protocols are commonly used in aviation, emergency service
and military settings.
Type of
Don’t Do
communication
Articulation Enunciate poorly. Articulate clearly.
Rate of speaking Speak rapidly Speak at a moderate pace
This example uses the
Communicating NATO/ICAO phonetic
item of Use everyday letter and
alphabet and number
equipment to number sounds.
pronunciation.
operate “Valve D B fourteen.”
“Valve Delta Bravo one
four.”
All critical Assume you have been heard Ask for the message to be
communication correctly. repeated back to you.
Communication Use the 12-hour clock. Use the 24-hour clock.
of times “Four o’clock” “Sixteen hundred hours.”
Use words with two or more
meanings.
Use words with only one
Words with “Please give me a ‘conservative’
meaning.
double estimate of the flow rate.”
“Please give me an estimate
meanings This can be interpreted as a
of the maximum possible
request to estimate the
flow rate.”
maximum or a low rate of
flow.
Precision Use imprecise terms. Use precise terms.
“The tank will be full ‘soon’.” “The tank will be full in two
to three minutes from now.”
Type of
Don’t Do
communication
Use words in a sentence
Use words in a sentence that that if partly heard cannot
if partly heard have a be misunderstood.
different meaning. “Pipe Victor Charlie one
Use of words in “It is not safe to open pipe eight is pressured. Keep
a sentence Victor Charlie one eight outlet.” valve closed.”
If the word “not” is obscured If “pressured”, “keep” or
the recipient may hear “It is “closed” are obscured, the
safe to open the pipe outlet”. sentence makes no sense
and must be repeated.
Be unambiguous and
Be oblique or circumspect.
Communicating clear.
“I think you might want to
specific “The level in naphtha tank
check the level on naphtha
warnings November Tango four two
tank November Tango four
has exceeded the high-level
two.”
alarm.”
Be unambiguous and
Be oblique or circumspect. clear.
Communicating “I think there may be a “There is a major fire at
an emergency problem with naphtha tank naphtha tank November
November Tango four two.” Tango four two. This is a
major incident.”
• Having a word and/or time limit for each safety critical communication,
such as 15 words or 30 seconds.
• Requiring long messages to be chunked, with each chunk recorded or
logged before saying the next chunk.
• The sender starts by saying the receiver’s name and then states
their message.
• The receiver repeats the message back.
• The sender confirms the accuracy of the repeat-back or repeats the
message if it is not accurate.
• The sender and receiver must say “ACKNOWLEDGED” each time they
receive a message, to be sure that they have heard and recognized the
message.
• If the repeat-back is wrong, say “THAT IS WRONG”. If the repeat-back is
right, say “THAT IS CORRECT”.
• State “SAY AGAIN” if the message is not clear.
• Say “ALL”, “FROM”, “BEFORE”, to indicate what should be repeated.
• Say “CORRECTION” if something has been said incorrectly by the person
sending the message.
• Say “OVER” to indicate a message has ended.
240 Human Factors Handbook
The goal is to ensure continuity of safe and effective working across shifts.
Poor shift handover was a factor in the Texas City refinery explosion [14]
summarized in B.1 (page 383). Another example is given in Table 19-2.
Handover time between shifts was unpaid. Staff tried to allow 15 minutes
for handover. This was thought to be too short.
242 Human Factors Handbook
The tank level gauge was stuck, and an independent high-level switch was
inoperative. The tank was being filled by manual control and being monitored
by operators. It overfilled, ignited, and the vapor cloud exploded. The fire spread
to 20 fuel storage tanks.
The level gauge had failed before. It was usually unstuck by raising and
lowering it. Sometimes the sticking of the level gauge was logged by supervisors,
and at other times it was not. There were 14 previous occasions when the gauge
had stuck that had not been logged [38].
The conditions that may create the potential for shift handover error are noted
in Table 19-3. A formalized shift handover process should be developed if one or
more of these risk factors exist during the process.
The main type of error that can occur during shift handover is the omission of
information, such as:
It is common practice to use formalized logs and shift handover forms, either
paper-based or electronic. The specific fields will be process specific. The
elements of an effective handover are summarized in Table 19-4.
The handover process should include information such as the reasons for
temporary bypasses, process state, and equipment faults. Good handover can also
include a checklist, especially those that highlight how the operating state of the
plant has changed. Failure to include relevant information in a clear and open way
will result in a poor shift handover. Other failings that result in poor shift
handovers can include not providing enough time on return to work situations or
poorly selected areas away from process (e.g., in the control room creating
distractions and providing on verbal cues alone).
managers should communicate their view of any issues that the next shift will need
to handle and, in particular, whether these issues limit what the next shift should
attempt to achieve. For example, if the plant is part way through a long start-up
procedure, do they think the plant state at the point of hand over means that start-
up can or cannot be completed in the next shift. Key operational goals and
requirements should be stated and reviewed to ensure a common understanding,
along with the next steps in the procedure.
The United Kingdom’s guide “Managing shiftwork: Health and safety guidance”
[82] is recommended further reading.
Repeat-back
Visibility of critical information Critical information should be
Priority information should be easy communicated verbally as well as in
to see and highlighted. writing and must be repeated back
by the oncoming shift leader.
Face-to-face
Formalized language
A social environment should be
Formal words should be used such
created that supports two-way
as ISOLATED and trip OVERRIDE.
communication and cross-checking.
19. Communicating Information and Instructions 245
Agile thinking follows awareness that the situation has changed or when
someone recognizes new aspects of a situation. It includes changing one’s
understanding of what is happening and changing one’s plans and decisions. It can
also include recognition that a plan of action is not having its intended effect and
that an alternative plan of action is required.
Figure 20-1: Stages of situation awareness
Situation awareness
Stage 2:
Stage 1: Stage 3:
Understanding the
Information gathering Anticipating events
situation
In offshore oil and gas explorations, especially when working in new locations,
drilling teams constantly monitor temperature, pressure, and drilling depth
during drilling operations to anticipate future conditions.
Comprehension
When drilling in oil and gas reservoirs, workers may experience higher
pressure than expected. The solution to their current mud weight calculation
may be related to the source of pressure. If they understand the pressure
source, they may be able to come up with a solution to manage the high
pressure.
Projection
Workers should also look ahead and predict what may happen. For example, if
they are experiencing high pressure at 6,600 feet (2,000 meters) down, then
what could happen when they get to 8,200 feet (2,500 meters)?
The team should make a decision: should they persist with the original plan or
do they need a new plan?
For any team to work effectively they must have “shared situation awareness”
– that is, a common understanding of the event.
252 Human Factors Handbook
Shared situation awareness is important where the task is large and complex,
and where it contains many sub-tasks split between individual team members. An
example of this would be a complex industrial process, where teams of operators
are needed to complete procedures such as start-ups. Process plants, such as
refineries, are also often spread over large areas. Team members and processes
may be physically separate from one another.
Poor situation awareness played a role in the Bayer Crop Science plant explosion,
2008 [83], as summarized in B.2
The newly installed operating systems played a key role in the Bayer Crop
Science plant accident. The new control systems significantly changed the interface
used by the board operator. This directly affected the operator’s situation
awareness. The new system presented many challenges due to the following:
The complexity of the system design coupled with human cognitive limitations
(limitations in perceiving and remembering a large amount of information
simultaneously) led to an incorrect assessment of the situation.
Mica Endsley and Debra Jones in their book “SA Demons: The Enemies of
Situation Awareness” [84] identified eight causes responsible for failures in
situation awareness. Those were termed as demons of situation awareness and
include:
Individual’s attention is
focused on and biased • Failure to review a
to information that decision that has
confirms the current already been made, in
Confirmation bias interpretation of the light of new information,
event, neglecting because the new
information that does information does not fit
not fit in. People “see the initial perception.
what they want to see.” • Incorrect action taken.
Tendency to recall
solutions from • Causes are not
Similarity bias situations that appear investigated.
similar to past • Incorrect decision and
experience. action may be taken.
When making a
judgement of the
likelihood of an
Representativeness
event, their • Assume a process upset is
heuristic
judgement is based caused by a similar reason.
on its similarity with
a common reason.
Continue investing
resources (time and
effort) into a course
of action that is • Missing vital clues from the
failing, as environment.
individuals believe • Important information may
Escalation of that with just a little be discounted and resources
commitment more time, they can misdirected.
figure out the • Incorrect/inappropriate
problem. They do decision and actions.
not wish to be seen • Losing track of time during
as inconsistent or to time critical actions.
waste previous
effort.
256 Human Factors Handbook
The cognitive biases noted in Table 20-1 are used subconsciously by individuals
to:
These cognitive biases can also lead to inadvertent errors. For example, the
“Escalation of commitment” bias can cause people to continue with a plan of action
despite information indicating it is ineffective or wrong. This is the opposite of agile
thinking.
• Handle and assess real-time data, process parameters, and alarms all at
the same time – individual situation awareness.
• Co-ordinate between field and control-room operators – which requires
shared situation awareness.
The decisions and actions operators take are of great importance as they can
either ensure smooth operations on the plant or lead to process upset and even
accidents.
o Information gathering.
o Comprehension.
o Anticipation of future events.
21.2.1 Overview
Training in situation awareness and agile thinking can form part of wider non-
technical skills training. This is sometimes known as Crew Resource Management.
Crew Resource Management helps a crew to work together and perform well.
Training guidance available to process industry roles is included in The
International Association for Oil and Gas Producers (IOGP) Report No. 502
“Guidance for Implementing Well Operations Crew Resource Management
Training.” [85]
The content of the training would include topics such as task management,
recognition of critical cues, development of comprehension, projection, planning,
information seeking, and self-checking activities [86].
Behavioral markers are used across many industries to describe the expected non-
technical behavior. Behavioral markers are descriptions of expected behavior that
can be used to measure how well an individual demonstrates that behavior in their
work. Behavioral markers should be industry and job role specific, as task
requirements for situation awareness vary. For example, control room operators
should maintain awareness of control panels and alarms, while engineers may
focus on process operations. Behavioral markers can be used to:
The IOGP Report No 502 [85] recommends refresher training at least every
three years. Employers may, however, consider more frequent refresher training.
Given the importance of situation awareness in many safety critical tasks, advice
can be offered on how to minimize the risks of reduced situation awareness.
Human performance tools are useful and can be used to reduce human error and
lead to various positive outcomes, such as:
These tools can also help foster agile thinking. For example, the Dynamic Risk
Assessment includes reviewing the effect of actions and adapting these.
Table 21-2: Human performance tools – examples
Table 21-2 continued
Human
performance tool Description Usage
(HPT)
STOP and seek This technique promotes awareness of workers’ knowledge limitations as applied to Especially when
STOP when unsure dealing with specific work situations, deviations, or uncertainties. workers operate in
PAUSE when unsure knowledge-based
Workers will seek help, usually from supervisors and/or co-workers, to continue work
modes.
and to deal with these uncertainties and/or lack of knowledge.
Pre-task briefings often follow the S-A-F-E-R pattern:
Human
performance tool Description Usage
(HPT)
Identify critical steps – any actions that will trigger immediate, intolerable, and Important in safety
Identify critical steps irreversible harm. Once critical steps are identified, workers can anticipate errors that critical tasks (high-
can occur at each critical step, estimate their consequences, then evaluate the risk).
existence of controls, contingencies, and stop work criteria.
Used in safety
A workplace risk assessment carried out by supervisors at the point of work, prior to critical, high-risk,
Point of Work Risk the start of an activity. It is used to identify those things, situations, processes, and complex tasks
Assessment (POWRA) activities that may cause harm to people. One completed POWRA form can apply to needing multiple
the whole team. checks.
Human
performance tool Description Usage
(HPT)
Stop-Think-Act-Review (STAR) assessment is often part of a Dynamic Risk
assessment. Following is a description of the S-T-A-R steps:
Stop
STOP
• Look for hazards
• Review hazards
• Has the situation changed?
Think
When operating in
Dynamic Risk
• Evaluate the situation skill-based and rule-
• Evaluate options based performance
Assessment
modes.
Particularly effective
Act
Act for repetitive tasks.
now • Apply safety measures
• Recommence the work
Review
Human
performance tool Description Usage
(HPT)
• Stop (or slow down) – pause to focus attention on the immediate task.
• Think – think methodically and identify the correct actions to perform. Consult
Dynamic Risk with others if further information is required. Understand what will happen
Assessment (cont’d) when a correct or incorrect action is performed.
• Act – perform the action.
It is important to recognize:
The main causes of situation awareness failures at different levels (L) are shown
in Figure 21-2.
When cutting a hole in a tank; the worker did not continuously monitor the
oxygen content in the tank. The oxygen in the tank was consumed by the heat from
the cutting operation and the worker passed out. The worker should have been
aware that oxygen could drop and should have had a ventilation fan providing
fresh air or monitoring CO/CO2.
The next few sections outline decision-making and agile thinking traps and provide
techniques on how to avoid these. These techniques aim to help people to
recognize and consider alternatives actions, to review these alternatives, and to be
open to changing plans and decisions.
Group-think can lead to poor decision-making. The group may ignore information
that contradicts their understanding. Individuals may self-censor. For example, if
they think a mistake is being made, they may keep this view to themselves.
If individuals are aware of confirmation bias and accept its existence, they are
then able to consciously avoid the biases. Confirmation bias affects people in four
primary areas of cognition. Table 21-5 provides examples of behaviors associated
with confirmation bias.
274 Human Factors Handbook
Some people are uncomfortable with changing their opinions, especially if they
feel they will be criticized for their initial opinion or decision being “wrong”. A
culture of trust and shared purpose is important. People need to feel that they can
change their opinions without fear. A common commitment to a shared purpose
can help people prioritize effective decision making over the avoidance of
individual fear about changing opinions and plans.
o 20 second scan;
o STOP and seek, STOP when unsure; and PAUSE when unsure;
o Pre-tasks and Post-tasks briefing;
o Point of Work Risk Assessment (POWRA);
o Dynamic Risk Assessment;
o Stop-Think-Act-Review assessment;
o Verbalize – Point – Touch;
o Shadow Boards.
• Group-think and other cognitive bias are common biases and should be
avoided. This is because they can contribute to poor decision-making
and may lead to severe consequences.
• Coaching, training and behavioral markers should also promote the
importance of changing opinions and decisions when new information is
received, if a plan of action is ineffective or when the situation changes.
Human Factors Handbook For Process Plant Operations: Improving Process Safety and System
Performance CCPS.
© 2022 CCPS. Published 2022 The American Institute of Chemical Engineers.
On July 24th, 1994, a large explosion occurred at Texaco Refinery, Milford Haven in
Wales, which caused injury to 26 people [87]. The blast from the explosion
damaged properties in a 10 mile (16 kilometer) radius and was heard 40 miles (64
kilometers) away. The site suffered severe damage to the process plant, the
building, and storage tanks. A summary of the event is given in B.4 (page 389).
During the sever electrical storm that proceeded the explosion, operators and
operations management failed to identify the underlying causes of the problem or
to recognize that they had the potential to lead to hazardous consequences,
despite these data being available to them. They continued to operate in a
disturbed environment for five hours prior to the explosion. All the information,
including alarms, was available to the operators via six distributed control systems
(DCS) screens, which were used to control the process and to diagnose faults.
Many alarms, in the plant were sounding simultaneously, all with the same -
high priority. In the 15 minutes before the explosion, operators were receiving
alarms at a rate of one every two seconds. Thirty minutes before the accident, a
critical alarm went off. Had the operators recognized the criticality of the final
alarm and taken appropriate action, the explosion may not have happened.
• A control valve shut when the control system indicated it was open.
• A modification that was carried out without proper assessment of
consequences.
• Control panel graphics that did not provide the necessary process
overview.
• Attempts to keep the unit running when it was supposed to be shut
down.
• Inadequate emergency management.
278 Human Factors Handbook
During the process upset, the actions taken by operators were reactive. There was
no assessment or management of the situation. People at senior level took on
operating roles during the upset. That is, they helped out, rather than taking an
overview of the complete process/overall perspective of the situation. Decisions
were made at an individual level with no coordination among team members.
A key issue identified by the official report into this accident was the
importance of emergency management training. One of the report
recommendations stated that training should include:
It is also important to put the process unit into a safe state or shut it down, in
high risk situation, and then aim to understand the problem.
22. Human Factors in emergencies 279
Error Recognition
What is the problem?
How much time is available?
How risky is the situation (present and future)?
YES NO
Execute
actions
Review
(Adapted from [88]) outcomes
In the Milford Haven Refinery explosion, three human errors were evident:
A refinery explosion occurred on the 21st June 2019 at the Philadelphia Energy
Solutions refinery (see Figure 22-3). The consequences could have been much
worse without prompt action by the control room operator.
At 04:00 am propane and some hydrofluoric acid escaped after an elbow joint
fractured in a hydrofluoric acid alkylation unit. The leaking vapor formed a ground
hugging vapor cloud around parts of the unit. Two minutes later the cloud ignited
causing a massive fire.
The control room operator quickly took steps to prevent the release of
additional hydrofluoric acid by rapidly draining the unit’s hydrofluoric acid to a
vessel designed to hold the acid in the event of an incident.
Hydrofluoric acid when released under pressure can form a toxic aerosol cloud
and travel for miles. This aerosol can immediately penetrate skin and cause
deaths. By draining the hydrofluoric acid, the operator greatly reduced the scale
of the accident. There were no serious injuries, due, in part to the operator’s quick
actions.
282 Human Factors Handbook
A preliminary video from the U.S. Chemical Safety Board had been issued at
the time of publication [90].
Even highly experienced and fully trained operators can experience skill fade.
This can occur due to lack of practice when carrying out low frequency emergency
response tasks.
Repeated exercises and drills promote the acquisition of skills (technical and
non-technical) through repetitive practice, until individuals have reached the stage
where their responses are “automatic”. This is
especially useful for emergency situations, as
Automatic responses are
they are rare and cannot be predicted and
being able to do tasks
people must make decisions in high stress
without having to think
situations. The individuals involved must be
about it.
able to react fast.
22. Human Factors in emergencies 283
Non-technical skills are social, cognitive, and personal skills that enhance the
way individuals carry out technical tasks and procedures. Examples of non-
technical skills include communication, decision-making, leadership, teamwork,
and situation awareness.
Poor non-technical skills can increase the likelihood of error, which can
increase the likelihood of an adverse event. Adverse events can take the form of
costly production failures without causing harm to anyone – for example, getting
a drill pipe stuck when exploring oil, or shutting down a manufacturing process by
mistake. They can also result in more severe consequences that lead to injuries
and fatalities.
The International Association of Oil and Gas Producers (IAOGP) [92] proposed
a basic Crew Resource Management (CRM) syllabus, with a focus on individual non-
technical skills. The proposed non-technical skills are transferable, and the
intention is that individuals can utilize them across a range of different teamwork
activities.
Non-technical
Performance support and error prevention
skills
Situation Awareness
• Enables identification of deterioration in
cognitive skills and interpersonal skills
• Monitors process performance
• Recognizes faults, upsets, and anomalies in
emergency situations
• Recognizes situations where decisions are
needed
• Engages in decision-making process:
• Assesses the situation;
• Considers various options/alternatives;
Decision-making
• Selects and implement most appropriate option;
• Reviews the outcome.
• Recognizes where a different approach to
decision-making is needed
• Recognizes where decisions may be affected by
cognitive bias (group think)
• Manages recognition-primed decision-making
• Enables effective information sharing and
Communication process status verification
• Allows the process of searching for clues
Leadership
• Centralizes decision-making
• Monitors effectiveness of actions, and assesses
the outcome of actions
Performance • Strengthens personal resilience and
influencing factors performance under pressure
(e.g., stress and • Enables coping in unplanned situations and in
fatigue) time-pressured emergency scenarios
286 Human Factors Handbook
The impact of stress was evident in the Piper Alpha disaster (1998). This was a
fast-developing situation. People on the platform were at imminent risk of major
injury. The situation was in many respects, uncontrollable as well as confusing and
complex. The Offshore Installation Manager (OIM) was the individual responsible
for organizing the response to the emergency. Evidence from the report [93]
highlighted the inability of the OIM to make critical decisions in this situation of
stress. Some excerpts from the report are shown next.
“The OIM had gone a matter of seconds when he came running back in what
appeared…to be state a panic…The OIM made no specific attempt to call in
helicopters from the Tharos (a rescue vessel) or elsewhere, or to communicate
with the vessels around the installation, or with the shore or other installations;
or with personnel on Piper…” (para 8.9 [93, pp. 152-153])
“The OIM did not give any other instructions or guidance. One survivor said
that at one stage people were shouting at the OIM and asking what was going
on and what procedure to follow. He did not know whether the OIM was in
shock or not, but he did not seem to be able to come up with an answer.” (para
8.18 [93, pp. 156-157])
Modeling
Learners are given the opportunity
to observe another team in a
stressful situation, and to then
assess the effectiveness of their
performance.
Individuals need to be aware of the loss of their own and/or others’ situation
awareness. To stay in control of a situation, various types of information need to
be processed. Shared situation awareness requirements [94] are shown in Table
22-3. In emergency situations, individuals should be aware of:
Useful in
emergency
situations to
DODAR provide the
DODAR is a cyclical model of decision-
steps of dealing
making, consisting of the following steps:
with abnormal
situations.
• Diagnosis – What is the problem?
• Options – What are the options?
• Decisions – What are we going to
do?
• Assign the tasks – Who does what?
• Review – What happened? and
What are we doing about it?
292 Human Factors Handbook
Aid or
Definition Use
mnemonic
This is a method that uses diagrams focused
on cause and effect. It also shows potential
blockers to solutions.
Evacuate
Delegating and communicating are two non-technical skills that are crucial in
emergencies. These skills are also vital for individuals assuming a leadership
296 Human Factors Handbook
position. Table 22-6 show how these two non-technical skills translate into practice
in abnormal situations.
The United States Chemical Safety Board reported that, on November 9th, 2010, an
explosion occurred at E.I. DuPont de Nemours and Co. Inc. (DuPont) Yerkes
chemical plant in Buffalo, New York [98]. This explosion occurred when a contract
welder and foreman were repairing the agitator support atop an atmospheric
storage tank containing highly flammable vinyl fluoride (a gas).
This incident had multiple causes. The Chemical Safety Board report noted that:
There were issues with the role of contractor management, which are
discussed next.
The contractor submitted a “hot work permit”. However, the section of the
permit which asked if flammable material would be within 35 feet (10 meters) of
the work was not completed. The hot work was within 35 feet (10 meters) of the
slurry flash tank that vented vinyl fluoride to the atmosphere.
“The contractors were unfamiliar with the Tedlar® process and the process
equipment involved. The contractors did not know what the slurry flash tank
was or which chemicals were present inside it.” p10, CSB [98]
The permit was signed off by the DuPont construction engineer and by the area
manager. It was reported that:
• The DuPont construction engineer for the slurry tank work had no
working knowledge of the Tedlar® process.
• The construction engineer expected the area manager would advise the
contractors of plant-specific process safety information for hot work.
• An area manager signature was obtained by someone in a service
department. The area manager lacked knowledge of the area and of the
Tedlar® process. In addition, they did not perform the required “walk
down” of the area before signing the permit.
• The area manager assumed the construction engineer was briefing the
contractors on-the-job and the hazards.
23. Working with contractors 303
“The contractors … were allowed to complete the hot work permit and begin
hot work without getting approval from any DuPont employee knowledgeable
about the process.” p11 CSB [98].
“…the potential lack of familiarity that contractor personnel may have with
facility hazards and operations, pose unique challenges for the safe utilization
of contract services”. CCPS, [24].
Working with contractors creates some Human Factors risks. These include:
The client organization personnel should recognize the risks, and proactively
offer support to help contractors perform tasks successfully. It is important to
verify the activities and stop points with contractors. The host employer (Company)
should ensure that the work is inspected and the work plan is being followed.
304 Human Factors Handbook
Some good practices that can help with the Human Factors of working with
contractors are outlined next.
1. To what extent are they familiar with the site, its hazards, job and task
specific hazards, safety management arrangements, and procedures?
What actions will they take in case of an incident?
2. Are they familiar with availability and functionality of controls?
3. Are they familiar with performance expectations, and operational and
safety standards?
4. To what extent do the contractors feel psychologically safe to challenge,
report problems, and speak up about safety concerns?
5. To what extent do the contractors feel they are part of a “one site, one
team” approach, with an expectation to share information, and to
coordinate and communicate with one another?
The joint assessment of jobs and shared task planning can help to ensure a
shared understanding of plans, roles and responsibilities, objectives, and risks. It
can also help to foster a “one site, one team” approach.
Mobilization
An explicit “readiness to commence work” review can help confirm that the
contractors:
The last step before signing a Permit to Work should be a field visit of
Operations personnel together with the person responsible of the contractor crew
to check the real conditions before any activity starts. This provides a last
opportunity to identify any unusual situation or hazard.
Coordination
Demobilization
An explicit demobilization activity can help to ensure that site restoration and
reinstatement is performed without omission or miscommunication.
23. Working with contractors 307
The Energy Institute has published guidance that covers the management of
major organizational changes entitled “Managing major accident hazard risks
(people, plant and environment) during organizational change” [99]. This is a
detailed and comprehensive guide. It covers the Human Factors process of change
management and provides checklists and assessment tools.
This Chapter focuses on operational level changes rather than corporate level
organizational change.
This Chapter does not cover the mental health aspects of change, or the
management of laying people off or terminating their employment.
“Front line” operational level changes mean any change that may occur within an
operation/ production, maintenance or logistics department, that may affect
human performance. This may include:
Many of the major accidents cited in this handbook were, in part, caused by
changes in process plant, staffing, or ways of working. The following examples
include introducing a new control system, increased production and alteration of
piping.
performance
Some changes and their potential impacts are more obvious and immediate.
These include the following examples:
• The relocation of staff from one part of a site to another may reduce the
level of contact between two teams. This can slowly erode the level of
teamwork and create an unintended obstacle for seeking help.
Do Do not
Most MoC procedures involve a number of steps as shown in Figure 24-2. Front
line managers and supervisors should be involved at every stage of this process.
Impacts
The Energy Institute guidance “Managing major accident hazard risks (people,
plant and environment) during organizational change” [99] provides a simple tool
for rating risk levels and identifying potential impacts.
24.5.2 Actions
Many changes can cause uncertainty among people, which can lead to stress.
This is more likely to happen if the change alters their role, adds new tasks, or
changes the people they work with. The potential for stress may also be related to
individual factors. The potential for stress, and its impact on performance, should
be recognized and proactively managed. This may include consulting people and
being able to recognize within them the signs of stress, uncertainty, or concern.
Talking to people to understand their concerns is vital. Examples of how to help
people include:
Frontline people will often recognize risks that have been overlooked. A
psychologically safe environment is required – this includes ensuring that people
are actively engaged. It also means ensuring they are asked to speak up about
potential risks and the adequacy of change plans. This will help to reassure people
that all risks have been recognized and are being managed, especially if plans may
change in response to feedback. Error precursors, such as, Time Pressure or Late
Information may influence the thoroughness and accuracy of the change effort.
Refer to section 3.3 for common precursors to error.
An option, especially for higher risk changes, is to formally audit and track
selected indicators.
24. Human Factors of operational level change 317
For example:
• Signs of stress and fatigue may indicate that new lower staffing levels
are too low.
• Low morale may indicate dislike of new roles.
• An increase in error may indicate a new control system is confusing.
This Chapter builds on the previous two Chapters about learning from error. By
the end of this chapter, the reader should be able to:
The CCPS “Guidelines for Risk Based Process Safety [5]” cites “Measurements
and Metrics” as an element.
It also notes the need for developing, managing, and using metrics.
This Chapter does not cover safety culture nor culture-related indicators, such
as safety culture surveys. These are covered by several CCPS guidance documents
[5] [101] [102], which also cite many Human Factors metrics.
The concept of leading and lagging indicators is well known. Most organizations
will already have measures in place for these indicators, which are largely
quantitative and applied at organizational level. Some examples of leading and
lagging indicators are shown in Table 25-1. The indicators are used to identify
issues (spot problems), and to prompt actions to improve performance.
322 Human Factors Handbook
• What are the critical tasks in a specific area of operations? For example,
blinding, handling a process upset, correct maintenance, correct fault
diagnosis.
• What are the key factors (e.g., competence, job aid, tools and physical
environment) influencing the performance of people in safety critical
tasks?
• What demands (job requirements) are placed upon people in
operational areas? For example, shift work, or tasks that are lengthy,
complex, or novel.
• What indicators can be used to demonstrate adequate performance of
these tasks?
Safety critical
tasks
Performance indicators in
safety critical tasks
324 Human Factors Handbook
People can provide feedback about the support they get when performing process
operational tasks. This helps improve performance. Individuals should provide
feedback for the “right” reasons (because they want to) rather than for the “wrong”
reasons (because they were told to).
• Adequacy of support:
The process of giving and receiving feedback is also called a “feedback loop”.
This is because the process involves gathering feedback on a performed action,
analyzing the feedback, and (if necessary) correcting the actions and gathering
feedback once more. The process of receiving feedback and using it to improve
performance is shown in Figure 25-2.
.
326 Human Factors Handbook
Action
Collect feedback
Review feedback
Positive Negative
feedback feedback
Monitor effectiveness
Improvement
of action
Operational debriefs provide rich information on what was done well versus what
could have been done better. Operational debriefs can look at the execution of the
tasks and at the non-technical skills exhibited during tasks.
Operational debriefs can take place after doing a process start-up, and after
process upset or abnormal events, for example. They should reflect on:
• Signs of fatigue:
o A team member demonstrates slow reflexes, tiredness, and
impaired judgment.
• Signs of stress:
o A team member struggles to maintain focus or to remember
information.
o A team member demonstrates low levels of confidence or
morale.
Environmental demands
Job demands
Not enough time to complete the task.
Insufficient training for the job. Stress Stress response- Behavioral
Boring/repetitive work. The working
mitigating symptoms consequences
environment – (e.g., shift work, temperature
extremes, noise).
Control
Lack of control over work activities –no • Prior • Cognitive
involvement in decision making, no control experience impairment
over pace of work, etc. • Training • Behavioral • Emotional
• Practice • Emotional imbalance
Supervisor/Manager • Social support • Somatic • Impaired
Lack of support from supervisor or manager. • Thought-based performance
• Coping
strategies • Undelivered
outcomes/ output
Role
Lack of clarity about responsibilities.
Uncertainty about work objectives.
Change
Fears about job security. Restructuring of job (Adapted from Cooper et al. 1988, stress at work [124])
role.
25. Indicators of human performance 329
• Task completion:
• Effectiveness of audits:
• Lessons learned:
Psychological safety is an important indicator of safety culture and allows for free,
open discussion about human performance. Indicators of psychological safety
include:
supervisor and/or manager may also implicitly show respect to the worker and
may also convey that his work is important. Also, for some workers, they are less
likely to escalate concerns in a large group, however, task safety observations
afford a more natural environment in which to escalate concerns since the worker
can directly show the supervisor and/or manager the challenges he faces in the
field.
Signs of mindfulness are shown in Figure 25-4. Mindful individuals exhibit alert
and perceptive behavior towards hazard and risk identification (chronic unease)
and they respond calmly in emergency situations, regularly engage in self-
reflection, exhibit natural curiosity by asking open questions, and perceive error
as a learning opportunity.
25. Indicators of human performance 331
Alert and
perceptive
to
identifying
risks and
hazards
Learning Calm
from error response
Mindfulness
Self-
Curiosity
reflection
• Work one task at a time – when multitasking (i.e., doing various things at
a time) individual’s focus switches back and forth and loses important
information.
• Take time out – when a person feels overwhelmed with things or a
situation, they should step back for few minutes and clear their mind
start the task again.
• Be fully “present” at meetings - individuals should be fully present
(physically and psychologically) and actively listening to what others are
saying and contributing to the discussion.
• Mindfulness scan (i.e., bring mind back into focus) – it is important that
individuals take regular short breaks (taking a minute or two break from
work) to allow them to re-assess the situation.
332 Human Factors Handbook
Lessons learned should be fed back into the wider organization, in order to
show others how to benefit from the experiences. In particular, where other
business units may experience similar issues or be exposed to similar error traps.
The effectiveness of lesson sharing should be evaluated, and if actions arising from
the lessons learned are appropriate to another unit or team, they should be
implemented. The action implementation cycle shown in Figure 25-5 should result
in improved company-wide performance.
Reflect on performance
Implement actions
25. Indicators of human performance 333
Error is a symptom, not a cause. The root causes of error should be resolved
by improving the level and type of support given to people.
o Task criticality.
o Key factors influencing performance.
o Environmental/work demands placed upon people.
o Characteristics that indicate effective performance.
o Operational debriefs.
o Observations of:
• People’s states e.g., stress, fatigue.
• People’s performance e.g., mistakes, tasks being
completed on time.
o Task safety observations of workers doing their actual work in
the field.
This Chapter provides an overview of learning from error and discusses the
importance of Human Factors in incident investigation. By the end of this chapter,
the reader should be able to:
Those elements of culture that affect learning from error are highlighted in this
Chapter.
People often assume that learning occurs automatically once an incident has been
analyzed and lessons have been drawn from it. This thinking excludes the most
important element of learning, which is change. This refers to the application of
learning (behavioral change) into the work environment.
Both understanding and behavioral change need to take place for learning
to be fully accomplished. This will ensure that the lessons learned are robust,
and sustainable across time and changing circumstances.
The possibility of human error occurring exists in every task performed by any
person working in the process industry. Error can happen, but it is manageable.
Incidents and errors offer valuable learning opportunities and lessons. If these
lessons are acted upon, they will help prevent reoccurrence of errors, and also
enable improvement in the way risks are managed.
Several steps are required to achieve learning. These steps are provided in
Figure 26-1.
338 Human Factors Handbook
Reporting
Monitoring
effectiveness of Analysis
improvements
Implementing Planning
improvements improvements
On April 20th, 2010, the Macondo well blew out [107]. Eleven fatalities resulted and
the Deepwater Horizon drilling rig sank and spilled an estimated four million
barrels of crude oil in the Gulf of Mexico. The spill disrupted the entire region’s
economy, and severely damaged fisheries and the eco-habitat.
A summary of key events preceding the blowout is provided in B.6 (page 395).
The technical cause of the blowout was that the cement that was pumped at the
bottom of the well did not seal off the hydrocarbons in the formation. Factors that
increased the risk of cement failure included:
Other blowouts had occurred in the offshore drilling industry around the time
26. Learning from error and human performance 339
of the accident, and recommendations had been made on the application of better
cementing practices [108]. For example, just one year earlier the Montara Oil Spill
(2009) accident had occurred. The inquiry report for this accident [109] noted that
a direct cause of the accident was the defective installation of a cemented shoe
casing, intended to operate as a primary barrier against blowout.
The root causes of the accident noted in the investigation report [110] were
cited as “organizational and safety management failures”, including:
From a Human Factors perspective (and learning focus) it was evident that
reporting systems were weak, which impaired lesson learning. Lessons learned
from a similar near miss (caused by a negative pressure test failure) which
occurred on December 23rd, 2009, in the North Sea [111], were not shared across
the wider organization soon enough.
The United Kingdom Health and Safety Executive was satisfied with the
corrective actions implemented by Shell and Transocean following the North Sea
incident. The Executive also noted that the shortcomings that had led to the
accidents had been addressed [112]. This suggests that the 2009 near miss lessons
may have been shared and applied in the North Sea. The fact that the 2010
Deepwater accident occurred suggests that this learning had not yet been shared
with the Gulf of Mexico site.
Lessons learned in the aftermath of the 2009 near miss and the subsequent
2010 accident, suggest that it is vital that systems to investigate accidents are
appropriately designed. Such investigation systems must be able to identify
340 Human Factors Handbook
relevant key lessons learned from other errors, incidents, and near misses. These
lessons should be communicated to all individuals and teams that do similar work
and/or are exposed to similar error traps.
Even high performing teams make errors. They are viewed as “high performing”
because they learn from errors. That is, they identify root causes of error, seek out
relevant lessons learned, and apply appropriate solutions.
Errors and mistakes are “windows into reality” and offer a unique opportunity
for learning and application of learning into the working environment. Lessons
should be drawn from both negative events (accidents) and positive events
(what went well). Self-directed learning behavior is an attribute of high
performing teams, as shown in Table 26-1.
.
26. Learning from error and human performance 341
• Overload of recommendations.
• Failure to check the effectiveness of implemented actions.
• Individuals may be sensitive about what they did or did not do (when
they should have), and why they acted as they did.
• Individuals may genuinely not remember the exact sequence of events
and actions that occurred. This can be due to a range of reasons
including stress, cognitive overload, or impaired memory.
Lack of
accountability
Blame Culture
Prevents
Impaired
problem
relationships
solving
Employee
disengagement
To get a fuller account of the incident and people’s actions, the following
techniques can be used:
• Interpret behavior
Human error is not the cause of incidents. Error is just a symptom of error-
inducing conditions or preconditions – that, is conditions that allow or cause the
error to happen. Examples include lengthy and difficult to follow procedures,
inefficient training, or conditions that lead to fatigue.
Reoccurrence of error is often associated with “error traps” – factors that make
errors more likely. The presence of “error traps” means that other people are more
likely to make the same or similar mistakes in the same situation.
Addressing root issues makes it more likely that similar accidents with similar
causes can be avoided in the future. The causal pathway including root and
immediate causes and preconditions to error is shown in Figure 26-4.
Fatigue should be evaluated during the root cause analysis. This can be done
by reviewing the work schedules of employees involved in the incident for a few
weeks before the incident, as well as conducting interviews with them.
348 Human Factors Handbook
Start Here
1 Was the individual instructed / influenced Yes The individual acted on the instructions or
to do this by the supervision or other figure
under the influence of an authority figure.
of authority? 1
No
3 Did they understand what was required, and No The individual did not have the capability
did they have the knowledge, experience, skills,
or the resources to meet the expectations.
physical capacity and resources to do it? 3
Yes
4 Yes
Did they intend to act with
The individual made an unintentional error.
company expectations, but made a mistake?
4
No
5 Yes
Were they following custom-and-practice which A custom-and-practice
was common among their peers? had developed among the team.
5
No
6 Substitution test: Could another person with Yes The individual found
the same knowledge, skills & experience have
themsleves in a difficult situation.
done the same thing in the identical situation? 6
No
7 Yes
Is there evidence to suggest they acted to The individual acted to
help self or company to save time and effort? benefit themselves of the company.
7
No
8 Yes
Is there evidence to suggest they This is a special case
intended to cause harm, damage or loss? - always consult Huam Resources.
8
No
Underlying causes
Preconditions
Policy, culture, Immediate causes
Organizational, ACCIDENTS,
design, training, Active failure:
environmental, and INCIDENTS, AND
supervision, and incorrect action or no
psychological BUSINESS UPSETS
operating action.
influences.
procedures.
ERROR
• Underlying causes e.g., work design, shift design, site and corporate
level processes, management systems.
• Preconditions e.g., tiredness, impaired cognition or focus, difficulties
with memory, poor situation awareness.
• Immediate causes e.g., individual action failures, slips, or lapses.
Task characteristics can often be contributing factors to action failures, such as:
Cause
Effect
Fish Bone
Diagrams
Alarm design Ambient noise Response time
Did not
respond
to alarm
1. “What happened?” – develop a diagram that shows the sequence of events in the accident.
2. “How did the incident happen?” – identify failed, inadequate, missing, and effective barriers. This is to
identify risk management measures that should have been in place.
3. “Why did the accident happen?” – create a causation path that identifies immediate causes and related
human failures of failed barriers, pre-conditions influencing the immediate causes, and underlying (root)
causes that created the preconditions.
Agent Event
Barrier
Object
Barrier
Analysis [31]
This handbook does not cover culture (e.g., Safety Culture or Just Culture).
However, many good Human Factors practices recommend a focus on culture and
its underpinning elements. This is because they contribute to a more effective
incident investigation, and they allow for application of lessons learned. “Good”
safety culture also reduces the likelihood of non-conformance or procedural non-
conformance, and increases understanding of risk.
One common pitfall of actions arising from investigations is to call for more
training for the personnel involved even when the incident has occurred as a result
of an action error.
26.7 Learning
Learning from incidents is not an easy process, especially if the incident had
significant consequences such as causing injury, fatality, or damage to property
and/or the environment. Individuals affected by the incident may carry a lot of
resentment or other strong feelings, which can act as a barrier to rational thinking
and taking forward (sharing and applying) lessons learned.
Restorative Just Culture aims to repair trust and relationships after an incident.
It allows all parties involved to discuss how they been affected and decide what
needs to be done to repair harm. The goals of restorative culture are shown in
Figure 26-6 [117].
Emotional healing
Moral engagement
Help to cope with guilt,
Do the right thing now
resentment, etc.
Restorative Culture
Learning from incidents is a crucial element of process safety. The learning does
not stop once the incident investigation is completed (i.e., root causes were
identified and improvements proposed). The lessons learned from error should be
shared and applied to ensure employees’ full understanding of the issues and in
order that change may begin.
.
Human Factors Handbook For Process Plant Operations: Improving Process Safety and System
Performance CCPS.
© 2022 CCPS. Published 2022 The American Institute of Chemical Engineers.
References
[1] Center for Chemical Process Safety (CCPS), “CCPS Process Safety Glossary,” 2021.
[Online]. Available: www.aiche.org/ccps.
[2] United Kingdom Government, “The Control of Major Accident Hazards Regulations
2015,” 2015. [Online]. Available: https://www.legislation.gov.uk.
[4] Health and Safety Executive (HSE), “Performance Influencing Factors (PIFs),”
www.hse.gov.uk, Undated.
[5] Center for Chemical Process Safety (CCPS), Guidelines for Risk Based Process Safety
(RBPS), Hoboken, NJ USA: John Wiley and Sons, 2007.
[8] Center for Chemical Process Safety (CCPS), Essential Practices for Creating,
Strengthening, and Sustaining Process Safety Culture, Hoboken, NJ, U.S.: John Wiley
and Sons, 2018.
[9] Center for Chemical Process Safety (CCPS), Process Safety Leadership from the
Boardroom to the Frontline, Hoboken, NJ USA: John Wiley and Sons, 2019.
[10] Center for Chemical Process Safety (CCPS), “Risk Based Process Safety Overview,”
AIChE/CCPS, www.aiche.org/ccps, 2014.
[11] Center for Chemical Process Safety (CCPS), Recognizing and Responding to
Normalization of Deviance, Hoboken, NJ USA: John Wiley and Sons, 2019.
[12] Center for Chemical Process Safety (CCPS), Human Factors Methods for Improving
Performance in the Process Industries, Hoboken, NJ USA: John Wiley and Sons, 2006.
[13] Center for Chemical Process Safety (CCPS), Guidelines for Investigating Process
Safety Incidents, 3rd Edition, Hoboken, NJ USA: John Wiley and Sons, 2019.
[14] U.S. Chemical and Hazrd Investigation Board (CSB), “BP America Refinery Explosion,”
U.S. Chemical and Hazrd Investigation Board, www.csb.gov, 2007.
[15] National Transportation Safety Board (NTSB), “Loss of Thrust in Both Engines After
Encountering a Flock of Birds and Subsequent Ditching on the Hudson River US
Airways Flight 1549,” US National Transportation Safety Board, Washington, 2010.
364 Human Factors Handbook
[16] The B.P. U.S. Refineries Independent Safety Review Panel, “The Report of the BP U.S.
Refineries Independent Safety Review Panel,” The B.P. U.S. Refineries Independent
Safety Review Panel, 2007.
[17] J. Reason and A. Hobbs., Managing Maintenance Error: A Practical Guide, Burlington,
VT USA: Ashgate, 2003.
[20] Longford Royal Commission, “The Esso Longford Gas Plant Accident, No. 61 - Session
1998-1999,” Government Printer for the State of Victoria, Melborne, Victoria
Australia, 1999.
[21] U.S. Chemical Safety and Hazards Review Board (CSB), “Formosa Plastics Vinyl
Chloride Explosion, Report No. 2004-10-I-IL,” U.S. CSB, Washington, DC U.S., 2007.
[22] Australian Transport Safety Bureau (ATSV), “An Overview of Human Factors in
Aviation Maintenance, ATSB Transport Safety Report, Aviation Research and Analysis
Report – AR-2008-055,” Australian Government, www.atsb.gov.au, 2008.
[23] G. Miller, “The magical number seven, plus or minus two: Some limits on our
capacity for processing information.,” The Psychological Review, vol. 63, pp. 81-97,
1956.
[24] Center for Chemical Process Safety (CCPS), Guidelines for Risk Based Process Safety,
Hoboken, N.J. USA: John Wiley and Sons, 2007.
[25] Center for Chemical Process Safety (CCPS), Guidelines for Writing Effective Operating
and Maintenance Procedures, Hoboken, NJ USA: John Wiley and Sons, 1996.
[26] U.S. Chemical Safety and Hazard Investigation Board (CSB), “Pesticide Chemical
Runaway Reaction - Pressure Vessel Explosion, Investigation Report No. 2008-08-I-
WV,” Chemical Safety and Hazard Investigation Board, www.csb.org, 2011.
[29] Energy Institute, “Guidance on human factors safety critical task analysis, 2nd
Edition,” Energy Institute, publishing.energyinst.org, 2020.
References 365
[31] Center for Chemical Process Safety (CCPS), Bow Ties in Risk Management: A Concept
Book for Process Safety, Hoboken, NJ U.S.: John Wiley and Sons, 2018.
[32] Center for Chemical Process Safety (CCPS), “What is LOPA?,” 2021. [Online].
Available: www.aiche.org/ccps.
[35] Human Performance Oil and Gas group, “Walk Through Talk Through Template and
Guide,” Human Performance Oil and Gas Group, https://www.hpog.org, 2020.
[36] U.S. Department of Energy (US DOE), “Writer's Guide for Technical Procedures
(Archived DOE-STD-1029-92 Chg Notice 1),” U.S. Department of Energy (U.S. DOE),
www.standards.doe.gov/standards-documents, 1998.
[37] M. Beychok, “Petroleum Refining Processes,” Tel Aviv University, 2011; This work is in
the public domain.. [Online]. Available: www.tau.ac.il.
[38] U. K. Health and Safety Executive (HSE), “Buncefield: Why did it happen?,” U.K. Health
and Safety Executive (HSE), www.hse.gov. uk, 2011.
[39] U.S. Chemical and Hazard Investigation Safety Board (CSB), “Key Lessons from the
ExxonMobil Baton Rouge Refinery Isobutane Release and Fire,”
https://www.aiche.org, 2017.
[40] CCPS, Images supplied for publication with permission by a CCPS Project Team
Member, 2021.
[43] International Standards Organisation, “ISO 6385: 2016. Ergonomics principles in the
design of work systems,” September 2016. [Online]. Available: https://www.iso.org.
366 Human Factors Handbook
[45] Engineering Equipment and Materials Users Association, “EEMUA Publication 201
Control rooms: a guide to their specification, design, commissioning and operation,”
Engineering Equipment and Materials Users Association, https://www.eemua.org,
2019.
[46] R. McLeod, Designing for Human Reliability: Human Factors Engineering in the Oil,
Gas, and Process, Massachusetts, U.S.: Gulf Professional Publishing, 2015.
[47] Engineering Equipment and Materials Users Association (EEMUA), “Alarm systems: a
guide to design, management and procurement, EEMUA Publication 191,”
Engineering Equipment and Materials Users Association (EEMUA), www.eeuma.org,
1999.
[48] A. Hopkins, “An AcciMap of the Esso Australia Gas Plant Explosion,” in Proceedings of
the 18th European Safety, Reliability & Data Association Seminar, Karlstad, Sweden,
2000.
[49] Center for Chemical Process Safety (CCPS), Guidelines for Managing Process Safety
Risks During Organizational Change, Hoboken, NJ, U.S.: John Wiley and Sons, 2013.
[50] Center for Chemical Process Safety (CCPS), Guidelines for Defining Process Safety
Competency Requirements, Hoboken, NJ, U.S.: John Wiley and Sons, 2015.
[55] Center for Chemical Process Safety (CCPS), Guidelines for the Management of
Change for Process Safety, Hoboken New Jersey, USA: John Wiley and Sons, 2008.
[57] K. Letrud, “A rebutal of NTL Institutes's Learning Pyramid,” Education, vol. 133, pp.
117-124, 2012.
[58] R. Flin, P. O'Connor and M. Crichton, Safety at the sharp end: A Guide to Non-
Technical Skills, Farnham: Ashgate, 2008.
[60] Health and Safety Executive (HSE), “RR446 - The development of a fatigue / risk index
for shiftworkers,” Health and Safety Executive (HSE), www.hse.gov.uk, 2006.
[61] International Association for Oil and Gas Producers (IOGP), “Report 626 – Managing
fatigue in the workplace,” International Association of Oil and Gas Producers,
www.iogp.com, 2019.
[62] Energy Institute, “Managing fatigue using a fatigue risk management plan (FRMP),”
Energy Institute, www.eneryinstitute.org, 2014.
[63] Canadian Centre for Occupational Health and Safety (CCOHS), “Fatigue,” Canadian
Centre for Occupational Health and Safety, https://www.ccohs.ca/, Undated.
[64] Energy Institute, “Human factors briefing note no. 23 – Workload and staffing levels,”
Energy Institute, www.energyinstitute.org, 2016.
[65] Energy Institute, “Locks removed on wrong blinds,” 21 February Undated. [Online].
Available: https://toolbox.energyinst.org.
[67] Health and Safety Executive (HSE), “The safe isolation of plant and equipment,” HSE
Books, https://www.hse.gov.uk, 2006.
[69] Energy Institute, “Confirming zero energy when pigging a pipeline,” 2 April 2019.
[Online]. Available: https://toolbox.energyinst.org.
[71] Energy Institute, “Draining pumps leads to product release,” 10 April 2109. [Online].
Available: https://toolbox.energyinst.org.
[72] J. R. Saward and N. A. Stanton, Individual Latent Error Detection (I-LED): Making
Systems Safer, Boca Raton: CRC Press, 2018.
368 Human Factors Handbook
[73] C. Wickens, Engineering Psychology and Human Performance, Columbus, OH, U.S.:
Charles.E.Merrill Publishing Co, 1984.
[75] J. Mitchell, “Lessons learnt from the introduction of human performance concepts
and tools on oil and gas platforms.,” in Hazards 27, Birmingham, 2017.
[76] M. Thomas, “Error management training: defining best practice. ATSB Aviation Safety
Research Grant Scheme Project 2004/0050 2007,” Autralian Transport Safety Board,
https://www.atsb.gov.au, 2004.
[77] M. Wright and S. Opiah, “Literature review: the relationship between psychological
safety, human performance and HSE performance,” Energy Institute, London, 2018.
[78] S. Dekker, Just culture: Balancing safety and accountability., London: Ashgate
Publishing Ltd.., 2012.
[79] T. R. Clark, The 4 stages of psychological safety: defining the path to inclusion and
innovation., Oakland: Berrett-Koehler Publishers, 2020.
[81] North Atlantic Treaty Organisation, “The NATO phonetic alphabet,” North Atlantic
Treaty Organisation, www.nato.int, Undated.
[82] U.K. Health and Safety Executive, “Managing shiftwork. Health and safety guidance,”
HSE Books, www.hse.gov.uk, 2006.
[83] U.S. Chemical Safety and Hazard Investigation Board (CSB), “Bayer CropScience
Pesticide Waste Tank Explosion,” U.S. Chemical Safety and Hazard Investigation
Board , Washington, DC., 2011.
[84] M. Endsley and D. Jones, “SA Demons: The Enemies of Situation Awareness,” in
Designing for Situation Awareness: An Approach to User-Centered Design, Boca Raton,
CRC Press, 2011, p. pp. 31–41..
[85] International Association for Oil and Gas Producers (IOGP), “Report 502 – Guidelines
for implementing Well Operations Crew Resource Management training,”
International Association for Oil and Gas Producers, www.iogp.org, 2014.
References 369
[86] M. Endsley and D. Garland, “Training for situational awareness in individuals and
teams,” in Situation awareness analysis and measurements, MAHWAH:NJ, LEA, 2000.
[87] U.K. Health and Safety Executive, “The explosion and fires at the Texaco Refinery,
Milford Haven, 24 July 1994: A report of the investigation by the Health and Safety
Executive into the explosion and fires on the Pembroke Cracking Company Plant at
the Texaco Refinery, Milford Haven,” HSE Books, www.hse.gov.uk, 1997.
[88] J. Orasanu, “Training for aviation decision making: The naturalistic decision making
perspective.,” in Proceedings of the Human Factors and Ergonomics Society Annual
Meeting, Los Angeles, 1995.
[90] U.S. Chemical Hazards and Safety Review Board, “Preliminary Animation of
Philadelphia Energy Solutions Refinery Fire and Explosions,” 16 October 2019.
[Online]. Available: www.csb.org.
[91] U.S. Chemical Hazard and Safety Investigation Board, “Fire and Explosions at
Philadelphia Energy Solutions Refinery Hydrofluoric Acid Alkylation Unit. Factual
update October 16, 2019,” U.S. Chemical Hazard and Safety Investigation Board,
www.csb.gov, 2019.
[92] International Association for Oil and Gas Producers (IOGP), “Report 501 – Crew
Resource Management for well operations teams,” International Association of Oil
and Gas Producers (IOGP), www.iogp.org, 2020.
[93] D. Cullen, “The Public Inquiry into the Piper Alpha Disaster: Volumes I and II.,” Her
Majesty's Stationery Office, London, 1990.
[94] M. McNeese, E. Salas and M. Endsley, “A model of inter- and intrateam situational
awareness: implications for design, training, and measurement,” in New Trends in
Cooperative Activities: Understanding System Dynamics in Complex Environments, Santa
Monica, Human Factors & Ergonomics Society, 2001, pp. 46-67.
[95] Center for Chemical Process Safety (CCPS), Contractor and Client Relations to Assure
Process Safety, Hoboken, N.J. U.S.: John Wiley and Sons, 1996.
[96] American Petroleum Institute, “Contractor Safety Management for Oil and Gas
Drilling and Production Operations,” American Petroleum Institute,
https://pslcolombia.com, 2005.
370 Human Factors Handbook
[97] International Association of Oil and Gas Producers (IOGP), “Report 423 – HSE
management guidelines for working together in a contract environment,”
International Association of Oil and Gas Producers, https://www.iogp.org, 2017.
[98] U.S. Chemical Safety and Hazard Investigation Board (CSB), “E. I. DuPont De
Nemours Co. Fatal Hotwork Explosion,” U.S. Chemical Safety and Hazard
Investigation Board, www.csb.gov, 2012.
[99] Energy Institute, “Managing major accident hazard risks (people, plant and
environment) during organisational change,” Energy Institute,
www.energyinstitute.org, 2020.
[100] Center for Chemical Process Safety (CCPS), “Introduction to Operational Readiness,”
Center for Chemical Process Safety, https://www.aiche.org, Undated.
[101] Center for Chemical Process Safety (CCPS), “Process Safety Leading Indicators
Industry Survey,” AIChE/CCPS, New York, NY USA, 2007.
[102] Center for Chemical Process Safety (CCPS), Guidelines for Process Safety Metrics,
Hoboken, NJ USA: John Wiley & Sons, 2009.
[103] U.K. Health and Safety Executive, “HSE Management Stanadards,” HSE Books,
https://www.hse.gov.uk, Undated.
[104] Center for Chemical Proces Safety (CCPS), Driving Continuous Process Safety
Improvement from Investigated Incidents, Hoboken, N.J., U.S.: John Wiley and Sons,
2021.
[105] Center for Chemical Process Safety (CCPS), Guidelines for Investigating Process
Safety Incidents, New York: Wiley Inter Science, 2003.
[107] British Petroleum (BP), “Deepwater Horizon - Accident Investigation Report,” British
Petroleum (BP), https://www.bp.com, 2010.
[110] “National Commission on the BP Deepwater Horizon Oil Spill and Offshore Drilling,
“Report to the President”,” 2011.
[111] M. Christou and M. Konstantinidou, “Safety of offshore oil and gas operations:
Lessons from past accident analysis.,” Publications Office of the European Union.,
Luxemburg, 2012.
[112] Offshore Engineer, “The Future of Offshore Energy and Technolog,” onedigital,
https://www.oedigital.com, Undated.
[113] E. Smith, R. Roels and S. King, “Guidance on learning from incidents, accidents and
events.,” in Proceedings of Hazards 25 Conference, www.icheme.org, 2015.
[115] D. R. Edwards, “Tripod Beta: Guidance on Using Tripod Beta in the Investigation and
Analysis of Incidents, Accidents and Business Losses.,” Energy Institute., London,
2017.
[116] United Kindgon Health and Safety Executive, “Human Failure Types,” HSE Books,
https://www.hse.gov.uk, Undated.
[120] Energy Institute: Hearts and Minds, “Making compliance easier (formerly Managing
rule breaking),” Energy Institute, https://heartsandminds.energyinst.org, Undated.
[123] International Association for Oil and Gas Producers (IOGP), “Introducing behavioural
markers of non-technical skills in oil and gas operations - Report 503,” International
Association for Oil and Gas Producers, https://www.iogp.org, 2018.
372 Human Factors Handbook
[124] C. C. R. Cooper and L. Eaken, Living with Stress, London: Penguin Books, 1988.
[125] Energy Institute, “HSE 184 Learning from incidents,” Energy Institute,
https://publishing.energyinst.org, 2015.
[127] BP, “Hot Work Safety Manual,” BP U.S. Pipelines and Logistics (USPL), BP,
www.bp.com, 2014.
[128] Energy Institute, “The Modern View of Incident Causation,” Energy Institute - Hearts
and Minds, https://toolbox.energyinst.org, 2020.
Human Factors Handbook For Process Plant Operations: Improving Process Safety and System
Performance CCPS.
© 2022 CCPS. Published 2022 The American Institute of Chemical Engineers.
• Unintentional errors:
The action was not intended, for example, pressing button A when the
intention was to press button B.
• Intentional errors:
“The operator intends to perform some act that is incorrect but believes it to
be correct or to represent a superior method of performance. In everyday
language, he has good intentions, but the effect on the system of his performance
may be undesirable.” (page 2-7)
A. D. Swain and H. E. Guttmann (1983, page 2-16 [118]) also used the following
categories of error:
Professor Reason authored in 1990 an accident causation model termed the “Swiss
cheese” model [19]. According to the model, hazards are prevented from causing
loss by a sequence of barriers, such as training, supervision and engineered
protection. Each barrier may have unintended weaknesses. These weaknesses are
represented as holes, such as with Swiss cheese.
374 Human Factors Handbook
• Active failures are the unsafe acts committed by people who are in
direct contact with the system. These include slips, lapses, or mistakes,
such as omitting an operational task or performing a task incorrectly.
• Latent failures are “resident pathogens” within a system caused by
decisions made by engineers, procedure authors, and management for
example. These can create “error provoking conditions” such as time
pressures and understaffing and poor procedures. They may lay
dormant for many years until a combination of events reveals them.
This model has been used to help understand accidents and the role that the
systems of management created the (hidden) conditions for human error. This
includes the notion that latent failures can cause multiple defenses to fail, and
thereby undermine “defense in depth” safety management systems. The model is
also used to prompt the identification and resolution of latent failures before they
contribute to an accident. The concept being that resolving one latent failure would
avoid many active failures.
A.2.1 “Violations”
Reason defined these as non-malevolent acts. The actions are intended but the
harmful consequences are unintended.
Appendix A - Human error concepts 375
A.2.2 Non-compliance
The Energy Institute’s ‘Hearts and Mind’ have issued extensive guidance on
“Making Compliance Easier” [120]. The guide provides an up to date view of ‘non-
compliance’ (p7), citing four forms of non-compliance. Their definitions are
reproduced in Table A-1. They note that reckless violations are considered to be
very rare. Their definitions focus on how the organization, the design of
procedures, team norms and knowledge of risks influence behavior.
a) Situational non-compliance
These happen when it is very difficult or impossible to get the job done by
following the procedures strictly. For example, there may not be enough people,
or the right equipment may not be available to follow the procedures as written.
b) Optimizing non-compliance
These happen when people think they can get the job done faster or more
conveniently by not following all the rules. There are two subtypes of optimizing
non-compliance:
Optimizing for organizational benefit: These happen when people take
shortcuts because they believe that it will help the organization achieve its goals,
e.g., achieve a performance target or meet a deadline. Non-compliance for
organizational benefits may show ways to improve productivity and safety if
brought out into the open, communicated, discussed and approved.
Optimizing for personal benefit: These happen when people take shortcuts
to reach a personal goal (e.g., leaving work on time, or meeting a target),
avoiding using complicated procedures, or because they have found a quicker,
easier or better way of doing the job.
c) Routine non-compliance
A non-compliance of any type can become routine.
These happen when people no longer appreciate the risk of the situation, or
when the rule no longer reflects reality, and not following the rule becomes the
accepted behavior. The rule may be seen as no longer relevant or important.
These non-compliances become routine, either by a whole group or just by one
individual. This indicates that there is an issue around a particular rule, or a
particular individual, or the effort required to follow the rule is perceived to be
greater than the benefits.
The Energy Institute’s ‘Hearts and Mind’ have issued extensive guidance on
“Making Compliance Easier” [120]. The guide states that:
Two figures from the guide are shown in Figure A-1 and Figure A-2.
Appendix A - Human error concepts 377
Human performance has a big role to play in incidents and accidents, and the
human performance of an organization arises from the interaction of people,
culture, equipment, work systems and processes.
The following principles of human performance embody the approach that
recognises the contribution of the system as well as the individual to errors,
mistakes and non-compliances in the organization (adapted from [106]).
70% due to
work
20%
systems
equipment
failure
80% human
error (mistakes
and non-
The modern view acknowledges that incidents are the result of complex
interactions in the system between people, plant and processes.
Therefore, improvements should focus on the system as a whole rather
than the individual, and on the need to reduce conditions and situations
that make mistakes and non-compliances more likely. Watch the video:
The Modern View of Incident Causation [128]
(Reproduced with permission from the Energy Institute: Figure 2 from [125])
Appendix A - Human error concepts 379
1. “People are fallible, and even the best people make mistakes.
Professor James Reason and Alan Hobbs, in their 2003 book “Managing
Maintenance Error, A Practical Guide” [17] offer 12 Principles of Error
Management. These twelve principles are as follows:
2. Errors are not intrinsically bad: Success and failure spring from
the same psychological roots. Without them we could neither learn
nor acquire the skills that are essential to safe and efficient work.
380 Human Factors Handbook
3. You cannot change the human condition, but you can change
the conditions in which humans work: Situations vary
enormously in their capacity for provoking unwanted actions.
Identifying these error traps and recognizing their characteristics are
essential preliminaries to effective error management.
5. People cannot easily avoid those actions they did not intend to
commit: Blaming people for their errors is emotionally satisfying
but remedially useless. We should not, however, confuse blame with
accountability. Everyone ought to be accountable for his or her
errors [and] acknowledge the errors and strive to be mindful to
avoid recurrence.
While these principles were cited in a book on maintenance error, they are
considered to be applicable to all error. Professor Reason stated that errors could
be anticipated, prepared for, and eliminated by the application of these principles.
The principles recognize that error will occur and needs to be guarded against
while also adopting a “systems” view that the organization should systematically
identify potential error and avoid the conditions that cause error.
Human Factors Handbook For Process Plant Operations: Improving Process Safety and System
Performance CCPS.
© 2022 CCPS. Published 2022 The American Institute of Chemical Engineers.
The 2005 Texas City refinery explosion, shown in Figure B-1, occurred during the
start-up of the isomerization (ISOM) unit, killing 15 people and injuring 180 [16].
The unit was being restarted after maintenance. The raffinate splitter tower
was overfilled. The raffinate flowed from the tower through a set of pressure safety
relief valves to a blowdown drum and stack, from which it was released into the
atmosphere and likely ignited by a nearby truck engine.
(from www.csb.gov)
• The ISOM raffinate section start-up began during the night shift and
stopped with the tower level control valve closed (this was unusual). The
operator did not use the start-up checklist and did not log his actions.
When the day shift started work, they had no record of the start-up.
• The bottom of the tower had been filled to 99%, which was not unusual
but was not consistent with the start-up procedures. Over time, this had
become an accepted deviation to the procedures. A high-level alarm set
at 72% activated and alarmed throughout the incident. A 78% high-high
level alarm did not activate.
• A poor shift handover meant the day shift was unaware that heat
exchangers, piping and other equipment had been filled in addition to
384 Human Factors Handbook
In 2008, a large explosion led to fatality of two workers at the Bayer Crop Science
plant in West Virginia, USA [26]. The fire burned for more than four hours. Two
contractors and six firefighters were treated for possible toxic exposure [83]. The
damaged plant is shown in Figure B-2.
The incident happened during the first methomyl restart after an extended
outage to install a new process control system and a stainless-steel pressure
vessel. The steps leading to this accident are outlined next:
not sampled, so the absence of the solvent was not identified. The state
of the vessel was also not discussed at shift handover.
• At 18:15, the outside operator started a recirculation pump as advised
by the board operator. At this time, the residue treater was not at its
optimal operating level, filled only to 30%. The temperature ranged from
140°F (60oC) to 149°F (65oC), which was below the critical operating
temperature of 275°F (135oC).
• It was recorded at 18:38 that the temperature began to steadily rise.
• At 22:21, a little under four hours later the vessel was filled to 51%
(normally filled to 50%). The temperature rose gradually from about
140oF (60oC) during this four-hour period (between 18:15 and 22:21),
well within the critical decomposition temperature of 275oF (135oC).
Thus, at this moment there was no indication of the filling failure.
• Recirculation then stopped due to an automated control system error.
• The temperature then rapidly rose to 286oF (141oC), within three
minutes to exceed the critical decomposition temperature of 275oF
(135oC).
• The board operator had observed that the residue treater pressure was
above the normal operating limits and still climbing but could not
understand why.
• At 22:25, this operator then heard the residue treater high-pressure
alarms sounding. The panel operator asked two outside operators to
check the vent system, as he suspected the vent line was blocked, while
he switched the residue treater recirculation system to full cooling.
• Due to the unusually high temperature, a ‘runaway thermal reaction’
occurred, i.e., the chemical reaction was accelerated by the high
temperature which in turn increased the temperature.
• Temperature continued to rise until it exceeded the safe operating limit
of 311oF (155oC).
• At 22:27, the gas produced by the thermal reaction overwhelmed the
emergency vent system. The vessel over pressured and ruptured. The
blast spilled approximately 2,200 gallons (8,300 liters) of flammable
solvent and toxic residues onto the road and into the unit, which
erupted into flames.
As shown in Figure B-3, the Esso Longford gas explosion in Australia was an
industrial accident with major consequences [20]. The incident resulted in two
fatalities and eight injuries. It also cut the gas supply for the state of Victoria for
two weeks.
A failure of a warm liquid (lean oil) system caused the temperatures of a heat
exchanger to drop and become intensely cold and therefore brittle. When
operators tried to reintroduce the warm oil, the brittle vessel fractured and
released large quantities of hydrocarbon vapor, which found an ignition source,
and exploded.
Some key events and failures are noted after Figure B-3.
• The lean oil flow in GP1 was stopped when pumps in GP1202 tripped
and were not restarted.
• GP905 experienced loss of lean oil flow. However, cold rich oil, and cold
condensate continued to flow through.
• GP1201 pump operations were disrupted for a few hours. Once it was
restarted there was some flow of warm oil into GP905 for a short period
of time. The flow of warm lean oil into a cold reboiler caused it to
become brittle and rupture. The reboiler temperature had fallen below
its minimum design metal temperature (MDMT). Metals below their
MDMT are susceptible to brittle fracture.
• The rupture of GP905 led to the release of a large volume of
hydrocarbon in the form of vapor, igniting, and consequently leading to
a series of explosions, and fire.
The impact of the incident was wide reaching, affecting operating personnel
and surrounding communities. The plant supplied heating gas to Melbourne and
other regions. The accident occurred during wintertime when the local
temperature was low and consumption of natural gas for heating was high.
Consequently, the resulting 20 days with gas shortage had a high impact on the
community.
Appendix B - Major accident case studies 389
On July 24th, 1994, a large explosion occurred at the plant of Texaco Refinery,
Milford Haven in Wales, which caused injury to 26 people [87]. The blast from the
explosion damaged properties in a 10 mile (16 kilometer) radius and was heard 40
miles (64 kilometer). The site suffered severe damage to the process plant, the
building, and storage tanks.
The event was preceded by a severe electrical storm that caused disturbance
to the plant, affecting the vacuum distillation, alkylation, and Butamer units, as well
as the fluidized catalytic cracking unit.
The explosion occurred some five hours later. The direct cause was a
combination of failures in management, equipment, and control systems during
the plant upset. These failures led to the release of approximately 22 tons (20
tonnes) of flammable hydrocarbon from the outlet pipe of the flare knockout
drum.
The released hydrocarbon formed a cloud of vapor and droplets that found a
source of ignition and consequently exploded.
Some key events and failures leading to the explosion are noted next:
A process diagrams illustrating the interaction of the key valves and vessels
that led to the explosion is shown in Figure B-5.
Figure B-5 Interaction of the key valves and vessels
The United States Chemical Safety Board reported that, on November 9th, 2010, an
explosion occurred at E.I. DuPont de Nemours and Co. Inc. (DuPont) Yerkes
chemical plant in Buffalo, New York [98]. This explosion occurred when a contract
welder and foreman were repairing the agitator support atop an atmospheric
storage tank containing highly flammable vinyl fluoride (a gas).
The plant had a Tedlar® process to convert vinyl fluoride into polyvinyl fluoride
(PVF), as shown in Figure B-6. The process includes the following stages:
• On October 21st, the process was shut down and all slurry was pumped
out of slurry tanks 2 and 1. The slurry tanks were locked out.
• On October 29th, a damaged agitator support on tank 1 was discovered
after the insulation was removed.
• On November 1st, repairs on tank 1 were delayed because materials
were unavailable. It was decided that slurry tank 1 repairs could be
completed after the process restarted on November 9th.
• On November 3rd, DuPont engineers discovered that the seal loop on
the flash tank overflow line had a split. They decided to return the tank
to service without repairing the split until the next outage. They
overlooked that the split provided a pathway for vinyl fluoride to enter
slurry tank 2.
• On November 6th, the Tedlar® process was restarted, with valves aligned
so that slurry went to slurry tank 3. Tank 2 had been returned to service.
• The equalizer line remained, connecting all three tanks.
• On November 7th, the reactor recycle compressor malfunctioned. The
unit was restarted without the compressor, doubling the vinyl fluoride
entering the flash tank.
• Later that day, a lock out card for tank 1 indicated that all valves to and
from the tank had been locked out, and the agitator was locked out. The
flash tank overflow line had no valves.
• Finally, on November 9th, A DuPont technician checked the atmosphere
around the tops of slurry tanks 1, 2, and 3. Continuous air monitoring
was arranged on tank 1.
• The atmosphere in tank 1 was not tested. This was not a requirement.
• Contractors completed a permit and started work on slurry tank 1, using
an arc welder.
The defective (split) seal loop had allowed vinyl fluoride and steam to flow from
the flash tank to tank 2, with some also entering tank 3. The vinyl fluoride and
steam flowed via the equalizer line from tank 2 into tank 1. Vinyl fluoride is heavier
than air and concentrated in the bottom of tank 1.
The top of tank 1 had an unsealed half-inch hole (for the agitator pipe). In
addition, the arc welder raised the temperature of the metal on top of the tank to
far above the ignition temperature of the vinyl fluoride. Either sparks entering the
tank or vinyl fluoride vapor contacting the hot metal, ignited the gas and caused
an explosion. A fire occurred in the tank and the overpressure ripped off the
majority of the top of the tank, causing the welder’s fatality. The foreman was
injured. The vinyl fluoride was consumed by the fire and the fire self-extinguished.
Appendix B - Major accident case studies 395
On April 20th, 2010, the Macondo well blew out [107]. Eleven fatalities resulted and
the Deepwater Horizon drilling rig sank and spilled an estimated four million
barrels of crude oil in the Gulf of Mexico. The spill disrupted the entire region’s
economy, and severely damaged fisheries and the eco-habitat.
Deepwater was operated by Transocean and had been under contract to BP.
A summary of events in the Deepwater Horizon Oil Spill (2010) are noted next:
The cement job was completed. This was followed by two pressure tests that
were also successfully completed. Later on, a decision was made not to run the
cement bond log.
levels could not be monitored. An unexpected loss of fluid was observed on the
riser pipe, which suggested leaks in the blowout preventer.
Negative pressure tests (where the crew reduced the fluid pressure to test for
leaks through the cement or well casing) showed unexpected results and raised
concerns over potential leaks.
A high-pressure pipe used to cut off the flow of oil fell to zero. Pressure in the
drill line increased to 1,400 pounds per square inch, indicating a buildup of natural
gas.
The internal blowout preventer and annular preventer opened. They started
pumping seawater down the drill pipe to displace mud and spacer from the riser.
The well pump was shut down for a sheen test, but the well continued to flow. The
drill pipe pressure increased. Abnormal pressure and more fluid returns were
observed.
Gas surged from the well and up the riser. Motor vessel Damon Bankston
reported drilling fluid spilling onto its deck. A warning was issued for Damon
Bankston to move 1,600 feet (500 meters) from the rig.
The rig then lost power. A few seconds later, the first explosion occurred. The
second explosion occurred 10 seconds later. A mayday call was made by
Deepwater Horizon, the emergency procedure was activated, and transfer
commenced of 115 personnel, including 17 injured, to Damon Bankston.
Performance/ Knowledge
HF Competency Level 1 - Operator Level 2 - Supervisor* Level 3 - Manager**
Criteria
Performance/ Knowledge
HF Competency Level 1 - Operator Level 2 - Supervisor* Level 3 - Manager**
Criteria
Performance/ Knowledge
HF Competency Level 1 - Operator Level 2 - Supervisor* Level 3 - Manager**
Criteria
Performance/ Knowledge
HF Competency Level 1 - Operator Level 2 - Supervisor* Level 3 - Manager**
Criteria
Procedures and job aids
Understand the concepts and Can recognize the need
various forms of procedures for application of various Able to lead on how to
and job aids procedures and job aids develop procedure and sign
Understands procedures and
Understands the importance Can recognize and select off/approve procedures
job aids
of procedures and job aids in the most suitable Able to appraise/critique
enhancing human procedure/job aid for a procedures and job aids
performance particular task
Performance/ Knowledge
HF Competency Level 1 - Operator Level 2 - Supervisor* Level 3 - Manager**
Criteria
Procedures and job aids
Performance/ Knowledge
HF Competency Level 1 - Operator Level 2 - Supervisor* Level 3 - Manager**
Criteria
Operational competency
Can determine
competency
requirements by
Supporting Is involved in the process of Able to review the
Understand the process of conducting task analysis,
operational determining competency for effectiveness of competency
determining competency perform learning needs
competency safety critical tasks process
analysis, and select
assessment learning
methods
Performance/ Knowledge
HF Competency Level 1 - Operator Level 2 - Supervisor* Level 3 - Manager**
Criteria
Operational competency
Task Support
Task planning and Can develop realistic task Can provide input and Can apply techniques of Can develop task planning
error management plans challenge to task plans “grounded” planning methods
Understands and can
Can help identify potential apply techniques to
Distractions and Can minimize distractions and Can review effectiveness and
distractions and minimize task
interruptions interruptions develop tactics
interruptions distractions and
interruptions
Can recognize which control
Can review effectiveness and
Can select, develop and apply of work processes apply and Can implement control
Control of work develop control of work
suitable control of work offer input to their of work processes
processes
development
Table C-1 continued
Performance/ Knowledge
HF Competency Level 1 - Operator Level 2 - Supervisor* Level 3 - Manager**
Criteria
Task Support
Can determine and
Understands the imperative Can review effectiveness and
Isolation and Can achieve high reliability apply suitable isolation
to apply isolation and develop isolation and
interlocks high Isolation and interlocks and interlock
interlock controls interlock requirements
requirements
Can identify and support
Can review effectiveness and
Communication Can reliably operate Can apply formal application of
develop communication
protocols communication protocols communication protocols communication
protocols
protocols
Can review effectiveness and
Can identify and support
Can apply shift handover develop shift handover
Shift handover Successful shift handover application of shift
procedures procedures
handover procedures
Can recognize how
fatigue contributes to Is able to assess the impact
Understands the concept of impaired performance of fatigue on performance
Can describe factors
fatigue management and its Can determine staffing and safety
contributing to fatigue
relation to performance needs for tasks Can design shift rotating
Can detect signs of schedules and staffing levels
Fatigue fatigue
Management Can identify factors (e.g.,
Is able to detect signs of excessive workload,
Can identify signs of fatigue Is able to assess level of
fatigue in self and others and understaffing)
and its impact on fatigue risk and identify
initiate course of action to contributing to fatigue
performance mitigating strategies
mitigate fatigue and suggests mitigating
strategies
Table C-1 continued
Performance/ Knowledge
HF Competency Level 1 - Operator Level 2 - Supervisor* Level 3 - Manager**
Criteria
Task Support
Can review task plans
Is able to lead discussion on
and evaluate task
Understands concepts and Can describe error traps and error management,
operations
importance of error the components of error including effective error
Can recognize
management management prevention and
occurrence of errors in
management
Error Management self and others
Is able to ensure that
Can recognize error traps
Can identify potential employees understand the
Is able to contribute and envisage potential
error traps and change principles of error
effectively to error consequence. Can identify
tasks and conditions to management and apply
management someone at risk of making a
minimize risk of error them in practice
mistake
Can support the
Can coach people in how development of new ways of
Can recover from error and Can recover from error and
Error recovery to recover from error recovering from error and
be resilient be resilient
and be resilient being resilient
Performance/ Knowledge
HF Competency Level 1 - Operator Level 2 - Supervisor* Level 3 - Manager**
Criteria
Task Support
Non-technical skills
Performance/ Knowledge
HF Competency Level 1 - Operator Level 2 - Supervisor* Level 3 - Manager**
Criteria
Non-technical skills
Understands conditions Can identify: a) factors Can recognize factors Is able to conduct an
contributing to loss of contributing to impaired that lead to assessment of situation
situation awareness and signs situation awareness; and b) impaired/loss of awareness and categorize
of impaired situation signs of impaired situation situation awareness in situation awareness as:
awareness awareness self and others maintained; impaired; or lost
Situation Awareness
(continued)
Performance/ Knowledge
HF Competency Level 1 - Operator Level 2 - Supervisor* Level 3 - Manager**
Criteria
Non-technical skills
Performance/ Knowledge
HF Competency Level 1 - Operator Level 2 - Supervisor* Level 3 - Manager**
Criteria
Non-technical skills
Performance/ Knowledge
HF Competency Level 1 - Operator Level 2 - Supervisor* Level 3 - Manager**
Criteria
Non-technical skills
Communication
Performance/ Knowledge
HF Competency Level 1 - Operator Level 2 - Supervisor* Level 3 - Manager**
Criteria
Non-technical skills
Performance/ Knowledge
HF Competency Level 1 - Operator Level 2 - Supervisor* Level 3 - Manager**
Criteria
Non-technical skills
Can demonstrate
effective leadership skills
(e.g., centralizing
Is able to assess
Understands the concept and Can identify characteristics of communication,
Leadership importance of leadership in effective leadership in coordinating tasks, effectiveness of leadership
emergency situations emergency situations managing teams skills in abnormal situations
understanding of the
situation etc.) in
emergency situations
Managing contractors
Performance/ Knowledge
HF Competency Level 1 - Operator Level 2 - Supervisor* Level 3 - Manager**
Criteria
Managing change
Performance/ Knowledge
HF Competency Level 1 - Operator Level 2 - Supervisor* Level 3 - Manager**
Criteria
Understand the importance of Can describe the importance Can recognize the need Is able to lead discussion on
lessons learning and sharing of lessons learning and for lessons to be learnt lessons learnt
in error prevention sharing across business units and shared across wider Is able to assess
Can contribute to lessons business effectiveness of lessons
Lessons Learning being captured and shared Can contribute to sharing strategies
discussions involving Is able to assess the
lessons learning and effectiveness of applied
their application in lessons across the whole
practice business
* / ** The definition of supervisors and manager may differ across organizations and countries.
Human Factors Handbook For Process Plant Operations: Improving Process Safety and System
Performance CCPS.
© 2022 CCPS. Published 2022 The American Institute of Chemical Engineers.
Medium
Safety criticality Monitor process indicators & keep within safe operating
(High, Medium, Low) limits.
Share information about equipment faults or failures.
Complexity of task
Low
(High, Medium, Low)
Freq. of task (High,
Medium
Medium, Low)
Time required to
complete task (Long, Medium
Medium, Short)
Knowledge
Function of equipment including safety devices.
Interpret signs, signals, & symbols.
Competency - Knowledge of safe operating limits, & process hazards.
Knowledge, Skills and Procedural
Attitudes required Establish procedures.
Skills
Alertness, communication, task planning, interpreting
info, basic reading & writing.
416 Human Factors Handbook
Knowledge
Function of equipment including safety devices.
Interpret signs, signals, & symbols.
Competency - Knowledge of safe operating limits, & process hazards.
Knowledge, Skills and Procedural
Attitudes required Establish procedures.
Skills
Alertness, communication, task planning, interpreting
info, basic reading & writing.
Skill; Rule; and
Mostly skill-based.
Knowledge-based
Some procedure-based.
activities
Training content
Memory-based Resource & application.
versus resource & Memory-based (partial).
application
Dependency N/A
Appendix D - Competency performance standards 417
Safety criticality
Medium
(High, Medium, Low)
Complexity of task
Medium
(High, Medium, Low)
Freq. of task (High,
High
Medium, Low)
Time required to
complete task (Long, Short
Medium, Short)
Knowledge
Function of equipment including safety devices.
Interpret signs, signals, & symbols.
Knowledge of safe operating limits, & process
Competency
hazards.
Knowledge, Skills and
Procedural
Attitudes required
Establish procedures.
Skills:
Communication, task planning, interpreting info,
decision-making, basic reading & writing.
Skill; Rule; and
Knowledge-based Procedure/ rule-based task.
activities
Training content
Memory-based Resource & application.
versus resource & Memory-based.
application
Dependency N/A
418 Human Factors Handbook
Safety
criticality (High, High
Medium, Low)
Complexity of
task (High, High
Medium, Low)
Freq. of task
(High, Medium, High
Low)
Time required
to complete
Medium
task (Long,
Medium, Short)
Knowledge
Function of equipment including safety devices.
Interpret signs, signals, & symbols.
Competency Knowledge of safe operating limits & process hazards.
Knowledge, Skills Procedural
and Attitudes Establish procedures.
required Skills
Problem solving, communication, task planning,
interpreting info, leadership, teamwork, decision-making,
advanced reading & writing.
Appendix D - Competency performance standards 419
Training
content
Memory-based.
Memory-based
Resource & application.
versus resource
& application
Dependency N/A
Human Factors Handbook For Process Plant Operations: Improving Process Safety and System
Performance CCPS.
© 2022 CCPS. Published 2022 The American Institute of Chemical Engineers.
Types of human
Task category Task types performance (skill, Learning method Form of learning
rule, knowledge)
• On-the-job training
Skill-based task,
requiring visual • Mentoring
Process Skill-based Practice-based
or auditory • Trial and practice
recognition skills • “Show me”
Types of human
Task category Task types performance (skill, Learning method Form of learning
rule, knowledge)
Information-based • Pre-course material, & classroom
Knowledge-based
Knowledge- learning
based • Formal qualifications
operational • On-the-job training, mentoring,
Procedural
planning, shadowing
Rule-based Demonstration-based
development of
• “Talk through” or “walk-through”
procedures
procedures
Types of human
Task category Task types performance (skill, Learning method Form of learning
rule, knowledge)
Procedural, Demonstration-based • On-the-job training
operational, or • Instructor led “walk-through”
Procedural Skill-based
maintenance Practice-based procedures
task
• On-the-job training
Skill-based task,
requiring • Mentoring/coaching
Psycho-motor Skill-based Practice-based
psychomotor • Trial & practice
skills • “Show me”
Table E-1 continued
Types of human
Task category Task types performance (skill, Learning method Form of learning
rule, knowledge)
Types of human
Task category Task types performance (skill, Learning method Form of learning
rule, knowledge)
Maintaining Information-based • Verbal instructions, on-the-job
Knowledge-based
Cognitive & awareness of training
Process process, and Practice-based • Shadowing
Skill-based
system state
Situation Awareness: Developing and maintaining a dynamic awareness of the situation and of the risks present during an
operation. This is based on gathering information from multiple sources from the task environment, understanding what the
information means, and using it to think ahead about what may happen next.
Table F-1 Situation awareness – behavioral markers for oil and gas industry
Elements Examples of behaviors reflecting good practice Examples of behaviors reflecting poor practice
Regularly checks key sources of information including Does not go to the effort to locate or confirm important
alarms and other prompts. information that is not readily available.
Makes use of all available information sources – Does not initiate prompt intervention at the activation
instruments, colleagues, and others – to check the of the alarm. Is unable to interpret signals or other
Actively seeks status of the operation, or to check assumptions about parameters to prevent a problem.
relevant the state of the world or the operation.
information Shows concern and takes action if important
information is not available when it is needed.
Asks for regular updates from colleagues who may have
relevant information.
Proactively addresses missing relevant information.
Table F-1 continued
Elements Examples of behaviors reflecting good practice Examples of behaviors reflecting poor practice
Challenges key assumptions that could impact safety, Does not evaluate the reliability of information that has
and regularly checks to confirm they are still the potential to create an unsafe condition.
reasonable. Makes statements, asks questions, or makes
Works to Challenges assessment of risk and the state of the suggestions that may indicate:
understand world.
information Shows unease or concern and checks if data or • Lack of awareness of what is happening.
information is not consistent with expectations. • That they have not understood the
Prioritizes actions, taking into account critical signals, significance of information.
avoiding “information flooding”. • That they have ignored the views of others.
Table F-1 continued
Elements Examples of behaviors reflecting good practice Examples of behaviors reflecting poor practice
Elements Examples of behaviors reflecting good practice Examples of behaviors reflecting poor practice
Maintains Steps back and checks that the situation or conditions Gives too much weight to expectations based on
awareness have not changed significantly with time. experience, rather than current information/opinion.
and respect If something unexpected happens, steps back and Shows a willingness to disbelieve data or information
of risk reassesses planned activity in consultation with peers. that conflicts with what is expected.
Checks the work environment to ensure it is as it should Shows a willingness to quickly accept data or
be before beginning a critical activity. information that backs up preconceptions.
Plans to make allowance for interruptions or Does not consider potential problems. Gives no insight
unexpected events. into expectations or actions if the situation changes.
Identifies and proposes alternative options if events do Acts in ways that knowingly goes beyond their
not go as planned. competence or experience.
Table F-1 continued
Elements Examples of behaviors reflecting good practice Examples of behaviors reflecting poor practice
• Have a different understanding of the current Reacts negatively or not at all, if other team members
state of the operation than they do. say or do things that suggest they have a different
Recognizes • Are not aware of the state of critical assessment of the situation, equipment, or risks.
mismatches equipment. Does not make others aware of:
between own • Have a different assessment of the key risks.
• Difficulties until after things have gone wrong.
SA and that • Have a different assessment of team goals and
held by • Information or events that are unexpected.
priorities than their own.
others Assumes, without checking, that others who need to
Interrelates different types and sources of information.
know are aware of the same risks as they are that
Willing to be challenged on their mental model and
could affect the safety of an operation.
change.
Does not hesitate to challenge other’s mental model
respectfully.
Adapted from [123]
Human Factors Handbook For Process Plant Operations: Improving Process Safety and System
Performance CCPS.
© 2022 CCPS. Published 2022 The American Institute of Chemical Engineers.
Staffing arrangements
Staffing arrangements
Creation of new
communication and
contact interfaces. Formalize communications.
Loss of in-house Provide team bonding
expertise. exercises.
Reliance on continuity of Determine the minimum
service. level of in-house expertise
Outsourcing.
Contractor does not to be retained.
adopt company safety Offer a cultural induction.
values. Provide contractor training
Contractor lacks and certification.
competence in site
management
procedures.
Human Factors Handbook For Process Plant Operations: Improving Process Safety and System
Performance CCPS.
© 2022 CCPS. Published 2022 The American Institute of Chemical Engineers.
Index
Shift handover, 124, 196, 202, 238, Teamwork, 32, 285, 411
241, 242, 243, 244, 383, 386, 404, Trust, 7, 191, 193, 224, 225, 275,
433 284, 346, 359
Shift Handover, 56, 63, 129 Usability assessments, 103
Shift system, 164, 337, 434 User acceptance testing, 104
Staffing level, 70, 174, 185, 432, 433, User-centered design, 103
434 Vigilance, xxiv, 36
Stress, 38, 218, 253, 286, 288, 327, Walk-through, 49, 55, 73, 74, 130,
328, 411 146, 421
Supervision, 13, 46, 79, 123, 134, Work instruction, 65
145, 311, 346, 433, 435 Workload, xxiii, 12, 35, 53, 158, 159,
Tabular Task Analysis, 131 174, 176, 177, 196, 219, 242, 253,
Task analysis, 53, 69, 128, 130, 136 256, 259, 295, 309, 404, 432, 434,
Task verification, 77, 202, 203, 213, 435
217, 230, 231