text
stringlengths 0
5.91k
|
|---|
Appendix A. Inherently Safer Technology (IST)
Checklist
A.1 IST Checklist Procedure
The following procedure can be used to answer each Inherently Safer
Technology (IST) question listed in A.2 Checklist Questions:
1.Determine the applicability of th e item to the process being
reviewed. Indicate “Y” or “N” in the appropriate column to
indicate applicability.
2.If an item is applicable to th e process under review, determine
whether there are any IST opportunities or applications available
that could potentially reduce the risk (consequences and/or
likelihood) of an accidental or intentional release. If an IST
opportunity is identified, it should be described in the
Opportunities/Applications column . If no new IST opportunities
are identified, this is noted in the Opportunities/Applications
column with supporting info rmation, including general
references to existing safeguards and IST already implemented,
where applicable.
3.If an IST opportunity is identifi ed, a screening evaluation of the
alternative(s) should be perfor med by the review team and the
results summarized in the Feasibility column. If the IST
alternative is determined not to b e f e a s i b l e ( d u e t o c o s t ,
technology limitations, security , operability, safety, or other
factors), this decision is noted as such along with supporting
information, including general refe rences to existing safeguards
and IST already implemented, where applicable. The current
status of the opportunity (e.g ., “to be evaluated further,”
“evaluation in-progress,” “implementation in-progress,” etc.) is
documented in the “Current Status” column.
4.If an IST opportunity is d eemed potentially feasible, a
recommendation for further eval uation or implementation is
442(VJEFMJOFTGPS*OIFSFOUMZ4BGFS$IFNJDBM1SPDFTTFT"-JGF$ZDMF"QQSPBDI
#Z$$14
¥5IF"NFSJDBO*OTUJUVUFPG$IFNJDBM&OHJOFFST |
significantly reducing the inventory in the transfer pipe. This alternative
reduced the distance to the ERPG-3 concentration of 20 ppm chlorine to
about 0.5 miles (0.8 km) for the re lease scenario of rupture of the
chlorine transfer line.
For this installation, alternative wa ter disinfecting systems were also
investigated. It was found to be fe asible to use sodium hypochlorite
treatment, and this essentially elimin ated the hazard of a chlorine vapor
cloud entirely.
15.3.4 Reduction of Chlorine Transfer Line Size A manufacturing process included a chlorination process using liquid
chlorine. The original facility used a 2-inch (5.1 cm) transfer line from the
chlorine storage facility to the manuf acturing building. A hazards review
questioned the line size, and it was de termined that it could be reduced
to 1 inch (2.5 cm) without impactin g the manufacturing process. Table
15.2 shows the impact of this reduct ion on the hazard zone resulting
from the potential failure of the transfer pipe. Several typical weather conditions are considered. For purpos es of this example, the hazard
zone was defined as the distance to the ERPG-3 concentration for
chlorine, 20 ppm. For all weather condit ions, the distance to the ERPG-3
concentration was reduced by a factor of two to three.
15.3.5 Substitution of Aqueous Ammonia For Anhydrous Ammonia Dilution of a hazardous material can be an important strategy for
improving the inherent safety of a chem ical process or storage facility. It
can reduce the storage or vapor pre ssure of a hazardous material and
reduce the atmospheric concentratio n of hazardous vapor from a spill.
Approximately 500,000 pounds (227,000 kg) of anhydrous ammonia
were stored in a large pressurized storage tank. The tank was rated for
75 psig (90 psia, 6.2 atm. absolute ) working pressure, and it had a
pressure relief valve set for 72 psig (87 psia, 6.0 atm. absolute). This value
compares to the vapor pressure of ammonia of about 93 psig (108 psia,
6.4 atm. absolute) at 60 °F (16 °C). The ammonia was kept under
refrigeration to maintain the stor age tank pressure below the tank
pressure rating and relief valve set po int. Occasionally, the refrigeration
system would fail, and the storage tank pressure would slowly increase 400 |
143
Explosions Vapor clouds
Confined deflagrations
Detonations
Exothermic runaway reactions
Physical overpressure of pressure vessels
Brittle fracture
Polymerizations
Decompositions
Boiling liquid, expanding vapor explosions (BLEVE)
Undesired reactions catalyzed by materials of
construction or by ancillary materials, such as pipe
dope and lubricants
Toxicity-related hazards Toxic to humans (acute) causing reversible or irreversible injury or fatality
Environmentally toxic to plant, animal or fish life
(large scale events)
Important flammability characteristics are the lower and upper
flammability limits, the flash point, the minimum ignition energy, the
minimum oxygen concentration, an d the autoignition temperature.
Values of these properties are availa ble in many different publications
and sources, including:
Safety Data Sheets (SDS). The SDS is a part of the Globally Harmonized System (GHS) of Classification and Labelling of Chemicals managed by the United Nations (Ref 8.76 GHS) that
was set up to replace the assortment of hazardous material |
PEOPLE MANAGEMENT ASPECTS OF PROCESS SAFETY MANAGEMENT 433
Lessons
Process Safety Culture. The characteristics of a good process safety culture include
maintaining a sense of vulnerability and establis hing a learning/questioning environment. The
Baker Panel report written after the BP Texas City Refinery explosion focused on process safety
culture. (Baker 2007) BP was in the process of implementing the recommendations of the
Baker Panel in 2010. In 2008 BP overhauled its management system and developed a new
system called the Operating Management Syst em Framework (OMS), and by 2009, OMS was
about 80% implemented. BP intended to have OMS applicable to drilling rigs. However, BPs
requirements were just being rolled out wh en the Macondo Well was drilled and were not
applied to the Macondo well.
The confirmation bias, which prevented the crew from recognizing the failure of negative
pressure test as valid, is another symptom of a lack of a learning/questioning environment and
a lack of a sense of vulnerability.
Further illustrating this point was the BOEMRE report statement that “in the weeks leading
up to the blowout on April 20, the BP Macondo team made a series of operational decisions
that reduced costs and increased risk” and that the investigation team “found no evidence that
the cost-cutting and time-saving decisions we re subjected to the various formal risk
assessment processes that BP had in place”.
Compliance with Standards. Neither BP nor Transocean implemented their own process
safety management policies. Bo th had MOC guidelines that were not followed during the
abandonment procedure.
The CSB noted Transocean’s “minimal guidance and unclear expectations of the risk
management tools its personne l should use”. The crew at Macondo well did not apply the
techniques identified as Transocean’s risk management tools; HAZID/HAZOP, Major Hazard
Risk Assessment, Safety Case, and Operation Integrity Case. These tools were supposed to
demonstrate the risk was As Low As Reasonably Practicable (ALARP), but Transocean did not
provide guidance on what tools to use.
Hazardous processes should be designed with multiple safeguards. BOPs are designed
with multiple rams that close in various ways and are intended to shut off the flow from the
well either through the annular space or the central drill pipe. At the time of the incident,
Transocean had BOPs with two BSRs on 11 out of 14 of its rigs, and BP had two BSRs on all the
other rigs it was leasing. The Deepwater Horizon rig only had one BSR. Although it was normal
industry operating practice at the time, relying on such a vulnerable layer of protection as the
final layer is an example of poor COO. One could argue that reliance on the BOP may have
reduced the crew’s “sense of vulnerability” as they believed it was the ultimate layer of
protection, when, in fact, it was a flawed safeguard.
Operating Procedures. The Deepwater Horizon crew were not supplied with a procedure
for testing the cement barrier. The crew did not, therefore, have a criterion for deciding if the
test was positive or negative, or actions to take following a negative test. (CSB 2014 a) The
abandonment procedure was written 24 hours in advance. This was partly due to the nature
of the strata at the bottom of the well, which could not be known until the well was drilled. No |
CON TIN UOUS IM PROVEM EN T 327
revalidated (5-year interval). This is an important means to institutionalize
the lessons learned from the inciden t, as discussed in Chapter 16.
15.1 REGULATORY COM P LIANCE REVIEW
Table 15.1 Requirement Complia nce Checklist (USA OSHA/ EPA)
Requirement Statement Com pliance?
The Investigation Itself: Yes No
1. An investigation must be performed for each incident in a covered process
that did or could reasonably have resulted in a catastrophic release of:
– a highly hazardous chemical per US OSHA PSM or,
– a regulated substance per US EPA RMP.
2. The investigation should start as soon as is reasonably possible, but must
start within 48 hours following the incident. (This requires documentation of
date and time at which the investigation began.)
3. The investigation team is to be composed of:
– at least one person knowledgeable in the process involved,
– a contract employee if the incident involved work of the
contractor,
– any other person with appropriate knowledge and experience that is
required to thoroughly in vestigate and analyze the incident.
The Report and Findings: Yes No
1. A report is required at the conclusion of the investigation and the report
must include:
– date of the incident
– date the investigation began
– a description of the incident
– the factors that contributed to the incident
– recommendations resultin g from the investigation
2. The report must be reviewed wi th all affected personnel whose jobs are
relevant to the investigation findings, including contract employees where
applicable.
3. A system must be in place and utilized to promptly address the incident
report findings and recommendations.
4. The investigation report must be retained for five years.
|
CONTINUOUS IMPROVEMENT 159
and Guide for Fire and Explosion Investigations (NFPA 2021). The metrics
guidance on classifying incidents can also be used to determine severity
(CCPS 2018d, API 2016). Whichever approach is taken, the goal should
be to identify the system weakne sses and conditions causing the
abnormal situation, and then im prove the system to prevent the
conditions that could otherwise lead to future abnormal situations.
6.4 AUDITING
Auditing process safety is anothe r way of identifying and correcting
management system issues and weaknesses before they result in
abnormal situations and, potent ially, serious incidents. Both
internal/self-audits and external audits should be conducted. Most, if not
all, of the elements of RBPS (CCPS 2007a) should be audited periodically,
and actions taken to improve the effectiveness of the management
system, with a particular emphasis on abnormal situations.
For example, an audit may find th at hazard identification and risk
analysis (HIRA) studies did not ad dress the circumstances of past
occurrences of abnormal situations. Such a key finding should spur
leadership to look closely at how HI RA studies are conducted, in terms
of scope, methodology, and team co mpetency / experience. Future HIRA
studies, especially Hazard and Op erability (HAZOP) Studies, should
specifically address the manage ment of abnormal situations.
An audit might identify that the safety systems in a plant are not
capable of performing adequately (if at all) during abnormal situations.
Field inspections of equipment may id entify several possible issues such
as isolation valves incorrectly closed or valves that are open but should
be closed, missing valve manual actuator s, leaking air lines to solenoids,
or open-ended lines. Auditors can review the alarm summary with
control room operators. Any standing /stale alarms can be discussed and
measured, providing an opportunity to review the operator training and
their level of understanding of the control/alarm systems and required
responses. These kinds of issues ca n then be addressed, and the data
used as a leading indicator in the metrics. Example Incident 6.1 and
Example Incident 6.2 are good exampl es of incidents in which audits
have identified deficienci es in safety systems. |
106 | 3 Leadership for Process Safety Culture Within the Organizational Structure
3.25 CCPS, Guidelines for Managing Process Safety Risks During
Organizational Change, American Institute of Chemical Engineers, New
York, 2013.
3.26 Schein, E.H., Organizational Culture and Leadership , 3rd Ed., San
Francisco: Jossey-Bass, 2004. |
472 PROCESS SAFETY FOR ENGINEERS: AN INTRODUCTION
14. Which statement best describes the reactivity of water with ethylene? Compatible, caution,
chemical reactive, or incompatible?
15. Which statement best describes the reactivity of ethylene with itself? Selective reduction,
selective reactivity, saturated radicals, or self-reactive?
16. During the release, your boss is concerned wi th potential ignition sources. Considering the
MIE, what do you tell him about the likelihood of ignition?
17. You look up the maximum intended inventory of the process for ethylene and see that it
is 20,000 lbm. If the process is in the Un ited States, would it be covered under OSHA
1910.119, the “PSM” or Process Safety Management regulation? There is no OSHA PSM TQ
for ethylene.
18. What is the damage estimate for common struct ures if the side on pressure is 2.0 psig?
19. If the wind speed during the release is 2 m/s on a still night, what is the atmospheric
stability class?
20. The ethylene release scenario is analyzed. It is determined that the consequence could be
an explosion causing $5 million in damages. Wh at severity category is this using the risk
matrix provided in section 14.5?
21. This ethylene release scenario involves an equipment failure that is estimated to occur
sometime in the life of the piece of equipment. What probability category is this using the
risk matrix provided in section 14.5?
22. Considering the severity and probability categori es from the previous two questions, what
is the risk using the risk matrix provided in section 14.5?
23. It is decided that this ethylene release risk should be reduced. What strategy should be
considered first to reduce the risk?
24. It is decided to install an independent protecti on layer (IPL) as a means to reduce the risk.
What functional criteria must this protection layer have to be considered an IPL?
Exercise 4: Wastewater Equalization Tank
You are assigned to conduct a HAZOP of a proc ess. Your colleagues help you to break the
process into nodes for the study. Node 1 is shown in Figure A.4 and the Node 1 intention and
parameters are described in Table A.3.
Table A.4 is the HAZOP Worksheet for Node 1. The “high level” deviation has been
completed. Complete the other deviations listed on the worksheet. Include the risk ranking for
both the unmitigated and mitigated scenario. Use the risk matrix provided in Figure A.5 and
Table A.5 and Table A.6. |
3
Guidelines for Hazard Evaluation Procedures, Third Edition, with
Worked Examples , 2007.
Guidelines for Engineering Design for Process Safety, Second Edition ,
2012, includes a section (Chapter 5.2) on inherently safer design.
Guidelines for Implementing Proc ess Safety Management Systems,
Second Edition , 2016.
Guidelines for Siting and Layout of Facilities, Second Edition , 2018.
Guidelines for Integrating Process Safety into Engineering Projects ,
2019.
1.3 ORGANIZATION OF THE BOOK
The book is written with the key princi ples for inherent safety in the body
of the book. Tools for implementing the approach, as well as indicative
examples and checklists, are included in the appendices. The guidelines
begin with an explanation of the concept of Inherent Safety (IS) and then
explain its role in Process Risk Management.
Chapter 2 introduces the topic of inherent safety. The key terms and
the philosophy behind inherent safety are described. The different ways
in which inherent safety is applied can be categorized into “strategies.”
These strategies— minimize, substitute, moderate, and simplify—are
discussed in detail in Chapters 3 through 6. Human factors are an extremely important subset of the simplify strategy, and Chapter 6.11
of this book presents a detailed disc ussion of human factors as related
to inherently safer design for the human-machine interface. Chapter 7
outlines how each of the four IS stra tegies can be applied to traditional
protection layers.
“Inherently safer” is a way of thin king and, to successfully implement
it, this thinking must be well-und erstood and continually employed
wherever possible. Improved underst anding of the pr ocess may result
in a better, more reliable, and even more productive and profitable
process that produces higher qua lity products. Processes should be
reviewed for hazards and risks periodic ally throughout their life cycle.
Chapters 8 discusses review methods to do this. Chapter 9 discusses the
role of inherent safety in chemical process security. Chapter 10 presents
available methods for implementing inherently safer strategies. These can either be independent, special stud ies, done periodically or before a
major project or change is undertaken, or opportunistically applied into |
338 PROCESS SAFETY FOR ENGINEERS: AN INTRODUCTION
Figure 15.1. Oxidation reactor after explosion
(Celanese)
Figure 15.2. One of several units impacted by explosion
(Celanese)
|
TOXIC HAZARDS 111
In the US, the Emergency Response Planning Guideline (ERPG) concentration values have
become broadly accepted within industry and government. ERPG values have been developed
for approximately 150 chemicals. The thre e ERPG levels are defined as follows.
ERPG-1: the maximum airborne concentrat ion below which nearly all individuals
could be exposed for up to one hour with out experiencing effects other than mild
transient adverse health effects or perceiving an objectionable odor.
ERPG-2: the maximum airborne concentrat ion below which nearly all individuals
could be exposed for up to one hour without developing serious health effects that
could impair their ability to take protective action.
ERPG-3: the maximum airborne concentrat ion below which nearly all individuals
could be exposed for up to one hour without developing serious health effects that
could impair their ability to take protective action.
Figure 6.4 Emergency Response Planning Guideline (ERPG) concentration
(NOAA)
Where ERPG criteria are not available, other criteria may be used. These may include the
following: The ERPG, AEGL, and TEEL are all listed in the Preventive Action Criteria database
listed in Section 6.7.
Acute Exposure Guideline Levels (AEGL) va lues published by the U.S. Environmental
Protection Agency (EPA). AEGLs are applic able to exposures ranging from 10 minutes
to 8 hours.
Temporary Emergency Exposure Limit (TEEL) values developed by U.S. Department
of Energy.
ERPG and AEGL values focus on short term exposures relevant to process safety and
emergency response. Other more appropriate pa rameters are used for longer term worker
exposures, as might occur during spill cleanup op erations including threshold limit value (TLV).
These are set by the American Conference on Go vernmental Industrial Hygienists. Permissible
|
326 | Appendix E Process Safety Culture Case Histories
term inal operated by another com pany. During the process, he
began to change the position of a figure-eight type line-blind
valve. Unfortunately, block valves upstream of the blind had been
opened out of sequence. As he swung the blind, a jet of gasoline
sprayed out at high volume. The worker was unable to stop the
release, and was soon overcom e by fumes.
A supervisor attempted to rescue the worker, but he too was
overcom e with fum es and barely escaped. A second worker also
attempted rescue and was also overcome. A third worker
norm ally on site was offsite for personal reasons, leaving no one
at the site to initiate further control actions. B y the time response
personnel arrived, vapors from the leak had engulfed the entire
site. Recognizing the danger of explosion, they retreated.
One hour and fifteen m inutes later, the vapor cloud exploded,
creating a fireball that engulfed entire site and broke windows in
the surrounding com munity up to 2 km away. The ensuing fire
soon spread to all the other tanks on the site and continued to
rage for eleven days. Due to the scale of the fire, responders
decided to allow the fire to burn itself out rather than try to control
it. Ultim ately six workers and five in the comm unity lost their lives.
The investigating comm ission (Ref E.6) determ ined that the
accident was caused by valves being operated out of sequence,
and was exacerbated by the absence of a remote isolation valve
and/or remotely operated shut-off. The comm ission noted that
there were no operating instructions for m aking the transfer,
leaving the procedure up to the operators who were not well
trained in this procedure.
While the commission did not comm ent specifically on
com pany process safety culture, several recom mendations show
they were clearly thinking about it. Among the recommendations,
the com mission recomm ended creating an independent process
safety function reporting to the CEO and that line managem ent
practice conduct of operations to ensure that all process safety |
Overview of the PHA Revalidation Process 5
recommendations to reduce the likelihood of the deviation or the severity of the
consequences.
Similar to the What-If/Checklist techniqu e discussed in this section, a HAZOP
is often paired with checklists or ot her “complementary analyses” to further
evaluate specific concerns (e.g., human factors, facility siting).
What-If/Checklist Analysis. The What-If Analysis technique alone, or the
Checklist Analysis technique alone can be an appropriate core methodology for
some processes. However, in most ca ses, the two are combined to take
advantage of both the creative, brainstorming questioning of the What-If
Analysis method and the systematic qu estioning of the Checklist Analysis
method. What-If/Checklist is often used for lower hazard processes, simpler
processes, and processes that do not involve chemical reactions, such as storage
or transfer systems. What-If/Checklist is also appropriate for processes and
activities that do not have rigidly defined material and energy flow paths or that
have hazards not directly tied to proc ess parameters such as flow, level,
temperature, and pressure. Laboratory and pilot plant operations are often
candidates for What-If/Checklist Analysis. This technique is often the first hazard
evaluation method performed on a process (e.g., during early design stage of a
new process), and as such, it can be a precursor for future, more detailed,
studies.
In the What-If portion of the analysis , the team asks questions or voices
concerns about possible undesired events. Their concerns (usually phrased in
the form of what if? questions) are primarily form ulated to identify hazardous
events or specific loss scenarios that could produce an adverse consequence.
There may be no specific pattern or orde r to these questions, unless the leader
systematically directs the team to focus on a particular portion of the process
and/or a particular category of questions (process variable deviations,
equipment failure modes, operating errors , etc.). The questions can also address
any abnormal condition related to the pl ant or procedure, whether caused by
human or equipment performance gaps, or by external events (e.g., floods, high
winds, vehicle impacts, dropped objects).
In the checklist portion of the anal ysis, the team primarily relies on a
standard list(s) of questions developed based on experience with similar
chemicals, equipment, processes, and/or activities. Frequently, checklists are
created by simply organizing inform ation from current relevant codes,
standards, and regulations. The standard questions may be modified to address
specific concerns about the subject process. |
OVERVIEW OF RISK BASED PROCESS SAFETY 51
Personnel and contractors can be multilingual, and communication requires extra
effort.
RBPS Element 18: Measurement and Metrics
Current trends are an important indicator of process safety performance. This
element addresses leading and lagging in dicators of process safety performance,
including incident and near-miss rates as well as metrics that show how well key
process safety elements are being perf ormed. Guidance is available on suitable
metrics from API for downstream (API 754) and IOGP for upstream (IOGP 456).
Upstream-specific metrics are defined such as loss of well control events or
maritime incidents.
Contractors perform much of the work in upstream, so it is important that their
process safety performance is included in the company collected metrics.
API 754 and IOGP 456 are closely aligned and have four tiers of process safety
events with Tiers 1 and 2 being actual losses of containment, fatality or multiple
injury events. Tiers 3 or 4 address demands on safety systems, near miss events,
delayed inspections and maintenance, and deficient safety management systems.
COS has defined additional categories of indicators such as loss of well control, loss
of station keeping, mechanical lifting, and other process safety related indicators.
RBPS Element 19: Auditing
Auditing provides for a periodic review of process safety management system
performance. The aim is to identify gaps in performance and identify improvement
opportunities, and subsequently to track closure of these gaps to completion.
Auditors may be second party (company personnel but independent of operational
roles at the site) or third party (independent company) depending on company
policies or local regulations.
Auditors must have the requisite skills for the upstream activities under audit.
While all audits should address the effective implementation of the safety
management system, there are technical aspect s that also need to be addressed such
as the process safety barriers status, and if offshore, key marine systems. COS
(2014) has developed auditing requirem ents as well as requirements for the
accreditation of Audit Service providers and auditors for SEMS audits, which have
been incorporated by reference by BSEE.
RBPS Element 20: Management Review and Continuous Improvement
Management review and continual improvement is the practice where managers at
all levels set process safety expectations and goals with their personnel and review
performance and progress towards those goal s. This may take place in a leadership
team meeting or one-on-one with the relevant line manager.
Management review and continual improvement are the closing aspect in the
so-called PDCA management loop of Plan-Do-Check-Act.
The review formalizes the link between goals set and results achieved. Where
goals are not achieved then additional actions or investment can be included in the |
2.10 Lear n to Assess and Advance the Culture |67
and Act Upon Hazards/Risks and Learn to Assess and Advance the
Culture.
To combat normalization of deviance, some com panies seek
to instill a keen sense of Operational Discipline (Refs 2.33, 2.34),
i.e., measures to ensure the performance of all required tasks
correctly every time. Operational Discipline may appear to apply
only to operators and mechanics, but in fact it applies to all
employees, including leadership. Leaders also have policies they
m ust adhere to and tasks they must perform . In operational
discipline, leaders have the added responsibility to assure that the
people they manage are perform ing their required tasks.
Establishing high standards of process safety performance is an
important aspect of operational discipline. These standards should
be stressed to new employees in their initial training, and then
reinforced during refresher training and regularly on the job.
Where changing circumstances warrant, standards are carefully
modified, but the normalization of deviance with respect to these
standards is not accepted, and there is zero tolerance for willful
violations of process safety standards, rules, or procedures. Zero
tolerance should be clearly and quickly dem onstrated when
violations occur. Effective Operational Discipline requires
employees at all levels to hold each other accountable for their
parts of the PSMS.
2.10 LEARN TO ASSESS AN D ADVAN CE THE CULTURE Longford, Victoria, Australia, Septem ber 25, 1998
B rittle fracture of a heat exchanger at a gas processing plant led
to an explosion and fire that killed two people and injured
several more (Ref 2.35). The dam age caused the facility to shut
down, leaving homes and businesses in the city of Melbourne
and the state of Victoria with no natural gas supply for weeks. |
5
The railroad desperately needed the explosive to maintain its
construction schedule in the mountai ns. Fortunately, a British chemist,
James Howden, approached Central Pacific and offered to manufacture
nitroglycerine at the construction si te. This is an early example of an
inherently safer design principle – minimize the transport of a hazardous
material by in-situ manufacture at th e point of use. While nitroglycerine
still represented a significant hazard to the workers who manufactured,
transported, and used it at the construction site, the hazard to the general public from nitroglycerine tr ansport was eliminated. At one time,
Howden was manufacturing 100 pounds of nitroglycerine per day at railroad construction sites in the Si erra Nevada Mountains. The Central
Pacific Railroad’s experience with th e use of nitroglycerine was quite
good, with no further fatalities directly attributed to use of the explosive
during the Sierra Nevada construction (Ref 1.22 Rolt), (Ref 1.2 Bain).
By today’s standards, little about 19th Century railroad construction
would qualify as safe, but the in-sit u manufacture of nitroglycerine by
the Central Pacific Railroad did repres ent an advance in inherent safety
for its time. A further, and probab ly more important advance occurred
in 1867, when Alfred Nobel invented dynamite by absorbing nitroglycerine on a carrier, greatly enhancing its stability. This is an
application of another principle of inherently safer design – moderate , by
using a hazardous material in a less hazardous form (Ref 1.9 Henderson).
A milestone event in pr ocess safety and inherent safety was the 1974
Flixborough explosion in the United Kingdom that caused twenty-eight
deaths. On December 14, 1977, inspired by this tragic event, Dr. Trevor
Kletz, who was at that time safety advisor for the ICI Petrochemicals
Division, presented the annual Jubilee Lecture to the Society of Chemical Industry in Widnes, England. His to pic was “What You Don’t Have Can’t
Leak,” and this lecture was the first clear and concise discussion of the concept of inherently safer ch emical processes and plants.
Following the Flixborough explosion, interest in chemical process
industry (CPI) safety increased, from within the industry, as well as from
government regulators an d the general public. Much of the focus of this
interest was on controlling the hazards associated with chemical processes and plants through improv ed procedures, additional safety
instrumented systems and improved emergency response. Kletz |
CASE STUDIES/LESSONS LEARNED 203
The 60 Still Base was fitted with a thermocouple, mounted in a
thermowell from the top of the ve ssel and operated under a vacuum
when the distillation process was underway.
Normal procedure was for distillation to be carried out in two stages.
Initially, the vessel was filled to de pth of some 2.16m (85 inches) and
operators took a sample for thermal analysis before the vacuum was
started and heat applied. Typically, the distillation continued until there
was a 50:50 mixture of the volatile MNTs to the less volatile DNT and
nitrocresols. The thermal analysis wa s used to check for thermal stability
of the residues that was used as a basis for evaluating how far the
distillation would proceed.
Once this first stage was complete, the vacuum would be broken,
and a second charge of Whizzer O il added before additional thermal
testing was done. If the results were acceptable, the vacuum would be
applied and distillation continued until, again, there was a 50:50 mixture
of volatiles to non-volatiles. The fi nal residue in 60 Still Base was taken
away in a tanker for disposal by off-site incineration.
7.3.4 History of Meissner Plant Prior to Incident
7.3.4.1 Process
There were two Meissner plants at the Castleford facility. Meissner I
dated from 1962 with a capacity of 20 tonnes per day and Meissner II
was commissioned in 1972 with a capa city of 60 tonnes per day. The
products from the nitration reaction were first washed with water, to
remove the residual acids, and up until 1988, this was followed by a
caustic wash stage to remove the byproduct nitrocresols, which was
treated and then discharged as an effluent to the local river.
In 1988, the caustic wash stage wa s removed for environmental and
safety reasons. This meant that th e nitrocresols were carried further
through the process. At the same time, they introduced the batch
distillation of the 40 % whizzer oil using 60 Still Base.
The hazards associated with the potential thermal instability of
substances such as DNT and nitrocresols were well known to the
management at the site. Company specialists developed thermal |
Pressure Relief Devices
223
on a P&ID. We however don’t go to detail because this is
a topic in the design stage of a project.
The last item won’t be touched here at all again because
it has nothing to do with P&ID development.
12.9 Locating PRDs
How do we know if a PRD is needed? Is there a legal obli-gation instated by your region’s regulatory body? If so, then determining whether or not you need a PRD is easy: “We have to have it, we are going to install it. ” If not, then the question becomes whether there is a technical obli-gation to include a PRD in the design or not. Therefore, there are two questions you should ask yourself to decide whether a PRD is needed. These two questions are out -
lined in Figure 12.7.
From a purely theoretical point of view, a PRD is
required if the following four criteria are met:
1)
Ther
e is at least one valid overpressure scenario for a
container with trapped fluid.
2) Thi
s overpressure scenario will increase the pressure
of the container to higher than the maximum allowa-
ble working pressure (MAWP) during the life of the container.
3)
Thi
s overpressure scenario does not void the integrity
of the container prematurely, before PRD action.
4) The r
isk of explosion (without a PRD) is higher than
what is tolerable.
However, from a practical standpoint, items 2, 3, and 4
are generally eliminated from the list of PRD requirement criteria. In the case of item 2, this is not always validated, irrespective of whether the overpressure criterion increases the pressure to higher than the MAWP or not; as long as there is one overpressure scenario (item 1) the require-ment for a PRD is confirmed. However, in some cases where the ultimate pressure can easily and accurately be estimated, this criterion can be validated. One example of such a case is thermal expansion of trapped liquid.
For item 3, generally nobody checks if the integrity of
the container is held intact before the opening of the PRD. This is because such an estimation would be dif -
ficult, if not impossible. If someone wants to challenge the elimination of item 3 from the list, he may ask an exaggerated question such as: “do we need to install a PRD for a cardboard container?” Installing a PRD for a cardboard container may seem funny, but in reality, and for general materials of construction in the pro-cess industries, we generally install a PRD for enclo-sures irrespective of their material. So, for cases such as a fiberglass tank in a fire scenario, we still need to install a PRD.
For item 4, we don’t go through the risk analysis to
check whether the risk is higher than our tolerable level to see whether or not we need to install a PRD.
Basically, from a technical point of view, as long as
there is a container with at least one overpressure sce-nario (meaning with trapped fluid – item 1), there is a need to install a PRD.
However, the regulatory body’s view could be different
from the purely theoretical viewpoint and the practical viewpoint. Some regulatory bodies have a practical view, but some of them have a stricter approach. Some regula-tory bodies expect to “see” a PRD on every single con-tainer. Their view is basically: “if there is a container, there should be a PRD. ”
12.10 Positioning PRDs
If needed, where should a PRD be physically placed? Directly on the system to be protected, or far from it? At the top or bottom of the container? If it is not directly on the system to be protected, then what is the maximum allowable connecting pipe length? Which pipe fittings may be installed on this connecting pipe? Vertically or horizontally?
All of these questions about PRD installation are
answered in this section, even though not all of these fea-tures are visible on P&IDs.
The short answer is that the PRD should be placed,
whenever possible, directly on the system to be pro-tected, vertically, upward, and at the top of the container (Figure 12.8).
To expand this short, simplistic answer it should be
said that a PRD should be placed as close as possible to the protected enclosure, and as far away as possible from flow disturbance‐causing elements such as elbows or static mixers. PRDs also should be protected from vibra-tions. Another important requirement is that a PRD should be easily accessible for inspection.
Now the question is, if – for whatever reason – we
cannot put t
he PRD directly on the system to be pro-
tected, where it should be placed? The answer is that it can be installed upstream or downstream of the system or “on” the system through a connecting pipe. In all these case the connection between the to‐be‐protected system
1.Is it illegal not to include a PRD?
2.Is it unwise not to include a PRD?
Figure 12.7 Tw o questions regarding inclusion of a PRD for a
system. |
34 Guidelines for Revalidating a Process Hazard Analysis
• Revalidating the PHA to ensure current company risk practices and
procedures are being followed (If an asset has been acquired by a
new company). This revalidation may be performed on an
accelerated schedule
• Confirming management agreement that the revalidation will use
the same consequences of interest as the prior PHA or revising as
necessary
• Looking for opportunities to maximize the value of PHA
revalidation activities beyond mi nimal regulatory requirements
• Ensuring outside personnel being used for PHA facilitation or as
subject matter experts (SMEs) are knowledgeable in all
requirements prior to initiating the revalidation sessions
Obstacles to Success:
• Trying to simply Update a PHA when fundamental regulatory or
organizational requirements have changed
• Trying to only revalidate a portion of the process when
requirements have changed
• Trying to revalidate a PHA when organizational requirements are in
flux due to a merger or acquisition; prior to initiating a revalidation,
the requirements should be defined
• Neglecting to identify and explain requirements to the revalidation
team (including contractors and SMEs) prior to the meeting |
11.6 Prepare | 151
Lucas said, “Charlotte, there’s a bit of a problem here. Two ten-hour work
shifts only add up to twenty hours. What happens to the other four hours in
the day?” Charlotte flushed red with embarrassment and said, “I guess in my
eagerness to come up with a solution, I didn’t think about that. I only ruled out
the eight-hour work shift because it increased our crew complement by fifty
percent.”
Lucas said, “Let’s not panic here. I think I have a solution. When I served in
the Royal Australian Navy, we broke the overnight watch into two shorter
periods we called the dog watches.” Lucas went on to explain, “From midnight
to 0400 we could have a small skeleton crew with work limited to pumping oil
and station-keeping. It might add to the crew complement, but maybe we can
assign them tasks that others do, and we’ll come out even. I think this could
work. What do you think, Oliver?”
Oliver looked at it carefully and said, “It doesn’t look like a big change, but
I’d like to try it out first before fully committing to it.” Lucas said, “That’s all
we’re asking for. I think you’ll find that with more rest, your crew will operate
more safely and more efficiently.” Oliver said, “I hope you’re right. To see if it
makes a difference, I’d like to see at how this new shift pattern compares with
the previous one.”
Charlotte chimed in, saying, “That’s a perfect segue into the next part of
my proposed plan. We don’t regularly do fatigue risk management surveys, yet
it’s one of the key elements in API 755. I suggest that we collaboratively
develop targets, like percent overtime and number of open shifts, so we can
measure the impact of the change. Then, we conduct the survey on an annual
basis, rather than the ad hoc way we’ve been doing it.” Charlotte continued,
“We should also look at other factors, such as healthcare costs and safety
data.” Oliver said, “Looks like you’ve come thoroughly prepared. I’m
comfortable bringing this to Mason.”
Now it was Lucas’s turn to talk about trying to keep safety in the forefront
of people’s minds. Lucas said, “Creating a sense of vulnerability is a big
challenge, but it’s imperative in keeping all of us safe on this rig.” He continued,
“Everybody retains information differently. Some people like to read, while
others like to watch a video or listen to a talk. I suggest we develop various
ways to remind people of what can happen if they don’t operate safely. But
first, we need some shock value to wake everyone up. I think watching a movie
on the Macondo catastrophe could do that. We’d need to have a follow-up
discussion to really get the message across.” Lucas then laid out the rest of his
plan: |
62 GUIDELINES FOR MANAGING ABNORMAL SITUATIONS
Example Incident 3.9 – Hydrocracker Operations
In an oil refinery hydrocrackin g unit, a feed pump failure had
historically been addressed by simp ly starting the spare pump, then
addressing any temperature excursio ns that might occur during the
event in the customary manner. Howe ver, a modification was made in
which the upper heater chamber was combined wi th the chamber
from another heater. The redesign was not approved by the licensor.
The effect of the change was to allow heat to flow from one side of the
heater to the other, so that in a tr ip situation, heat continued to flow
where it was not required. The first time the feed pump tripped, and
the standard pump trip response was performed, the result was an
extreme reactor runaway condition (from 400 oC to 1000 oC). This
showed the need for an all-purp ose emergency shutdown procedure
that could be applied under any circumstance.
Lessons learned in relation to abnormal situation management:
Management of Change: Implementing a change without
considering input from the licensor on the potential effects of the
change.
Root Cause Analysis: Less severe versions of this event had
occurred at other similar units at other locations having the same
design change in the preceding mo nths. An analysis of the failure
causes and an adjustment to the default response would have
mitigated the damage that resulted.
Procedures: The procedure for star ting the spare pump was simply
based on historical experience wi th a more conventional heater
design. This should have been re-examined as part of the
Management of Change process.
|
EQUIPMENT FAILURE 211
Fired Equipment
Overview. Fired equipment, such as furnaces, flares , incinerators, thermal oxidizers, or heat
transfer fluid heaters, are used to provide heat to processes, and dispose of combustible waste
streams. For example, natural gas is converted to hydrogen in a reformer, which is a series of
catalyst-packed tubes heated to several hundre d degrees centigrade by a burner. Other uses
of fired equipment include steam generation and heating of distillation column reboilers.
A common failure mode is accumulation of unburnt fuel due to loss of flame, too much
fuel being fed, or insufficient air (oxygen) as examples. The unburnt fuel can then ignite and
cause a fire or explosion. The second most co mmon failure mode is tube failure, which can be
caused by overheating, flame im pingement, improper firing, th ermal cycling, thermal shock,
or corrosion. This can also result in fires an d explosions. Liquid carry -over into flares and
incinerators can also cause expl osions as experienced in the Te xaco Milford Haven Explosion.
(HSE 1994)
Example 1. A heater was heavily damaged during startup as a result of a firebox explosion
(Figure 11.26). The operator had some difficult y with the instrumentation and decided to
complete the start up by bypassing the safety interlocks. This allowed the fuel line to be
commissioned with the pilots out. The main ga s valve was opened, and gas filled the heater.
The heater exploded destroying the casing and several tubes. Fortunately, no one was injured.
(CCPS a)
Figure 11.26. Damaged heater
(CCPS c)
Example 2. An explosion destroyed the furnace and adjacent column at a NOVA Chemical
Bayport, TX plant (Figure 11.27). Before the ex plosion, an operator noticed flame stability
problems with the low NO x burners and began to manually adjust the airflow. During the few
minutes that adjustments were being made to manage the burners, a loud puff was heard
followed by a major explosion in the furnace. It appears that the explosion was caused by
clogging in the nozzles on the burners resulting in an unstable flame. (CCPS 2004)
|
263
Consequence Analysis
Learning Objectives
The learning objectives of this chapter:
Understand basic concepts of source, rele ase, dispersion, consequence and impact
modeling, and
Know where to access tools to support consequence analysis.
Incident: DPC Enterprises L.P. Chlorin e Release, Festus, Missouri, 2002
Incident Summary
On the morning of August 14, 2002, a chlorine transfer hose failed releasing 21,772 kg (48,000
lb) of chlorine over a 3-hour period during a railroad tank car unloading operation at DPC
Enterprises, L.P., near Festus, Missouri. The facilit y repackages bulk dry liquid chlorine into 0.9
metric ton (1 ton) containers and 68 kg (150 lb) cylinders for commercial, industrial, and
municipal use in the St. Louis metropolitan area.
Key Points:
Asset Integrity and Reliability – Did you get what you purchased? It is
often difficult to simply visually determine if that pipe, hose, or valve is what
you thought you were purchasing. Posi tive Material Identification (PMI)
should be used to verify that material s are delivered as specified, especially
where use of an incorrect material may lead to failure.
Emergency Management – We are in it together. Emergency response
plans and drills should recognize and test the assets and limitations of the
neighboring emergency response capabilities.
Asset Integrity and Reliability – Will your Emergency Shutdown (ESD)
system work in an emergency? ESD system design should consider the
operating and environmental conditio ns, including that of upstream
equipment that might impact the system. ESD system testing should verify
that the entire system works - from a sensor or button to the closing of a
valve.
Chlorine is a toxic chemical. Co ncentrations as low as 10 parts per million are classified as
“immediately dangerous to life or health”. The wi nd direction on the day of the release carried
the majority of the chlorine plume away from neighboring residential areas; however, some
areas were evacuated. Sixty-three people fr om the surrounding community sought medical
evaluation at the local hospital for respirator y distress, and three were admitted for overnight
observation. The release affected hundreds of other nearby residents and employees, and the
community was advised to shelter-in-place for 4 hours. Traffic was halted on Interstate 55 for
1.5 hours. Three DPC workers received minor skin exposure to chlorine during cleanup
activities. (CSB 2003) |
288 | Appendix E Process Safety Culture Case Histories
The PSMS does not address:
How the significant risk of the site is m anaged; the focus is
solely on com pleting the required docum entation.
Procedures and policies exist where specifically required
by the regulation.
Hazardous m aterials that are not covered by the
regulation
Parts of the process outside the covered boundaries that
could impact risk, such as such as cooling water, power,
and nitrogen.
Equipment for processing the final products, which are not
addressed in the regulations.
PHAs are performed using sim ple checklists because the
regulations allow it, and result in little m ore than short m emos
with brief checklists attached. Audits are com pleted relatively
quickly, and produce short reports with no m ore than three
findings. The incident investigation file contains no investigation
reports and no metrics are collected.
Do you believe the facility has had no incidents? How could
they avoid them ? Such an approach m ay reduce the regulatory
exposure for a time, and it certainly may seem sim pler and
cheaper. However, it ignores significant risks inherent to the
com pany’s processes. The notion that strict com pliance with
regulations will reduce the process safety risk to a low level is a
false belief and an indicator of a poor culture.
Establish an Imperative for Safety, Provide Strong Leadership,
Maintain a Sense of Vulnerability, Understand and Act Upon
Hazards/Risks.
E.2 – Peer Pressure to Startup
A facility has a com bined MOC/PSSR process (a
com mon practice). This process is m anaged
electronically, routing the MOC package via e-m ail B ased on
Real
Situations |
Piping and Instrumentation Diagram Development
58
These actions can be done manually after an operator
or an automatic control system checks the quality of the
product. Any of the aforementioned solutions can be used. However, the more important point is that the cases of low‐efficiency operation should be considered for each unit that creates a physical or chemical change on the process stream.
5.4.4
Star
t‐Up Operations
There are at least two types of start‐up operations: the
first start‐up of a plant after its construction, which is called commissioning, and the start‐ups after each
shutdown.
The start‐ups after shutdowns can be in two different
types: start‐up after a planned shutdown and start‐up after emergency shutdowns. A start‐up after an emer -
gency shutdown may have more steps than a start‐up after a planned shutdown.
Start‐up operations can be assumed to be a severe
capacity reduction case. In this situation, not all the instruments will work properly because process param-eters during the start‐up are not necessarily within their range. However, the operating personnel who are attending the plant are the resources available and expected to compensate for the lack of instrument operability. They will be present as a larger group dur -
ing normal operation.
The commissioning operation is a specific case of a
start‐up. In addition to all issues related to a start‐up, a commissioning has many other problems related with poor construction and installation. During the plant construction, the electric motors may be connected to
Mechanical r elief action
Mechanical relief actionSlS action
SlS actionAlarm
AlarmBPCS actionsHigh–high flowHSl le vel
High flowRangeability (R)
Normal flow
Low flow
Low–low flow
LSI le vel
Figure 5.19 Conc ept of rangeability.
Unit
UnitVs.
Unit
Figure 5.20 Using par allel equipment to provide a required
turndown ratio.
FCSplit range 0–50%
50–100%
Figure 5.21 A con trol valve arrangement in wide rangeability
requirement.
Vs.
Recirculation pumpUnit
Unit
Figure 5.22 Recir culating for increasing TDR.
FC
Figure 5.23 A cen trifugal pump minimum flow recirculation to
increase TDR. |
Piping and Instrumentation Diagram Development
190
However it should be realized that in this arrange-
ment, the fluctuation in the pump service impacts the
mixing and its effectiveness.
10.7 Liquid Mo vers: PD Pumps
Positive displacement pumps are not as common as centrifugal pumps in industry. When there is a need for a pump in industry the first choice is a centrifugal pump by default. PD pumps are not very user friendly and operators generally don’t like them. In similar capacities they are more expensive. However, they have several advantages over centrifugal pumps. They have the least sensitivity toward cavitation. This is the most important advantage of PD pumps over centrifugal pumps. The other advantage is that they are not sensitive toward a low flow rate. They are also good if the viscosity of fluid is high or very high and/or when a very high discharge pressure is needed. There are two main classes of PD pumps: reciprocating pumps and rotary pumps.
Even though PD pumps provide a lower flow rate than
centrifugal pumps in comparison with reciprocating pumps and rotary pumps, reciprocating pumps are gen-erally used when a higher flow rate than a rotary pump is needed. However, the reciprocating pumps generate lower discharge pressure than rotary pumps.
One weakness of reciprocating pumps is that they
create a pulsating flow. This pulsation is because of the forward and backward movement of the pump’s motive member (Figure 10.29). For example, when you are buying a reciprocating pump whose operating pressure is 500 psi, the pressure that is sensed by the elements installed on the discharge side of the pump could be between 400 and 600 psi. This is a very important point when you decide to install sensors on the
dis
charge side of reciprocating pumps. The pulsation
can be mitiga
ted by installing a “pulsation dampener”
on the discharge side of the reciprocating pump; how -
ever, the issue still exists in the pipe upstream of the pulsation dampener.
The issue of pulsation also exists on the suction side of
reciprocating pumps, and a pulsation dampener is some-times needed on the suction side of them too.
Rotary pumps can work in very high viscosity services
and/or when a high discharge pressure is required. Actually, amongst the different types of pumps, rotary pumps can generate the highest discharge pressure but they have the lowest capacity. One famous application for rotary pumps is oil pumps in cars. In cars, the circu-lation of oil is necessary for lubrication and cooling purposes, and usually a pump is used for this. Because the flow rate is very small and at the same time the pres-sure needs to be high to be able to send the oil through the capillary pipes around the cylinder blocks, the best choice of pump here is a rotary type pump.
Rotary pumps are available in different types includ-
ing: screw pump, lobe pump, and peristaltic pump (hose pump).
The various types of PD pumps are shown in
Figure 10.30.
However, it is important to know that not all PD pumps
can be clearly classified in one of these two classes of reciprocating pumps or rotary pumps. One exception is progressive cavity pumps. Progressive cavity pumps are basically a hybrid type of PD pump. These pumps are basically a mixture of reciprocating pumps and rotary pumps. As these pumps were primarily invented by a French engineer named Mono these pumps were named by people as the Mono pump.
Inert ga sMembrane
System /f_luid
MT
Figure 10.29 Pulsa tion in reciprocating pumps.
Reciprocating pumps Rotary pumps
Piston pumps
Plunger pumps
Diaphragm pumpsGear pump
Peristaltic pump
Progressive cavity pump
Figure 10.30 Symbols of v arious PD pumps.
M
Figure 10.28 Cen trifugal pump also working as a mixer. |
Part 6: Non-technical skills Human Factors Handbook For Process Plant Operations: Improving Process Safety and System
Performance CCPS.
© 2022 CCPS. Published 2022 The American Institute of Chemical Engineers. |
APPENDIX D – EXAM PLE CASE STUDY 369
The selected team initially established a specific plan of investigation
procedures for this occurrence. This st rategy session list ed priorities and
necessary actions to ensure that all required information was obtained in a
prompt manner. Needless delays in ev idence collection were avoided by the
use of this plan.
The investigation team visited the scene of th is incident before the
physical evidence could be distur bed. The maintenance foreman was given
the duty of taking photographs of the damaged area. He was careful to
obtain overall views of the scene and individual equipment and logged
where each photo was taken. All team members were provided with a field
investigative kit and appropriate safety protective gear. Important evidence
was gathered, preserved, and identified using a written log and tagging
system. A plot plan was posted and the location of each physical piece of
evidence was noted on the plan along with the tag number.
On completion of this task, preparatory work was performed by the team
members for preliminary witness intervie ws. The importance of focusing on
confidentiality and fact finding, wh ile avoiding assigning blame, was
emphasized to team members prior to conducting interviews. Two team
members, the safety supervisor and one other as available, were chosen to
meet with the witnesses. A small conf erence room in the Administration
Building was allocated for this project. The setting was arranged informally
to allow the person involved to feel at ease. After considerable debate within
the team, a conclusion was reached to not use a tape recorder during the
witness interviews. The interview proc ess was started early the evening of
incident and was continued throughout the next two days. At the end of
each day, the investigation team met to discuss the information obtained from the interviews and other activities.
The catalyst preparation area superv isor, on-duty control room operator
for the catalyst operation, and ma intenance superintendent were key
sources of information. Their written records and logs were examined in detail. Other personnel who were interviewed included two outside operators, fire brigade members, and associated maintenance employees.
During
these conversations, spec ial attention was paid to nonverbal signals.
The interview process generated seve ral unanswered questions about
operational and maintenance procedu res that required further study.
Second interviews, further evidence collection and examination, and
thorough evaluations of operatio nal and maintenance records were
conducted to try to find explanatio ns for the question s created by the
preliminary witness interviews. Due to a high pressure alarm occurring at
Kettle No. 3 in the catalyst preparation area prior to the fire, an analysis of the software and hardware for the cont rol panel that oversees this process
was deemed essential for this study. |
11 Determining operational co mpetency requirements
11.1 Learning objectives of this Chapter
This chapter focuses on competency of people performing safety critical tasks.
Being competent to perform all tasks, not solely the safety critical tasks, is
important, as noted in the CCPS “Gui delines for Defining Process Safety
Competency Requirements” (2015) [50]. Th e “Guidelines for Risk Based Process
Safety” [24] puts forth the approach of a llocating effort commensurate to the level
of risk. The IChemE “Process Safety Competency Guidance” [51] also offers useful
information on how to build and develop process safety competency.
There may be many tasks in a large proc ess plant. It is appropriate to spend
more effort on tasks that, if conducted poorly, could result in greater potential
consequences. Therefore, these Chapters on competence focus on safety critical
tasks.
The success of Competency Management is supported by the identification of
safety critical tasks, operators’ competency, and the performance standards
required for these tasks.
By the end of this Chapter, the reader should be able to:
• Identify the level of learning for low and high-risk tasks.
• Explain how to identify required competency for safety critical tasks.
• Define performance standards.
11.2 Identify and define safety critical competency: overview
Activities in the process industry are of ten carried out in difficult conditions.
Hazardous environments, complex processe s, and production pressures demand
higher levels of competency.
Prior to defining competency requirem ents, tasks and activities should be
ranked in terms of their safety criticality. This can be done by conducting Safety
Critical Task Analysis and Hazard Identifi cation and Risk Analysis, as outlined in
Chapter 6.
Human Factors Handbook For Process Plant Operations: Improving Process Safety and System
Performance CCPS.
© 2022 CCPS. Published 2022 The American Institute of Chemical Engineers. |
312 Human Factors Handbook
• 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.
24.4.4 Tips for recognizing chan ge and potential impacts
MoC procedures typically rely on someone recognizing that a change is planned or
is occurring, and then initiating the MoC process. All managers, team leaders and
supervisors of operations, production, maintenance and associated functions
should be conscious of what constitutes change and aware of the potential impact
of change. This is true regardless of whether they initiate change or whether
change is initiated by others.
A proactive approach is required, and it is important to keep an open mind
about potential impacts on human performa nce. Some tips on recognizing change
are given in Table 24-1.
Table G-1 in Appendix G, provides a list of typical changes, examples of their
Human Factors impacts, and typical Human Factors actions. This table can be used
to identify changes and recognize their potential Human Factors impacts.
|
340
A key step in selecting a decision aid is to understand the aspects of
the problem to be addressed. These aspects include:
Resource Availability: The time an d analytic resources to address
the problem
Problem Complexity: The number of alternatives, the complexity
of the system, and the amount of uncertainty in the problem
Importance/Scrutiny: How sensitive decision makers and
stakeholders are to the deci sion, and how extensively the
decision will be reviewed
Group Involvement: The desire of the organization to involve multiple decision makers or incorporate input from multiple stakeholders in the decision
Need for Quantification: The desi re to be able to point to a
quantitative basis for choosing one alternative over another.
In addition to these five problem aspects, there may be constraints
that affect the selection of a deci sion aid. Constraints can include
organizational guidelines on the type of analysis to be performed and
the types of issues to be addressed.
Specific criteria can be used to evaluate the appropriateness of
various decision aids to a given prob lem. Some key characteristics that
differentiate decision aids include:
Resource Requirements : The time, budget, and effort required to
use the decision aid
Depth of Analysis/Complexity : The detail and explicitness with
which important aspects of the problem are addressed, and the
complexity of doing a complet e and thorough analysis
Logical Rigor : The mathematical soundness and logical rigor of
the analysis
Group Focus : The ability to incorporate group opinions, handle
problems with more than one decision maker, and address
competing objectives
Quantitativeness : The ability to provide a quantitative basis for
the decision, accommodate sensitivity analyses, and address inherently quantitative decisions such as resource allocation problems |
HUMAN FACTORS 371
Explain the three parts of situational awareness.
In times of financial challenges, companie s may try to “do more with less”. Which
human factors topics should be co nsidered? Explain your answers.
Consider an engineering design project. What human factors topics can influence the
human performance on the project?
Describe at least 3 ways human factors coul d contribute to process safety incidents
during a) a unit startup, b) normal operation, and c) unit shutdown.
References
Bridges 2010, “Human Factors Elements Missing from Process Safety Management (PSM)”,
American Institute of Chemical Engineers 2010 Sp ring Meeting, San Antonio, Texas, March 22-
24.
CAA, Fundamental Human Factors Concepts, CAP 719, Civil Aviation Au thority Safety Regulation
Group, http://publicapps.caa .co.uk/docs/33/CAP719.PDF.
CSB 2007, “Vinyl Chloride Monomer Explosion, Investigation”, Report No. 2004-10-I-IL, U.S.
Chemical Safety and Hazard Investigation Board, March.
CCPS Glossary, “CCPS Process Safety Glossary ”, Center for Chemical Process Safety,
https://www.aiche.org/ccps/resources/glossary .
CCPS 2007, Human Factors Methods for Improving Performance in the Process Industries , Center
for Chemical Process Safety, Jo hn Wiley & Sons, Hoboken, N.J.
CCPS 2012, “Building Process Safety Culture Tool Kit, Incident Summary: Challenger Case”,
https://www.aiche.org/ccps/topics/eleme nts-process-safety/commitment-process-
safety/process-safety-culture/challenger-case-history.
CCPS expected 2022, Human Factors for Process Pl ant Operations: A Handbook , Center for
Chemical Process Safety, John Wiley & Sons, Hoboken, N.J.
Endsley, https://upload.wikimedia.org/wi kipedia/en/6/61/Endsley-SA-model.jpg.
HSE a, “Leadership and worker involvement toolkit, “Understanding human failure”,
https://www.hse.gov.uk/construction/lwi t/assets/downloads/human-failure.pdf.
HSE 2000, “Human Factors Assessment of Safety Critical Tasks”, Offshore Technology Report
– OTO 1999 092, U.K. Health and Safety Executive, 2000,
https://www.hse.gov.uk/research/otopdf/1999/oto99092.pdf
HSE 2009, “Review of human reliability assessment methods”, RR679, U.K. Health and Safety
Executive, https://www.hse.gov.uk/research/rrpdf/rr679.pdf.
INPO, Human Performance Reference Manual , INPO 06-003, Institute of Nuclear Power
Operations, October 2006.
Miller 2019, “Helping humans get it right”, Miller, L. and Grounds, C., Process Safety Progress ,
38: e12003. Viewed December 9, 2020, https://doi.org/10.1002/prs.12003
NOPSEMA 2020, “Critical Task Analysis”, N-06300-IP1704 A500978, Australia National
Offshore Petroleum Safety and Environmental Management Authority. |
Appendices 195
Example Facility Siting Checklist 2
OBJECTIVE: To aid in identifying facility siting issues relevant to process safety
issues within the scope of the PHA. Th e PHA team completing this checklist
should seek assistance from subject matter experts as necessary.
Q T R
I. Process/Unit Assessment
Have the following areas been evaluated for potential explosion, fire,
and toxic hazards in the overall facility siting study
• Control rooms? (See Section II)
• Operating units?
• Equipment (including flare and fired heater systems)?
• Tank farms?
• Process sewers?
• Motor control centers (MCCs)?
• Maintenance shops and established areas for cutting and
welding?
• Transportation routes (roads, rail spurs, etc.) within the site?
• Pipe bridges located over occupied structures?
• Fire protection equipment (including firehouse)?
• Other occupied buildings (i.e., offices, warehouses,
laboratories, analyzer rooms)?
Are there vents or relief devices that discharge to the atmosphere? If
yes, evaluate the following items:
• Are employees or equipment near the discharge point,
including downwind locations (e.g., on the ground, on
platforms, on tanks or towers, on roofs)?
• Are occupied building or control room air intakes in the
vicinity of release points?
Has the potential for vapor or liqui d accumulations been considered
for inert, toxic, or flammable chem icals? The following areas should
be considered:
• Drains
• Spill containment areas (e.g., dikes, berms, ponds, remote
impoundments)
• High points in buildings or canopies over lighter-than-air
flammables |
32 | 3 Obstacles to Learning
cost and business pressures can make less thick-skinned individuals reluctant
to seek knowledge from past incidents and instead accept the status quo.
Reverse Incentives
The best process safety management systems use metrics to track
performance and guide improvement. It can be tempting to also use metrics
to offer leaders and workers incentives towards better process safety
performance. However, incentives can have reverse effects if they are not
carefully designed.
For example, imagine that a plant had the goal of reducing incidents by
closing the PSMS gaps that led to near-misses. If personnel were given an
incentive based on decreasing the number of near-misses, it could drive them
to hide near-misses. This would almost certainly lead to increased incidents.
Similarly, the plant might have a goal to increase the closure of action
items from process hazard analyses (PHAs), audits, incident investigations,
and so on. Closure of action items should reduce the incidence of both near-
misses and incidents. However, unless the plant also implements a system for
verifying that action items are closed properly, personnel could be tempted to
check the box rather than close action items with effective measures.
Leaner Organizations
In the continuing effort to improve efficiency, plants and corporate groups
continue to grow leaner. Companies expect fewer personnel to cover a
broader range of responsibilities, often aided by automation. Often, as
organizations grow, the existing personnel must cover expanded
responsibilities. This leaves personnel with less time to study the literature and
think strategically about improvement. Even when they have time, personnel
may be reluctant to consider improvements, especially if it will mean
additional work for them. If the company wishes to continuously learn from
process safety incidents, it must provide key personnel the time they need to
study past incidents and develop ways to integrate those learnings into the
way the company operates.
Risk Misperception
The risk of process safety incidents is driven by low probabilities and high
consequences. Most other risks that companies manage, including
occupational safety, have higher probabilities but relatively lower
consequences. Process safety incidents are relatively rarer than other adverse |
168
and pipe routing. Pipe size should be sufficient to convey the required
amount of material and no larger. Ho wever, small bore piping is less
robust and less tolerant of physical abuse when compared to large
diameter piping, and additional attention to proper support and installation is required. In some cases, such as chlorine for water
treatment applications, it may be possi ble to transfer material as a gas
rather than a liquid, greatly reduci ng inventory in the transfer line.
Chemical inventory can also be minimized by removing dead leg
piping that is no longer needed in the process. Dead legs may experience
enhanced galvanic cell corrosion. Other damage mechanisms, e.g.,
corrosion under insulation are usually more pronounced in piping dead
legs. Additionally, dead leg piping ca n trap heavier substances (by means
of density or solid settling) that can subsequently create unknown
hazards, such as freezing and expansio n of water in dead leg piping that
can cause pipe failure upon thawing (Ref. 8.77 CSB).
Piping systems should be desi gned to minimize the use of
components which are likely to leak or fail. Sight glasses and flexible
connectors such as hoses and bellows should be eliminated wherever
possible. Where these devices must be used, they must be specified in detail, so they are structurally ro bust, have the same temperature and
pressure ratings as the fixed piping (or as close as possible), are
compatible with process fluids, and ar e installed to minimize the risk of
external damage or impact. In additi on, these more fragile elements of
the piping system should be inspected on a more frequent basis than the rest of the piping system. Caps an d plugs should be utilized for vents
and drains that are used infrequently.
Where flanges are necessary, spir al wound gaskets and flexible
graphite type gaskets are preferred. The construction of these gaskets
makes them less likely to fail catastro phically resulting in a large leak.
Proper installation of spiral wound ga skets, particularly the torquing of
the flange bolts, is import ant in preventing leaks.
|
5 • Facility Shutdowns 78
personnel who will have to use the new systems every day. Control
system upgrades tend to add anxiety about starting up the process on
the new system. The situation tends to be worse for individuals not
savy enough to adopt to the newer technology. Thus, it is critical that operators and engineers receive si mulator-based training rigorous
enough to create confidence abou t using the new capabilities.
5.3.4 Mothballing projects
Mothballed processes and equipmen t are temporarily shutdown for
an unknown period of time, requ iring some form of preservation
during the shutdown period. Special procedures and checklists may be
required, including lists of the steps required for cleaning or isolating
the equipment before idling the pe riod for an unknown period. The
main challenge for mothballed equipment is preventing deterioration so that the facility may be safely put back into production. Sometimes
efforts on the larger capital proj ects currently underway should be
stopped temporarily with the project completed at a later date. Such
costly delays may be due to issues beyond the company’s control, including unexpected, sudden, or adverse economic conditions, or
significant technology, construction, or installation issues which were
not anticipated in the original sc ope (and, hence, were not budgeted
when the capital project wa s originally approved).
The type of preservation techniqu es needed will depend on the
stage that the mothballed equipmen t is in when the mothballing will
occur. Preservation procedures will differ if the equipment is still in
construction stage, if the equipment has been received and is staged
for installation, if the equipment is part ially or fully installed in the field,
or if, for existing processes, the equipment has been operating for some time and needs to be cleaned before being mothballed. Depending on the equipment metallurgies and location (especially |
DETERM INING ROOT CAUSES 239
could have contributed to the inciden t causation and progression. Some of
these sequences acted with di rect impact on the trigger event, the pipe
failing and initial fire, while others acted to increase the severity.
The incident investigation team’s complete report is attached in the
Appendix D, and details of the root causes are di scussed. The root causes of
the incident were related to severa l process safety management areas:
• Asset integrity and reliability
• Contractor management
• Emergency management
• Hazard identification and risk analysis
• Management of change
Take the time to review the comp lete example in the Appendix. Look at
the trees and think about what the ro ot causes might have been if the chosen
top event had been the release of isopentane. Would the team have made a
recommendation about escorting and training contractors?
10.6.2 Data-Driven Cause Analysis
Another approach to root cause determination is to use historical data to
infer or identify potential causes. In this case, the investigators use past
experience to look for patterns that support or refute failure hypotheses. The
technique is only as good as the records, and if da ta have not been put in
the files or are in error, then misleading inferences may result. In addition, if
this type of event has not occurred before, the approach cannot be applied.
Failure data for the system under investigation are presented in a timeline
that can be correlated with overall plan t history. Two types of evidence are
sought:
Evidence for correlation with pl ant state, plant condition, or
external environmental effects.
Evidence that indicates a failure pattern that may correlate with
maintenance activities.
The following case study illustrates data-driven cause analysis using
historical data to identify potential causes.
|
Pressure Relief Devices
229
PRDs with different sizes in parallel and all functioning
should be used.
●When one PSV doesn’t provide the required “probabil-ity of failure on demand” for the system two (or more) PRDs with different sizes in parallel and all functioning should be used.
●When more than one scenario is valid and each of them calls for a different type of PSD two (or more) PRDs with different types in parallel and all function-ing should be used.
●When one PSV cannot satisfy the required reliability of the system. In this case, again, we are going to have spar -
ing philosophies like 2
× 50% or 3 × 33% for t
he parallel
PRDs. The spare PSDs are used where there is a need to inspect, recalibrate, and repair a PSD during plant oper -
ation and the “manual safety system” is not acceptable.
When there are spare PRDs each of them should have
isolation valves. The critical items are the isolations valves upstream of each PRD. The functioning PRD should have car seal open CSO attributes and the spare PRD car seal close (CSC). We have to make sure that when one upstream isolation valve is open the other one is closed. Sometimes this is done by placing one “change‐over valve” to make sure no mistakes happen (Figure 12.18).PSDs can be placed in series mainly for one reason: to
isolate the main PSD from the aggressive or dirty pro-cess fluid. In this arrangement a rupture disk is almost always placed upstream of (below of) the main system protector, which could be a PSV or a rupture disk (Figure 12.19(a)).
If there is rupture disk leakage the main purpose of the
rupture disk is missed; the connecting pipe of rupture disk to the PSV should be “checked. ”
As there is always a chance of rupture disk leakage that
the operators are unaware of the “space” should be monitored.
The most common solution is installing a pressure
gauge on the space. A more advanced solution is install-ing a pressure switch or pressure sensor to warn of the increase in pressure (because of the rupture disk leakage) to the control room (Figure 12.19(b)).
The other solution is to connect the rupture disk to a
burst sensor. The burst sensor sends a signal to the con-trol room in the case of rupture in the rupture disk (Figure 12.19(c)).
In Figure 12.19(d) a pressure alarm warns the opera-
tors if high pressure is created in the pipe between the rupture disk and the safety valve.
For more severe leakages a pipe can be connected to
the rupture disk–PSV pipe. On this piece of pipe there should be some systems. One example is placing an excess flow valve on it (Figure 12.20).
Each of above solutions has advantages and disadvan-
tages and one may decide to use a combination of them (Figure 12.21).
In less severe cases where the fluid is only mildly pre-
cipitating, a “flush ring” can work perfectly without the need to go to the more complicated case of using the combined rupture disk and PSV (Figure 12.22).
If there is a chance of contacting aggressive fluid on the
downstream of the PSV, there could be another rupture disk installed on the PSV outlet (Figure 12.23).
CSO, FP
When relief value is functioning During relief value inspectionNCCSO, FP ClosedClosedFigure 12.17 A PRD with pro visions for inline
care for the PRD.
CSO2×100%
CSCActive2×100%
Back up
Change-over valve
Figure 12.18 2 × 100% spare PRDs. |
7.8 Naturalistic Intelligence | 101
Another form of intrapersonal learning uses games. To address learning
related to person-down incidents in this way, the person-down exercise at the
end of Section 7.5 could be turned into an online game.
7.8 Naturalistic Intelligence
Employees with naturalistic intelligence may view the process, PSMS,
standards, and policies as comparable to living, breathing organisms. By
extension they would see process safety as the care, feeding, and well-being
of those organisms and view lessons learned in the context of process
health. People whose strength is naturalistic intelligence receive messages
better in a natural setting or when expressed using animals or plants.
Figure 7.5 shows how the person-down message might be communicated
effectively to these learners.
Figure 7.5 Example Naturalistic Communication About Person-Down
The poster created for the case presented in Chapter 9 uses a monkey to
remind employees to not climb on equipment, piping, and hoses.
Naturalistic learners may not be as interested in numerical details as
logical-mathematical learners but will still be interested in how the
components of the system work together and how the lessons learned will
help keep the overall system healthy.
|
7.3 Process Safety Culture and Organizational Excellence |249
achievable. With engaged and committed leadership, effective
processes and an OE culture, we can achieve our objectives in
operational excellence.”
ExxonMobil (Ref 7.4)
“All operating organizations are required to maintain the systems
and practices needed to conform to the expectations described in
the OIMS (Operations Integrity Management System) Framework.
To drive continuous improvement, the Framework is periodically
updated. This revision strengthens Framework Expectations with
respect to leadership, process safety, environmental performance,
and the assessment of OIMS effectiveness and is intended to: reinforce our belief that all safety, health and environmental
incidents are preventable; and to
promote and maintain a work environment in which each of
us accepts personal responsibility for our own safety and that
of our colleagues, and in which everyone actively intervenes to
ensure the safety, security and wellness of others.”
DuPont (Ref 7.5)
“Operational Excellence (OE) is an integrated management system
developed by DuPont that drives business productivity by applying
proven practices and pr ocedures in three ‘foundation blocks’ –
Asset Productivity, Capital Effectiveness, and Operations Risk
Management.
“The OE management system gives a company the benefits of
lower costs, increased efficiencies, fewer injuries, maximum
sustainable returns on operating assets, and an enhanced
competitive position.
“Our integrated OE management system can be applied to existing
facilities, new facilities, and facility expansions. OE gives an
organization these advantages: •
• |
386
improved tools and metrics for evaluati ng the contribution of IS to risk
reduction and econom ic assessments.
14.5 REFERENCES
14.1 California Code of Regula tions, California Office of
Emergency Services, Accidental Release Prevention (CalARP) Regulations ,
Title 19, Division 2, Chapter 4.5.
14.2 California Code of Regulations, Process Safety Management
for Petroleum Refineries , Title 8, Section 5189.
14.3 City of Richmond, CA, Richmond Industrial Safety Ordinance
(RISO), Municipal Code, Chapter 6.43.
14.4 Contra Costa County (CCC), California, Industrial Safety
Ordinance (ISO), County Ordinance Chapter 450-8.
14.5 Contra Costa Health Services (CCHS), California, Industrial
Safety Ordinance Guidance Document, Section D, June 2011.
14.6 Contra Costa County, California, Industrial Safety Ordinance
Annual Performance Review & Evaluation Report , Contra Costa
Community Health Hazardous Materials Programs, 2017.
14.7 Department of Homeland Security, Interim final rule
implementing the Chemical Facility Anti-Terrorism Standard, April 9,
2007, 72 Fed. Reg. 17718.
14.8 Environmental Protection Agency, Final Risk Management
Rule, 40 CFR 68, January 13, 2017, 82 Fed. Reg. 4594.
14.9 Hendershot, D.C., Safety through design in the chemical
process industry: Inherently safer process design Presented at the Benchmarks for World Class Safe ty Through Design Symposium,
Institute for Safety Through Design, National Safety Council, 1997.
14.10 Mannan, S. White Paper- Challenges in Implementing
Inherent Safety Principles in New and Existing Chemical Processes, Texas A&M University, 2002.
14.11 New Jersey Administrative Code, Toxic Catastrophe
Prevention Act (TCPA) Regulations , N.J.A.C. 7:31. |
Figure 15.7: Nitrogen supply line routed over the vessel manway. The
nitrogen must be disconnected to open the manway.
409 |
EVIDEN CE ANALYSIS & CAUSAL FACTOR DETERM IN ATION 191
9.5.2 Computational Modeling
Numerical models may be used to eval uate hypotheses in a variety of
problems. Numerical modeling refers to computer modelin g that involves
time-stepping to simulate the behavior of a system. Types of numerical
models that may be used in an incident investigation include:
Finite element analysis (FEA) – calculation of stresses, motion,
deformation and other properties of mechanical or structural
components subjected to forc es during an incident.
Computational fluid dynamics (CFD) – analysis of fluid flow,
including the thermodynamic conditions of the fluids
Fire and explosion CFD – these mo dels incorporate combustion and
explosion simulations.
Process simulation –process cond itions are calculated using models
of process equipment such as distillation towers, heat exchangers,
etc. The simulation can be stea dy state or dynami c (time varying).
Numerical models have become quite so phisticated. A qualified analyst
is needed to properly use numerical mo dels. There are many non-physical
parameters in numerical models such as time step and mesh (grid) size that
can change the results of a simulation. The numerica l model selected for the
simulations should be suitable for the type of event and the range of input
parameters associated with the event. Numerical models should be used
with caution if the numerical model has not been validated for the input
conditions.
Numerical models are not first-prin ciples models. Many numerical
models contain approximations, tuning coefficients, and numerical methods so that the models run in a stable fash ion and produce reasonable results.
While numerical models can provide valuab le insights into the behavior of a
system during an incident, the results must be reviewed carefully to ensure
that they are reasonable. Comparison to empirical data and first principles
calculations can be useful to check the numerical model.
9.5.3 Reconstruction
In some major investigations, reconstruction of a piece of equipment or
system may be required to understan d failure patterns, the physical
relationships between the various items that are recovered, the functionality
of equipment, and equipment behavi or under certain conditions. A
dedicated area or warehouse space may be required to effectively |
FIRE AND EXPLOSION HAZARDS 77
CCPS Guidelines for Fire Protection in Chemical , Petrochemical, and Hydrocarbon Processing
Facilities . This book consolidates fire prevention and protection tools and resources in a single
document. (CCPS 2003)
CCPS Guidelines for Pressure Relief and Effluent Handling Systems, 2nd Edition . Providing in-
depth guidance on how to design and rate emer gency pressure relief systems, Guidelines for
Pressure Relief and Effluent Handling Systems in corporates the current best designs from the
Design Institute for Emergency Relief Systems (D IERS) as well as American Petroleum Institute
(API) standards. Presenting a methodology that helps properly size all the components in a
pressure relief system, the book includes soft ware; the CCFlow suite of design tools and the
new SuperChems™ for DIERS Lite software, making this an essential resource for engineers
designing chemical plants, refineries , and similar facilities. (CCPS 1998)
CCPS Guidelines for Safe Handling of Powders and Bulk Solids . Powders and bulk solids are
handled widely in the chemical, pharmaceutical , agriculture, smelting, and other industries.
They present unique fire, explosion, and toxici ty hazards. Substances which are practically
inert in consolidated form may become quit e hazardous when converted to powders and
granules. This book discusses the types of haza rds that can occur in a wide range of process
equipment and with a wide range of substanc es and presents measures to address these
hazards. (CCPS 2004)
CCPS Guidelines for Vapor Cloud Ex plosion, Pressure Vessel Bu rst, BLEVE and Flash Fire
Hazards, 2nd Edition . This guide provides an overview of methods for estimating the
characteristics of vapor cloud explosions, fl ash fires, and boiling-liquid-expanding-vapor
explosions (BLEVEs) for practicing engineers. It has been updated to include advanced
modeling technology, especially with respec t to vapor cloud modeling and the use of
computational fluid dynamics. The text also reviews past experimental and theoretical
research and methods to estimate consequenc es. Heavily illustrated with photos, charts,
tables, and diagrams, this manual is an essentia l tool for safety, insurance, regulatory, and
engineering students and professionals. (CCPS 2010)
Crowl and Louvar, Chemical Process Safety Fundam entals with Applications 4th Edition, This
book provides a compilation of many process safety topics. It links quantitative academic
concepts to industrial process safety. (Crowl 2019)
Crowl , Understanding Explosions . This book describes the fundamentals and types of
explosions as well as methods to pr event their occurrence. (Crowl 2003)
FM Global property loss prevention data sheets . FM Global is an insurance company that
has used its loss experience to generate data sheets, which are essentially standards, on
variety of topics. These data sheets are intended to reduce the chance of property damage.
Topics of interest include industrial boilers, gas turbines, and extinguishing systems. (FMG)
|
APPLICATION OF PROCESS SAFETY TO OFFSHORE PRODUCTION 109
Figure 6-2. Example of offshore layout
(Piper Alpha, McLeod and Richardson, 2018)
reinjected if there is no export facility . Some gas treatment may occur onshore to
simplify offshore processing. Similarly, oil is sometimes only partially stabilized
with final stabilization carried out onshore.
One environmental operating factor of note for deepwater is cold sea
temperatures near the wellhead, often around 4°C, even in warm climates. This can
cause paraffins, waxy components, or hydr ates to form as solids thereby creating
plugging issues in the export line and risers. A process of flow assurance is used to
keep the oil flowing. This may include injecting solvents or other inhibitors, or some
combination of heated or insulated lines.
Historical Incidents
There have been several significant offshore incidents, and these have had a
substantial impact on facility designs, safety and environmental management, and
operational discipline as the industry responded. Two early events in the US affected
designs and led to the stoppage of well construction off the coast of California. These
were the Santa Barbara well control incident in 1969 and the Bay Marchand well
control incident off Louisiana in 1970. Deta ils of these events are readily available
online. The Santa Barbara incident created a large oil spill zone affecting coastal
locations and wildlife. There was a serious public outcry and the incident became a
trigger for subsequent major environmental legislation. The Bay Marchand incident
in 1970 caused four fatalities, lasted for three months, and created a large oil slick.
In the North Sea, the Piper Alpha disaster in 1988 led to the far-reaching Cullen
inquiry (available on the UK HSE website) and many changes to safety measures
for offshore installations. The P-36 disaster in 2001 off Brazil highlighted additional
safety issues and the need for operations excellence around maintenance isolation.
The Piper Alpha and P-36 incidents are discussed later in this chapter.
|
2.4 Ensure Open and Frank Communications |37
as body language, vocal inflections, and facial expressions can
either reinforce or contradict the m essage (Ref 2.12) and help
confirm that the message was received as intended. This process
is summ arized in figure 2.2.
Figure 2.2 Anatom y of a Comm unication
The link to trust should be clear. Trust helps the receiver of the
m essage tune-in to the deliverer and provides the environment
that encourages the receiver to provide feedback. However, the
non-verbal cues of the deliverer may be even more im portant. If
a leader does not in their heart believe in process safety, their
non-verbal cues will contradict anything positive they say.
Com munications should of course lay out a logical argument, be
unam biguous, and use a speaking or writing style that is easily
understandable to the audience.
Com munication should be accurate. Accurate
com munications avoid confusion, wasted time, and incorrect
decision m aking, and are m ost easily understood. Ensuring that
all comm unications are accurate is almost a hopeless task, given
norm al hum an error. Occasionally, people will communicate
information they believe to be correct, but later turns out not to
be. If a culture of open com m unication has been established, the
incorrect inform ation can be corrected through open and frank Non-
verbal
cuesMessage
sent
Message
received
Message
believedMessage
confirmed |
APPLICATION OF PROCESS SAFETY TO OFFSHORE PRODUCTION 123
Additional factors to consider include ignition probabilities, personnel presence
factors, escalation outcomes, etc. The ri sk contributions from all scenarios are
summed and a location specific individual risk (LSIR) developed for all modules,
as well as a combined group risk metric, th e Potential Loss of Life (PLL). LSIR and
PLL results are probed to identify the most hazardous locations and the major
contributor events so that risk reduction measures can be targeted. Moreover, these
measures are evaluated for the risk redu ction achieved. Some companies use the
ALARP principle for this determination. In simple terms, ALARP tests whether the
cost of a measure in terms of time, money, and effectiveness is commensurate with
the risk reduction achieved.
Software is available for performing an offshore QRA. A simple approximate
spreadsheet method widely used for the UK Sector of the North Sea was developed
by Spouge (1999). More detailed software tools are available which carry through
the calculations and escalation potentials w ithout the simplifications of spreadsheet
methods (e.g., SAFETI Offshore described in Pickles and Bain, 2015). BP has
developed its own screening offshore QRA tool, OMAR. These software tools
generally do not include CFD directly, due to the computational burden, but
summary CFD results can sometimes be imported. The tool used should be
commensurate with the type of decision, data available in the project stage, and the
type of result and precision required.
Fault and Event Tree Analysis
Fault trees are a form of risk analysis and are part of the RBPS element Hazard
Identification and Risk Analysis . Fault trees are useful to establish causal links in an
event sequence and are most commonly app lied offshore to understand BOP failure
modes and overall reliability.
The method is described in CCPS CP QRA (1999) and the Hazard Evaluation
Procedures Guide (CCPS, 2008a), and in more detail in NASA (2002). Fault trees
start with initiating or base events and show how a combination of events or barrier
failures allow the initiating event to progress upwards to the next step, and ultimately
to the top event (usually a loss of containment or system failure). Event
combinations are achieved using AND or OR gate logic. AND logic requires both
inputs to be true to progress upwards, while an OR gate allows the progression if
any one event is true. An example is shown in Figure 6-4. In this figure, boxes are
events, and boxes with circles below are base events. AND gates are represented as
an arch with a flat bottom and OR gates as an arrowhead shape.
Fault trees help identify common mode fa ilures, where a single failure disables
multiple barriers and greatly increases th e risk in a manner that can be hard to
visualize without this analysis. Another FTA application, termed minimal cut set
analysis, helps identify the fewest number of failures that lead to the undesired top
event occurring. Single event cut sets are very serious as a single failure results in
the top event. |
Heat Transfer Units
213
exchangers are expensive pieces of equipment. When
a heat exchanger is out of operation, it generally impacts the operation.
11.12 Fired Heaters and Furnaces
Furnaces are the least favorite heat transfer units for dif -
ferent reasons; they are expensive and they have a bunch of safety issues to be addressed. They may have names other than furnace; when the process fluid is water, they are named boilers or steam generators.
The main component of every furnace is the heating
system. There are mainly two types of furnaces: fired heaters where the heating system is fire, and electric heaters where the heating system is an electrical element. Electrical furnaces are expensive from a capital cost and operating cost viewpoint and are not used unless a very high temperature is needed. Therefore our discussion is limited to fire heaters.
Fire heaters use fire to increase the temperature of pro-
cess fluids. Fire is generated in burners where fuels and oxidant are brought together. The fuel could be anything from oil, gas, or coal to some specific sludge. The oxidant is mainly air and sometimes other oxidants like oxygen.
The result of burning fuel is flue gas. There are some
furnaces where their source of energy is not burning fuel but electrical energy.
Therefore, a furnace has at least three systems that
should be taken care of during the P&ID development: the process fluid side, the firing side, and the flue gas side (Figure 11.17).
11.12.1
Pr
ocess Fluid Side
As it was mentioned, for efficient heating, the fluid
should be exposed to heat with narrow thickness.
However, there are some cases where a huge stream
needs to be heated in a furnace. In such cases the stream is split into several narrow bore pipes before entering the furnace. This multi‐pass arrangement is needed to make sure the coils inside the fired heater have a small enough diameter (generally less than 4–6
in.) t
o make sure heat is
absorbed effectively by the process fluid. Therefore each fired heater may warm up a stream, not through a single coil, but rather through several narrow bore coils of gen-erally less than 4–6
in. Eac
h coil circuit in a furnace is
named one “pass. ” We may have a furnace with 2, 4, 6, etc. passes. Therefore, a large‐bore pipe for a process fluid is split into 2–6 (or more) small‐bore pipes outside the fired heater, and then all of them enter the fired heater chamber. These multiple passes will merge together again on the other side of the fired heater after heating the process fluid.
One responsibility of the P&ID developer is to distrib-
ute the stream of interest for heating amongst several coils evenly. This can be done by a control system or by symmetrical piping. These solutions are briefly discussed in Chapter 6 and the control system is discussed in more depth in Chapter 15.
Furnaces are designed and built to handle large
thermal expansion of its elements because of large temperature changes. However furnaces cannot handle the thermal shock. Introducing cold process fluid into the furnace coils may cause thermal shock and initia-tion of cracks and rupture in the furnace coils. Because of that when the temperature difference between the cold process fluid and the furnace target temperature is more than 100
°–150 °C, pr
eheating may be needed.
This preheating could be necessary to bring the cold process fluid to a temperature closer to the furnace temperature. The preheating is generally done by a heat exchanger and more commonly this heat exchanger uses the hot process fluid exiting the furnace to heat up the cold process fluid entering the furnace. Sometimes preheating is done through heat exchange from the hot flue gas of the furnace.
11.12.2
Flue G
as Side
Flue gas is generated inside the furnace and directed to
the atmosphere through the stack.A fired heater is a heat transfer unit that uses fire to heat
up the temperature of a process fluid.
If a fired heater is intended to be used to change the
process fluid physically or chemically to other fluids by heating up, it can be named a furnace.
Air FuelFiring SystemProcess fluidFlue Gas
Figure 11.17 Thr ee fluid streams within a fired heater. |
RISK ASSESSMENT 331
Tools
Risk analysis requires failure frequency data. Frequency analysis and risk
analysis can be conducted manually or using simple spreadsheets; however,
they are greatly facilitated using software packages. Resources include the
following.
HSE Failure Rate and Event Data for use within Risk Assessments. The U.K. Health and
Safety Executive Chemicals, Explosives and Mi crobiological Hazardous Division 5, has an
established set of failure rates that have been in use for several years. This is available at
https://www.hse.gov.uk/landus eplanning/failure-rates.pdf. (HSE Failure Rate)
IOGP Risk assessment data directory - Process release frequencies. The International
Association of Oil and Gas Producers Report 434- 01 provides frequencies of releases from
process equipment. They are intended to be app lied to equipment on the topsides of offshore
installations and on onshore fa cilities handing hydrocarbons an d can be applied to onshore
facilities where hydrocarbons are processed. (IOGP 2019)
Offshore and Onshore REliability DAtabase . OREDA is a project organization with oil and
gas companies as members. It is a comprehensiv e databank of reliability data covering process
hardware, control systems, and electrical equi pment that has been collected on offshore &
onshore operations globally for 35 years. For non-members, access to selected data is
available at https://www.oreda.com (OREDA)
Guidelines for Process Equipment Reliability Data. This book contains failure rate data for
use in supporting QRAs. (CCPS 1989)
Fault Tree and Event Tree software. Many software packages are available on the internet
to create fault trees and event tr ees, including one from Isograph.
LOPA software. LOPA may be conducted using worksheets to document scenarios and
spreadsheets to calculate the values. Software tools are available such as LOPA Works from
Primatech and AEShield from AE Soluti ons. (Primatech and AE Solutions)
SAFETI. This is a QRA software package available from DNV GL. It can be used to conduct a
QRA of onshore process facilities as well as pipe line risks. SAFETI includes failure rate data
similar to the IOGP data directory. (DNV GL)
Other options. Other software options include integrated PHA/ HAZOP and LOPA studies
Summary
This chapter discusses risk analysis, from qu alitative to a fully quantitative. The methods
include the risk matrix, LOPA, and QRA. Each of these methods has its own attributes that
make it better suited for various needs. All of the methods share the purpose of assessing the
risk with the intent to reduce it to a tolerable level.
The risk matrix is easy to use and easy to und erstand. It involves the least amount of time
and resources in developing risk es timates but is a coarse approach.
LOPA can be used to analyze higher risk, or consequence, scenarios and is helpful in
identifying if additional layers of protection are required or if a scenario is over-protected and |
Piping and Instrumentation Diagram Development, First Edition. Moe Toghraei.
© 2019 John Wiley & Sons, Inc. Published 2019 by John Wiley & Sons, Inc.Companion website: www.wiley.com/go/Toghraei_PID
Chapter No.: 1 Title Name: Toghraei c06.indd
Comp. by: ISAKIAMMAL Date: 25 Feb 2019 Time: 12:19:32 PM Stage: Proof WorkFlow: <WORKFLOW> Page Number: 71
71
6.1 Fluid Conductors: Pipes,
T
ubes, and Ducts
In process industries, to transfer fluid from point A to
point B (e.g. equipment A to equipment B), a fluid
conduc
tor is needed. Pipe is the most common type of
fluid conductor as it can transfer liquid, vapor, gas, or flowable solid. The fabrication process of pipes is differ -
ent from tubes. For a small fluid flow, tubes are used, while for low‐pressure gases or air, ducts. The other fluid conductors are channels, chutes, and so on (Table 6.1).
It is not difficult to transfer fluids, liquid, gases, and
vapor, but flowable solids is. The word flowable is used for a noncontinuous version of solids. These solids could be in the form of granules, pellets, flake, powder, beads, chunk, and so on. They could be sugar or tomato in food industries or ore in mineral processing. Usually the smaller the solid particles, the easier they are to trans -
port. To transfer a stream of iron ore in chunk form, we may need a bucket elevator rather than piping. Chutes are used for vertical or downward transfer of solids and semisolid materials. To summarize, transferring solids are generally done by equipment rather than by a simple conductor.
In this chapter, the main focus is on pipes.
6.2 Pipe Identifiers
As discussed in Chapter 4, the identifiers of pipes on P&IDs are pipe symbols, pipe tags, and stream names on the off‐page connectors.
6.2.1
Pipe S
ymbol
The symbol for pipe is a line (Table 6.2). On P&IDs it is
preferable to show pipes as vertical or horizontal lines rather than oblique. There are two features related to lines: their thickness and their arrowhead. On P&IDs, the thickness of lines gives more information about the pipe. A thick line represents a primary pipe, and a thin line means a nonprimary pipe. The decision whether a pipe is primary depends on the purpose of the plant. Generally, primary pipes are used in the main feed and product of the plant, while nonprimary pipes are for other streams. It is important to know that the thickness of the line does not necessarily specify the diameter of the pipe. A thin line could be a 2‐inch pipe, a 6‐inch pipe, or even a 20‐inch pipe. However, primary pipes do gen-erally tend to be large bore pipes.
The other feature of line for pipe symbols is the arrow -
head. The arrowhead shows the direction of flow in pipes during the routine operation. However, the question is, in which situations should the arrowhead be shown? The general rule is that the arrowhead should be shown in two cases:
1)
Where t
here is a change in the direction of the line
(Figure 6.1).
2) In the inle
t of equipment (Figure 6.2).
It is important to consider we generally do not put
an arrowhead on the inlet of valves or instruments
(Figure 6.3). Some companies also do not put arrow -
heads on short loops around process or instrument items (Figure 6.4).
There are cases in which a pipe has different flow
directions during different phases of routine operation. Because the arrowhead shows the direction of flow dur -
ing routine operation, normal operation could be one stage of routine operation. For example, a sand filter that works in a semicontinuous mode may have different operation phases: filtration, backwashing, and retention. Then there are pipes that carry flows in a different direc -
tion in different phases of operation. Thus bidirectional lines may exist (Figure 6.5).
6.2.2
Pipe T
ag
The second pipe identifier is a pipe tag. Pipe tags should
be assigned to each piece of pipe and shown on P&IDs. The anatomy of pipe tag is specified by the project 6
Pipes |
74 | 2 Core Principles of Process Safety
2.11 Morris, W., et. al., The American Heritage Dictionary of the English
Language, New College Edition , Houghton Mifflin Co, 1976.
2.12 CCPS, Recognizing Catastrophic Incident Warning Signs in the Process
Industries, American Institute of Chemical Engineers, New York,
2012.
2.13 Mathis, T., Galloway, S., STEPS to4Safety Culture ExcellenceSM, Wiley,
2013.
2.14 Canadian National Energy B oard (CNEB), Advancing Safety in The Oil
and Gas Industry - Statement on Safety Culture , 2012.
2.15 National Aeronautics and Space Administration, Columbia Accident
Investigation Board Report , Washington, DC, August 2003.
2.16 J ones, D., Kadri, S., Nurturing a Strong Process Safety Culture , Process
Safety Progress, Vol. 25, No. 1, American Institute of Chemical
Engineers, 2006.
2.17 Rogers, W.P. et al., Report of the Presidential Commission on the
Space Shuttle Challenger Accident, Washington, DC , J une 6, 1986.
2.18 UK HSE Health and Safety Laboratory, Safety Culture: A Review of the
Literature , HSL/2002/25, 2002.
2.19 Throness, B , Keeping the Memory Alive, Preventing Memory Loss That
Contributes to Process Safety Events , Proceedings of the Global
Congress on Process Safety, 2013.
2.20 Murphy, J., Conner, J., Black Swans, White Swans, and 50 Shades of
Grey: Remember ing the Lessons Learned from Catastrophic Process
Safety Incidents , Process Safety Progress, American Institute of
Chemical Engineers, 2014.
2.21 American Petroleum Institute, Management of Hazards Associated
with Location of Process Plant Buildings , API RP-752, 1st Ed., 1995.
2.22 CCPS, Guidelines for Evaluating Process-Plant Buildings for External
Fires and Explosions, American Institute of Chemical Engineers ,
1996.
2.23 American Petroleum Institute, Management of Hazards Associated
with Location of Process Plant Permanent Buildings , API RP-752, 2nd
Ed., 2009.
2.24 American Petroleum Institute, Management of Hazards Associated
with Location of Process Plant Portable Buildings , API RP-753, 1st Ed.,
2012. |
12 Guidelines for Revalidating a Process Hazard Analysis
Reasons why today's understanding of the risks associated with the process
might differ from the understanding that existed at the time of the prior PHA
include:
• Process changes (including decommissioning of equipment or
application of inherently safer design principles) have introduced
new hazards or affected existing hazards
• Changes in on-site or off-site facility occupancy or siting have
changed the at-risk populations
• External factors, such as the potential for flooding, have changed
• New knowledge is available to better understand the hazards
• Incidents (including events with actual losses or near-miss
incidents) in the process or in other similar processes have revealed
hazards or specific routes to hazardous events not previously
addressed
• Safeguards previously credited in the PHA have been removed,
compromised, rendered ineffective, or discredited
• New safeguards have been added
• Applicable codes, standards, or engineering practices have
changed
• Risk management systems have been implemented, changed, or
discontinued
• Retirements, transfers, or resignations have significantly changed
the overall skill levels of the staff
PHAs are an essential element of an y process safety management (PSM)
system, and they must provide current, accurate information to serve their
purpose successfully. Five primary reason s to revalidate PHAs on a periodic
basis are:
1. Evaluate the cumulative effect of changes in the process,
equipment, or personnel
2. Correct gaps and deficiencies in the prior PHA
3. Incorporate new knowledge and operating experience
4. Comply with frequency requirements per regulations or company
policy
5. Comply with content requirements per regulations or company
policy
|
314
Table 12.3: Examples of Potential Accident Consequence Analysis as
a Measure of Inherent Safety
Inherent
Strategy Description Consequence
Minimize Chlorine transfer line rupture —
distance to 20 ppm atmospheric
concentration
D atmospheric stability
3.4 mph wind speed 2-inch line - 3.4 miles To
1-inch line - 1.2 miles
Explosion overpressure 100 feet from a reactor 3000-gallon batch reactor - 1.1 psig
To
50-gallon continuous reactor -
0.3 psig
Substitute Concentration in the atmosphere
500 feet downwind from a large spill from a storage tank
D atmospheric stability
3.4 mph wind speed Methanol - 1000 ppm
To
Butanol - 130 ppm
Moderate Distance to 500 ppm atmospheric concentration for a large spill of
Monomethylamine
D atmospheric stability, 3.4 mph
wind speed Stored at 10ºC - 1.2 miles
To
Stored at 3ºC — 0.7 miles
To
Stored at -6ºC — 0.4 miles
Concentration in the atmosphere
500 feet downwind from a large
spill from a storage tank containing
methyl acrylate
D atmospheric stability, 3.4 mph
wind speed 2500 sq. ft. concrete containment dike —
1100 ppm
To
100 sq. ft. concrete
containment pit —
830 ppm |
80 INVESTIGATING PROCESS SAFETY INCIDENTS
management responds promptly and effect ively to reports of incidents, the
workforce will realize that their concerns are taken seriously. This will
encourage continued incident reporting an d drive a positive safety culture.
The supervisor, when informed of an incident, would customarily be
responsible for initiating further action to alert management, investigate the
incident, and take required action. Firs t notification may also need to follow
company protocol to report details of the incident to specific individuals or
organizations internally or externally including regulatory agencies (see
Section 5.3 below).
Not only does incident reporting allo w management to initiate remedial
measures, it can help to instill a sense of vulnerability to keep the workforce alert to potential hazards and their proper management. Furthermore, the
lessons learned from incidents can be shared more widely within the
company, and, if appropriate, externally.
All incidents and near-misses should be entered into the company’s
incident reporting system (such as a database or log). As a minimum, the
database or log may record the type of incident, date/tim e, description, and
circumstances of the incident. Additi onal information could include the
stakeholders notified and the incident cla ssification. Other fields may be left
blank at this time if the information is not yet available. Examples of types
of incidents that may be recor ded include, but are not limited to:
Injury (e.g., first aid, non- disabling, disabling, etc.)
Fatality
Occupational illness
Release of hazardous material from primary containment, i.e., vapor
release, liquid spill, solid release (including dust)
Fire
Explosion
Process upset (e.g. flaring, off-spec product/effluent, etc.)
Property damage (at or abov e a certain cost level)
Environmental damage
Security (trespass, theft, bomb-threat, etc.)
Community complaint (odor, noise, etc.)
Near-miss
Challenge to a safety system (e.g. re lief valve discharge or safety trip)
|
APPLICATION OF PROCESS SAFETY TO OFFSHORE PRODUCTION 111
The 100 Largest Losses in the Hydrocarbon Industry (Marsh, 2020) lists several
major process safety events in offshore production or well construction that have led
to large insurance claims. Many also involved multiple fatalities. Upstream incidents
(all offshore) account for 23 of the total list of 100 major (financial) losses applying
also to refineries, petrochemical plants, gas processing and terminals.
While there have been serious incidents offshore, industry data collection
initiatives focusing on process safety as well as occupational safety are providing
better transparency so that companies can better focus on process safety incidents –
“what gets measured, gets done”. The COS (2020) trend data for ten large companies
and six contractors in the Gulf of Mexico for 2018 shows some improvement but
also flat trends in some indicators. Specific Tier 1 and 2 process safety event
statistics are available from IOGP (2019a) with both categories showing some good
decline over the period 2011-2018 related to offshore activities.
6.2 OFFSHORE PRODUCTION FACILITIES: RISKS AND KEY
PROCESS SAFETY MEASURES
Process safety risks on offshore production facilities are typically due to the inherent
hazards of flammable reserv oir fluids or other chemicals onboard and the activities
carried out to support production. These include normal equipment operations and
asset integrity activities, mechanical lifting, support vessel operations, mechanical
degradation, etc. Improper management of these risks has the potential to cause a
process safety event.
There are important aspects of offshore operations that increase risks compared
to many onshore production facilities. These aspects include the following.
●The potential for fire or blast escalation given the limited footprint offshore
that requires processing units/areas be located adjacent to one another or
stacked vertically (as is apparent in Figure 6-2). The closer spacing creates
a greater potential for even a small pr ocess safety event to escalate to a
larger one by involving nearby process equipment. The facility design
should account for these conditions through the use of Hazard
Identification and Risk Analysis and consider emergency isolation or the
addition of fire or blast walls to minimize the likelihood of escalation. The
addition of walls can also trap hydrocarbon liquids and vapors, so the
installation of these walls requires car eful analysis and modeling. Weight
restrictions offshore can limit the number and scale of fire and blast walls.
●Personnel often live onboard near pr ocess equipment, increasing their
exposure to hazards an d associated risks.
●Mechanical lifts are inherently more hazardous offshore due to lifting from
different deck levels, supply vessels subject to sea conditions, and lift paths
potentially over process equipment on deck and production infrastructure
on the seabed. An advantage offshore, however, is that crane bases are
fixed structurally, and hence toppling events are less likely than onshore. |
Piping and Instrumentation Diagram Development
384
On the potable water distribution network a small sym-
bol of safety shower or eye washer is enough (Figure 18.3).
There would be another drawing based on the plot
plan to show the location of safety showers and eye
washers in a plant.
18.3 Dealing with Environment
A plant is located in an environment. Therefore it is exposed to the changes of the environment’s parameter.
There are plenty of parameters associated with environ-
ment including atmospheric pressure, ambient tempera-ture, relative humidity, precipitation, radiation, dust, etc.
First of all, for obvious reasons, it is tried to establish a
plant in the areas that have the least environment param-eter harshness and swing.
If a plant, or a portion of a plant, should be established
in a “bad” area and it cannot tolerate the environment parameter swinging, they can be “isolated” by locating them inside of a building. Indoor plants are very expen-sive and they are not always justifiable. However it is not rare to see fully indoor plants of food or pharmaceutical industries.
Table 18.3 summarizes the impact of several environ-
mental parameters on process plants.It can be seen that if the temperature swings are high
this affects every single item in a plant. Heat transfer from the process to surroundings (or reverse) is not desirable and should be minimized. This is because heat transfer means a waste of energy. Therefore “isolating” the plant from the ambient temperature needs more thought.
To study the effect of temperature change on a process
plant it would be better to study it in each stage of opera-tion. The operations of a process plant can be normal and nonroutine operations. Nonroutine operations could be reduced capacity, upset, shutdown, and start‐up operations.
Isolating a plant from environment in upset operation
is no different from any other nonroutine operations and won’t be discussed separately.
The requirements of start‐up operation are generally
considered in shutdown operation to make sure that after each shutdown, the unit or plant can be started up without any hassle.
In this context shutdown could be divided into short,
medium, or long term shutdown. The different require-ments of plants to be thermally isolated from the environ-ment are shown in Table 18.4.
Table 18.4 shows that thermal insulation and tracing
are used to minimize “communication” between the plant and its surroundings. In the next section we will explain the two main types of “thermal isolation”: heat conservation and winterization.
18.3.1
Arr
angements for Maintaining
the Temperature of the Process
One type of insulation is the type where the goal is keep-
ing fluids at the temperature they are intended to be.
Eyewasher Safety shower Combined safety shower
and eyewasher
Figure 18.3 P&ID symbols for personnel emer gency washers.
Table 18.3 Impac
t of environmental parameters on process plants.
Environmental
parameterMagnitude of swing Process impact Stack up strategy
Atmospheric
pressureSmall swing Units using ambient air
(e.g. blowers)Considering it in process design
Ambient temperatureLarge swing All equipment: heating up,
aerial coolersConsidering it in process design (summer/winter designs)
Relative humidity Large swing Units using ambient air
(e.g. blowers, burners)Using dehumidifier or considering it in process design and/or P&ID development
Precipitation Non‐marginal Open‐top equipment (e.g.
open‐top tanks and basins)Considering it in process design (e.g. closed‐top tanks)
Sun radiation Non‐marginal but
effectiveAll equipment: heating up Considering it in process design (PSV for blocked fluids
under sun radiation)
Dust Depends on the locationAmbient air using units, open‐top equipmentUsing filter |
Executive Summary | xxv
Recalling Experiences and Applied Learning (REAL) Model for learning
from incidents is introduced.
• Chapter 6 describes the REAL Model in detail, providing a roadmap for
learning from incidents.
• Chapter 7 discusses how to use the full range of learning styles in your
efforts to keep lessons learned fresh and to prevent erosion of knowledge
and normalization of deviance.
• Chapter 8 describes landmark incidents that contributed significantly to
our understanding of process safety. In your efforts to drive process
safety improvement, it is important to ensure that the lessons learned
from these incidents have been well implemented and maintained
throughout your organization.
• Chapters 9–14 present a set of fictional but realistic scenarios that show
how you might apply the REAL Model in your own workplace to translate
the findings from incidents to corporate lessons learned.
• Chapter 15 summarizes the concepts described in the book and challenges
the reader to drive continuous improvement.
One noteworthy feature of this book is the Index of Publicly Evaluated
Incidents, presented in the Appendix. Use this index to identify specific
incidents with relevant findings that can help you advance your corporate
improvement goals. Incidents described in the book that are included in the
index are flagged with a text box, as shown at right. You
can consult the Index of Publicly Evaluated Incidents to
find a link to the incident report, along with an
assessment (by a book committee member) of the most
important findings.
Thank you for reading this book. There is really no business priority that
comes ahead of protecting workers, facilities, communities, and the
environment. Process safety depends on all of us performing our roles with
professionalism and competence, using the best knowledge available. Just as
excellence in process safety cannot be achieved by one person alone, process
safety knowledge cannot be held only by company experts. When it comes to
process safety knowledge, the only thing that matters is what we know, and
do, and manage collectively.See Appendix
index entry XYZ |
Part 7: Working with contractors and managing change Human Factors Handbook For Process Plant Operations: Improving Process Safety and System
Performance CCPS.
© 2022 CCPS. Published 2022 The American Institute of Chemical Engineers. |
316
H Record of Foreseeable Hazards
I ISHE Performance Indices
I.1 Fire and explosion hazard index
I.2 Acute toxic hazard index
I.3 Inherent health hazard index
I.4 Acute environmental hazard index
I.5 Transport hazard index
I.6 Gaseous/atmospheric emissions environmental index
I.7 Aqueous emissions environmental index
I.8 Solid emissions environmental index
I.9 Energy consumption index
I.10 Reaction hazard index
I.11 Process complexity index
J Multi-Attribute ISHE Comparative Evaluation
K Rapid ISHE Screening Method
L Chemical Reaction Reactivity - Stability Evaluation
M Process SHE Analysis - Process Hazards Analysis, Ranking
Method
N Equipment Inventory Functional Analysis Method
O Equipment Simplification Guide
P Hazards Range Assessment for Gaseous Releases
Q Siting and Plant Layout Assessment |
46 INVESTIGATING PROCESS SAFETY INCIDENTS
3. “W hy” it happened?
A component for systematically investigating the management
and organizational factors that a llowed the critical events and
conditions to occur (root causes identification).
Finally, in selecting an appropriate incident investigation methodology,
consider whether the method facilitates the identification of management
system and organizational inadequacies and oversights. The methodology
should specifically identify factors that influence and control an
organization’s risk management practices and procedures.
3.4.1 Methodologies Used by CCPS Members
The Center for Chemical Process Safe ty (CCPS) conducted a survey of its
membership and other chemical processing companies in preparation for
the second edition of this book in 2003. Based on the responses, some
general observations can be made about incident investigations:
• Companies reported using an average of two or three different
methodologies for both major an d minor incidents. The surveyed
companies used both public domain and proprietary tools and
methodologies.
• The most popular methodologies use different combinations of
the tools described in Table 3.1 as well as proprietary tools.
The methodologies used today provi de improved results over simplified
techniques such as in formal, one-on-one interviewing. Most current
methodologies have adopted a battery of tools for application at particular
stages of the investigation process. As a minimum, a tool representing the
incident sequence is used prior to iden tifying causal factors (also known as
critical factors), to which root cause analysis is subsequently applied.
In general, the companie s surveyed use one of two approaches to
determine root causes. The first involv es timeline construc tion followed by
logic tree development. The second involves timeline construction,
identification of causal factors, fo llowed by the use of predefined trees or
checklists. |
REACTIVE CHEMICAL HAZARDS 103
Exercises
List 3 RBPS elements evident in the T-2 Laboratories reactive chemicals explosion
incident summarized at the beginning of this chapter. Describe their shortcomings as
related to this accident.
Considering the T-2 Laboratories reactive chemicals explosion incident, what actions
could have been taken to reduce the risk of this incident?
The U.S. Chemical Safety and Hazards Investigation Board conducted an investigation of
reactive chemical incidents. What were the 3 reactive hazards types they identified?
Provide an example of each.
What can result if an organic peroxide is mixed with a hydrazine?
Can a metallic alloy pipe be used to safely transport sulfuric acid?
Using the CRW worksheet, develop an Inter-Reac tivity Chart for the following chemicals:
Styrene, Ethylbenzene, Aqueous Sodium Hydroxide, Aqueous Hydrogen Chloride.
Submit a “screen shot” of your chart. What are the binary combinations of concern and
why?
Describe the chemical reactivity that occurr ed in the Napp Technologies explosion in
Lodi, New Jersey in 1995. What steps could have been taken to prevent this event?
References
ABET 2015, Criteria for accrediting engine ering programs, Accreditation Board for
Engineering and Technology, Baltimore, MD,
https://www.abet.org/accreditatio n/accreditation-criteria/criteri a-for-accrediting-engineering-
programs-2021-2022/
Nouryon, “Safety of Organic Peroxides”,
https://www.nouryon.com/globalassets/inriv er/resources/brochure-safety-of-organic-
peroxides-lowres-en_us.pdf
Barton & Rogers 1997, Chemical Reaction Hazards: A Guide to Safety , Institute of Chemical
Engineers, U.K.
Bretherick & Urban 2017, Handbook of Reactive Chemical Hazards, 8th Edition , Elsevier,
Netherlands.
CCPS, “Chemical Reactivity Worksheet 4.0”, https://www.aiche.org/ccps/resources/chemical-
reactivity-worksheet .
CCPS Glossary, “CCPS Process Safety Glossary ”, Center for Chemical Process Safety,
https://www.aiche.org/ccps/resources/glossary .
CCPS 1995 a, Guidelines for Chemical Reactivity Eval uation and Application to Process Design ,
Center for Chemical Process Safety , John Wiley & Sons, Hoboken, N.J.
CCPS 1995 b, Guidelines for Safe Storage and Handling of Reactive Materials , Center for
Chemical Process Safety, John Wiley & Sons, Hoboken, N.J.
CCPS 1998. Guidelines for Safe Warehousing of Chemicals , Center for Chemical Process Safety,
John Wiley & Sons, Hoboken, N.J. |
Heat Transfer Units
211
need to provide air and drain valves on the connecting
pipes of an aerial cooler. If this is the decision it is appro-priate to provide one set of vent and drain valves on each header of a tube bundle. This means that each tube bundle has two vent valves on the top of each header and two drain valves on the bottom of each header.
An aerial cooler can also be equipped with isolation
valves.
As the tube bundle of aerial coolers is considered as an
enclosure, it should be protected from over‐pressuriza-tion by installing PSDs.
11.9.2
Dealing with Ex
treme Temperatures
As aerial coolers use ambient air generally without any
pre‐treatment, they thus should be robust. There are, however, some extreme cases that should be considered in the design of aerial coolers and/or in the P&ID devel-opment of them. The main important issue is extreme ambient air temperatures.
The ambient temperature change depends on seasons
and going from night to day or vice versa. The control system around an aerial cooler tries to compensate the ambient air temperature so that there is a fairly fixed temperature of process fluid exiting the aerial cooler. Such a control system will be discussed in Chapter 13. Here, however, we talk about methods to deal with “extreme” ambient temperatures; how handle the cases where the ambient air is extremely high or extremely low, which may causes frosting if there is enough vapor in the air. If such conditions are forecastable in the area, it should be considered in the design.
●Dealing with extremely low temperatures: extremely low temperatures (say less than 15–20
°C) c
ould be
problematic, mainly where there is a chance of frosting on the aerial cooler tubes. There are mainly two ways to handle such conditions: using a steam coil and warm air recirculation.
In the steam coil solution, a bundle of steam coils is
superimposed on the aerial cooler bundles. By doing this, the air moving through the aerial cooler bundle has to go through the steam coils at the beginning. This warms up the cooling air and prevents frosting.
In the warm air recirculation solution, instead of
“virgin” ambient air, the warmed ambient air is used for cooling purposes in the aerial cooler.
There could be two sources of warmed up air.
In “internal air recirculation” the source of warmed up air is the already used ambient air after some cooling effect on the aerial cooler. In “external air recircula-tion” the source of warmed up air is partially from the “air conditioned air” of the adjacent building. Obviously the external air recirculation is available only for aerial coolers in the vicinity of air‐conditioned buildings.
Implementing the concept of air recirculation needs
putting several louvered walls around the aerial cooler in a way that it looks like the aerial collar is placed inside of a “box” with perforated walls.
One example of a boxed aerial cooler is shown in
Figure 11.15.
A combination of steam coil and recirculation air
can also be used.
In the P&ID, the louvers should be shown and
tagged in a similar way to control valves.
●Dealing with extremely high temperatures: I have seen in extremely hot days of summer that water hoses have been placed on the aerial cooler bundles! However a factory‐built system to deal with extremely high temperature is a “humidifier aerial cooler. ”
11.9.3
Aerial C
ooler Arrangement
Aerial coolers in series are generally avoided. This is because of difficulties in the symmetrical distribution and collection network between two aerial coolers in series considering the limited available space. However it is not difficult to avoid aerial coolers in series. By using several aerial coolers in parallel (which is very easy) and also using a multiple pass configuration, series aerial coolers can easily be avoided.
As aerial coolers have a modular structure, in the major -
ity of cases they are installed in a parallel arrangement.
Represent distribution header
Represent Collection header
Figure 11.14 P&ID detail of an aerial c ooler. |
KEY RELEVANCE TO PROCESS PLANT OPERATIONS 47
Example Incident 3.5 – B uncefield Explosion, 2005 ( cont. )
Design: Other important considerations.
o Consider installing gas detect ors in bunds that have the
potential for large quantities of flammable materials to
be released.
o Consider improvements to managing the integrity of
overfill protection systems.
o Ensure that the site receiving the fuel (rather than the
transmitting location) has safe, ultimate control of tank
filling and is not reliant on a third party to safely
terminate or divert a transfer.
Process Monitoring and Control:
o Understand and define the ro le and responsibilities of
the control room operators (including automated
systems) in ensuring safe transfer processes. Improve
procedures for product movements and record keeping.
o Provide suitable information and system interfaces for
frontline staff to enable them to reliably detect,
diagnose, and respond to potential incidents. The HMI
should be designed so that the operator has multiple
ways of viewing a transfer status, improving situational
awareness that would make it easier to identify level
gauge failure.
One of the outcomes of the Buncefie ld incident was an update to the
API 2350 standard and the application and use of Automatic Overfill
Protection Systems (AOPS) for high-risk scenarios.
Some principles for the effective management of abnormal
situations are included in several elements of the CCPS Risk Based
Process Safety framework: Hazard Identification an d Risk Analysis
(HIRA); Operating Procedures, including safe operating limits, and
consequences of deviation from safe limits; Training and Performance
Assurance; Asset Integrity and Relia bility; Conduct of Operations and |
10.2 Seek Learnings | 133
Antônio agreed to check the public literature for similar cases and
anything else that might inform their path forward. He quickly identified
several cases involving hose leaks and failures:
Kawasaki, Kanagawa, Japan, 1989
A hose used to connect a fuel oil tank to a boiler
broke, leaking oil into the sea. The hose was
attached to the tank and boiler, hanging freely with no support. The
weight of the filled hose exceeded the tensile strength of the hose, leading
to failure.
Festus, MO, USA, 2002
A braided metal hose being used to unload chlorine
from a rail car ruptured, releasing more than 20,000
kg of chlorine. Three workers and 63 nearby residents sought medical
treatment. The hose was stamped Hastelloy C, the intended metallurgy.
However, the hose was found to be stainless steel.
Belle, WV, USA, 2010
A short segment of small diameter PTFE-lined
braided stainless-steel hose was being used to
transfer phosgene from cylinders to the process. The hose burst, spraying
a worker with phosgene. The worker died several hours later. The
phosgene hoses were supposed to be replaced monthly. However, the
hoses had not been replaced in nearly six months. Further investigation
revealed that when the asset integrity management system was updated,
the replacement notification had been inadvertently cancelled.
Atchison, KS, USA, 2018
Raw material sulfuric acid was being unloaded by a
supplier into a customer’s plant. The driver attached
the hose to the wrong connection and pumped sulfuric acid into the
sodium hypochlorite tank. This created a large release cloud of chlorine See Appendix
index entry J57
See Appendix
index entry C23
See Appendix
index entry C25
See Appendix
index entry C49 |
Piping and Instrumentation Diagram Development
318
There are two options for a control loop here: flow
loop or pressure loop, and as flow and pressure are
interrelated, both of them are theoretically acceptable. The decision of whether to use a flow control loop or a pressure control loop could be based on some other parameters. For example, pressure sensors generally have higher rangeability and if your flow surges a lot a pressure loop may be a better option.
The other case is when you are dealing with very high
discharge pressures. In such cases, high design pressure flow meters could be overly expensive and not justifiable. This may direct you to use a pressure control loop instead of a flow control loop.
●Speed control (VSD). This method can be used with all gas movers.
Figure 15.45 shows the control of a gas mover by VSD
when its drive is an electric motor.
Some companies use the symbol “SC” in their sche-
matics, meaning “speed control. ”
Some other companies use the acronym “VFD. ”Figure 15.46 shows the same VSD control on a steam
turbine‐driven gas mover. Here the device that works to change the speed of turbine shaft is a “governor. ” ●Suction throttling. Suction throttling can be used for all types of gas movers but is really only popular with centrifugal gas movers (Figure 15.47).
●Discharge throttling. Although this method can be used for all dynamic type gas movers, it is not economical. By adjusting the valve on the discharge side of a gas mover, we are partially expanding a gas that we had already com-pressed (and spent money to do so) through the gas mover (Figure 15.48). This is a waste of energy. However, this is not much of a factor when dealing with gas movers with small compression ratios, such as fans and blowers.FC
SC
SCPTPT
FTM
MFigure 15.45 Speed con trol of a motor‐driven gas
mover.
T
FTFC
Gov
Figure 15.46 Speed con trol of a turbine‐driven gas mover.
M
FTFCPT PC
MFigure 15.47 Suction‐thr ottling control of a gas
mover.
M
FTFC
PTPC
MFigure 15.48 Discharge‐thr ottling control of a
gas mover. |
13.2 Seek Learnings | 171
When an operator removed the wrong bolts, the bonnet came off, causing
a major isobutylene release and fire which resulted in four fatalities. This
valve design met the API standard when the plant was built but the
standard was updated in 1984 to require the gearbox to be separately
mounted. The standard allowed for legacied valves to be considered
compliant.
In his search for other relevant cases, Rakesh found many incidents in the
public domain where a current standard needed to be upgraded. Chana’s
standards all needed to be updated. He was sure he would also find many
incidents caused by not following standards. However, he could not find
another incident caused by failure of a legacied process component.
He decided to focus on the other factors in the Baton Rouge incident and
found case studies that represented each factor.
• Generic procedure (procedure didn’t state clearly how to safely remove the
gearbox):
Port Neal, IA, USA, 1994
A massive explosion resulted in 4 fatalities and
injured 18 when an ammonium nitrate
neutralization reactor exploded. Offsite releases of anhydrous ammonia
continued for the next 6 days. The reactor’s pH control system failed, but
operation continued with manual sampling. Later, the reactor was idled
due to a shortage of nitric acid feed. Because it was winter, operators
attempted to keep the reactor warm, using a sparge of high-pressure
steam instead of applying low pressure steam to the reactor jacket. The
pH drifted low, destabilizing the ammonium nitrate. Then the
temperature rose to the decomposition temperature and the ammonium
nitrate detonated. Operating procedures did not match the equipment in
the field, didn’t specify clearly what to do during winter shut down, or how
to control pH when the sensor was out of service.
• Error traps (routinely doing a task one way, but occasionally needing to do
it a different way):
Anonymous 2 (from chapter 10):
A driver was directed by an operator to connect
the hose for a shipment of sodium hydrosulfide
to the wrong connection. This material was rarely delivered, and the See Appendix
index entry S14
Process Safety
Beacon Mar. 2009 |
Piping and Instrumentation Diagram Development
236
The other, less common, way to protect non‐pressure
vessels are “seal locks. ” Seal locks are a U‐shaped pipes filled
with water (or a freeze resistance liquid) in it (Figure 12.34). The height of the liquid column works in a similar way to a “dead weight” in dead weight PRDs. The problem of such PRDs is the necessity of frequent inspection of the seal lock to make sure the liquid is in working conditions.
12.18 Merging PRDs
Merging PRDs can be discussed in two contexts: merg-ing two (or more) PRDs from different enclosures and putting a single PRD instead and merging two nozzles on a container and using a dual purpose device instead.
Let’s start with the first topic as it is a heated topic
because it is of a technical and legal nature. In a nutshell, the PRDs of two separate enclosures can be merged together and only one PRD is used if there is no “obsta-cle” between the two enclosures.
In Figure 12.35 a “story” is presented to explain this
concept. Let’s start the story from a purely technical viewpoint (not legal).
In Figure 12.35(a), a single enclosure exists and it has
its own PRD.
In Figure 12.35(b), the above single enclosure is
turned into an enclosure with two compartments. There is narrower enclosure between the two compartments. Table 12.14 Tw o types of pressure in enclosures.
Vacuum
(external pressure)Pressure(internal pressure)
Schematic
Needs Vacuum protection Pressure protection
Pressure vessels PSV, mandatory Vacuum breaker
●No needed if the container is
designed “FV”(full vacuum)
Atmospheric tanks Emergency Normal Vacuum relief valve
Emergency vents or PSVs PRV
Could be merged as PVRV or PVSV
Pressure ValveV acuum Valve
Figure 12.33 PVRV symbol .Figure 12.34 Seal lock . |
CONSEQUENCE ANALYSIS 285
Mathematical . The most common mathematical appr oach has been the box model (also
known as top-hat or slab model), which estimates overall features of the cloud such as mean
radius, mean height, and mean cloud temperature without calculating detailed features of the
cloud in any spatial dimension. Gaussian (P asquil-Gifford) models can be used but the
mechanisms are incorrect and thus models wi th dense gas mechanisms are preferred. The
other form of mathematical model is the more rigorous computationa l fluid dynamics (CFD)
approach that solves the complete three-dime nsional conservation equations. The CFD model
is typically used to predict the wind velocity fiel ds, with the results coupled to a more traditional
dense gas model to obtain the concentration pr ofiles (Lee 1995). An advantage of CFD models
is the ability to account for changes in te rrain, buildings, and other irregularities. A
disadvantage is they require substantial defini tion of the problem including a representation
of the release environment in three-dimensional space in order to start the CFD computation.
The method requires moderate computer resource s. CFD modeling is typically used to analyze
offshore installations where this level of detail is required close to the potential release source
and for forensic investigations where the analys t tries to link observed outcomes to several
possible release scenarios.
Physical . Physical (scale) models employing wind tunnels or water channels have been
used for dense gas dispersion simulation. An advantage is the ability to model specific
situations with obstructions or irregular terrai n. A disadvantage is the inability to achieve exact
similarity in all scales and re-creation of at mospheric stability and velocity distributions. The
use of scale models is not a common risk assess ment tool in consequence analysis although it
has been applied successfully for specific problems such as dispersion of LNG plumes resulting
from spills into containment.
|
14 Guidelines for Revalidating a Process Hazard Analysis
process safety regulations. New inform ation may result in new or different
conclusions or recommendations in the PHA revalidation report. Each
revalidation offers an opportunity to review past risk judgments in light of
current knowledge. In particular, information gleaned from incident
investigations may confirm the credibility of hazards/hazardous events that
seemed implausible to the prior PHA team.
Comply with Frequency Requirements. The schedule for revalidating a PHA is
often dictated by a maximum allowable time interval, as discussed in Section 1.6.
Comply with Content Requirements. The need to revalidate a PHA may be
driven by changes in the required content. Examples of this include the revision
of the company procedure or policy for th e conduct of PHAs, or applicability of
new RAGAGEPs to an existing process. In the United States, state requirements
may be more stringent than federal government requirements, and in the
European Union, member states may impo se more stringent requirements than
the current revision of the Seveso Directive. The revalidation team must ensure
the PHA meets current regulatory content requirements.
Note that an organization’s understan ding of its process risks can change
(increase or decrease) when a PHA is r evalidated. In those cases where perceived
risks exceed the organization’s current ri sk tolerance, the revalidation team may
propose risk reduction measures as discussed in Section 8.3.
1.5 PHA REVALIDATION CONCEPT
The goal of a revalidation is to verify that the PHA accurately reflects the current
hazards, consequences, and risk controls for the subject process and that it
meets current PHA requirements (e.g., ri sk judgments are in accordance with
current company policy and content conf orms to applicable regulations).
The terms “Update” and “Redo” are used consistently throughout this book
to describe two basic approaches to a revalidation:
Update . This is an incremental approach wh ere specific changes that have been
recorded and information that has been gained since the prior PHA are reviewed
to produce a revalidation report that documents revisions to the prior PHA. The
Update effort may be relatively modest for some processes (especially for
established, mature processes su bject to infrequent change).
Prior PHA + PHA Update = Revalidated PHA |
366 PROCESS SAFETY FOR ENGINEERS: AN INTRODUCTION
English speakers to forget that English is no t their colleague’s first language. Colloquialisms,
slang phrases, poor pronunciation, and speaki ng quickly can cause communication problems.
Air Traffic Control communications are a good example of excellent communications. Each
communication is acknowledged. They are clear and brief and use a standard glossary of
words and phrases.
Mental models. A mental model is a human’s understanding or mental image of a
situation, task, or environment. The same situat ion can look different to two people, depending
on what each senses and understands. In proc ess safety, making sure people have the same
mental model with respect to critical safety elements is important. Communicating well and
having access to the same data are important in sharing a common mental model. A sobering
example of this is the NASA Challenger incide nt where managers perceived the engineers as
raising unwarranted concerns and the engineers perceived the managers as disinterested. The
investigation concluded that NASAs safe ty system had been “silent” including
misrepresentations of criticality (of known safety problems) and lack of involvement in critical
discussions. (CCPS 2012)
Human Factors in the Process Workplace
The goal is to improve human performance as a means to reduce process safety incidents. As
humans are involved in all aspects of the work place, many opportunities arise to use the
human factors that influence individuals and work teams in the workplace. A few examples are
described here.
HAZOP Study. The HAZOP study was discussed in Section 12.3.4. Human factors appear in
two ways in a HAZOP. First is that they can be a potential cause of process deviations analyzed
in a HAZOP such as a valve being inadvertently left open or a safeguard where human response
to an alarm is required. Checklists are often used to address human factors in a HAZOP. Other
approaches include having a walk-through of the unit being reviewed in advance of the study
to help the HAZOP team develop a common mental model of the process. If it is a study of a
unit yet to be built, a 3D model review can be used to provide this common understanding.
During the study itself, particular focus on safeguards that are entirely human dependent
when considering if recommendatio ns are warranted is desirable.
The second way is that human factors impact the conduct of the HAZOP study. The HAZOP
is conducted by a team, thus the factors influe ncing both the individual and the work team
should be kept in mind. Understanding human factors can lead to a more effective study.
HAZOPs can be a significant time commitmen t. Consider the participants’ workloads
and set the schedule to minimize stress an d fatigue. Manage the time and stick to
the agenda, providing appropriate meeting breaks
Select a location to minimize distractio ns and allow people to focus attention.
Clearly state the scope and objective. With a shared mental model of the, the study
will be more effective and efficient.
|
PREPARING THE FINAL REPORT 299
incident,
• Describes the management systems th at should have prevented the
occurrence,
• Identifies factors that contributed to an escalation of the incident
consequences,
• Details the system root causes, and
• Provides management with suggested recommendations to
prevent or lessen recurrence an d/or associated consequences.
The report should include relevant information, stated factually and
accurately. If there is uncertainty ab out an event sequence or some other
aspect of an investigation, the uncertainty should also be conveyed in the
report tone and the choice of words used in the report should reflect the
attitude of preventing a similar inci dent rather than affixing blame.
13.4 SAM PLE REPORT FORM AT
The report format and content will depend on the needs of an organization (NFPA 921, 2017) and the complexity of the investigation (API RP-585, 2014).
Because there is no single universal re port that simultaneously satisfies all
needs of all organizations and potential readers and users, the sample format presented below provides a variety of content and detail.
Organizations can select
the format best suited to their needs. Table 13.1
includes a list of sections that may be included in an incident investigation
report. Most sample reports answer basic questions such as:
• What happened?
• How did it happen?
• Why did it happen?
• What were the multiple managem ent system-related root causes?
• What can be done to prevent a repeat or lower the risk?
The subject matter of the report may influence aspects of a report’s
layout. The guidance below presents a lo gical sequence of se ctions that allow
someone reading the report to under stand the circumst ances, findings,
causal factors, root causes, and re commendations. Some organizations
develop standardized report formats, whic h can vary by categorization of the
incident (API RP-585, 2014). |
Management of P&ID Development
13
At the beginning of the P&ID development, there is
no need to keep track of changes as there can be huge
additions or changes on P&IDs, and also those changes generally do not have any impact on other groups’ schedule and budget. This period could be the pre‐IFA period.
After getting close to a stage that a “skeleton” of a
P&ID is developed, the program of keeping track of the markups should be started. For example, it could be instructed that every change on P&ID should be logged with initials and the date. Or the procedure could be that every single change should be marked up and ini-tialed, but it will remain there if it is approved by the Process group or by the Process group and the project
engineers. This type of tracking program could be implemented for the P&ID development between IFR and IFA or IFA and IFD.
After issuing an IFD begins a critical period of P&ID
development. In this stage, all other groups start to do their work. Then, every change on the P&ID may impact them. Therefore, a more stringent tracking program should be implemented. In this stage, no P&ID change can be added unless it is approved and signed off by other groups including the project engineer or manager.
Therefore, for P&ID additions, deletions, or changes
after IFD, a document package should be prepared and the change title logged in a log sheet by the proposer of the change for every change. There is usually a P&ID change meeting on a weekly or biweekly basis to discuss and approve or reject every proposed change. In this stage, no change on a P&ID can be marked up before approval by the group lead. In later stages of project (e.g. post‐IFC stage), every single change may need to be filed in a change notice.
This concept is shown in Figure 2.4.
2.8 Required Man‐Hours
for the De
velopment of P&IDs
It is difficult to introduce a methodology for predicting the required man‐hours for P&ID development because each company may use its own methodologies for pre-dicting the required man‐hours. The required man‐hours depends on the available tools in companies, the maturity and completeness of their P&ID development guidelines, and the skill of the designers.
The estimation of man‐hours is normally the respon-
sibility of each group. Process, Instrumentation and Control, Drafting, Piping, Mechanical, Civil, and other groups estimate their own hours for P&ID development. However, in some engineering companies, only two groups (Process and I&C) consider and assign man‐hours under the category of P&ID development. The other groups generally do not assign specific separate man‐hours for their P&ID involvement; it could be because their involvement is not as large as that of the two main contributors of P&ID development.
The Piping group has a specific role. On the one hand,
they are mainly the “users” of P&IDs wherein one may say they do not need any assigned man‐hours for P&ID development. They, however, may need to contribute to the P&ID development during detail stage of project. There are cases in which the Piping group design affects the P&ID. In such cases, the Piping group may need to incorporate changes on the P&ID.
The goal of man‐hour estimation for P&ID develop-
ment is coming up with required man‐hours for develop-ing each single P&ID for each incremental development or for development from the beginning to the IFC revi-sion of P&ID. For example, a process engineering lead Table 2.1 Change mark ups on P&IDs.
Intention Color used
Addition Blue
Deletion Red
Comment Black
Marking up without trackingIFRI FA IFDI FC
Marking up with initial
Marking up with logging and team approval
Document change notice before marking up
Figure 2.4 Managemen t of change on P&IDs. |
Manual Valves and Automatic Valves
113
7.8 Tagging Automatic Valves
Generally speaking, all ROT should be tagged on P&IDs.
Table 7.9 summarizes the methodology for tagging automatic blocking valves and throttling valves.
7.9 Tagging Manual Valves
Tagging manual valves is not as common as tagging automatic valves. Some companies do not tag their manual valves because they consider manual valves to be less important than automatic valves. However, in process plants that produce critical products, such as nuclear power plants, explosive material plants, and herbicide production plants, it is common to see tagged manual valves.
The manual valve tag anatomy is generated on a per‐
company basis, but it generally has an alphabetical acronym and a sequential number.
7.10 Valve Positions
When talking about valve positions, there are two differ
ent positions: regular and fail. Regular position is the position during normal operations of plant, and the fail position is the position when losing the driver of “operator. ”
The regular position of a valve is an attribute of block
ing valves, either manual or automatic. Regular position refers to whether the valve is fully open during normal operation of the plant. To show this position, an acronym is placed under the valve on the P&ID. For example, NO means normally open When a valve is NO, this means that during normal operation of the plant, the valve should be fully open, closing only during plant maintenance, shutdown, and so on.
The fail position of a valve is an attribute of automatic
valves (either throttling or blocking). It refers to the position of the valve when the “driver” of the valve is lost (fail condition). For example, a control valve (automatic throttling valve) could be called fail open (FO); this means that if the instrument air to this valve is lost, then the valve will go to its open position and stop in that position. Table 7.10 shows the assigning of the regular position and failure position to the respective valves as a clarification to the preceding discussion.
In the next two sections, regular positions and failure
positions will be detailed more fully.
7.10.1
Regular Position of Blocking Valves
and Decision Methodology
Regular positions of blocking valves are classified in
three pairs: normally open/normally closed, locked open/locked closed, and car seal open/car seal closed. Each could be used in a condition, but all pairs represent a valve that has a regular position of being open or closed.
Deciding on the position of block valves is done by
evaluating three items: safety, equipment protection, and process smoothness (Table 7.11). If a valve needs to be opened or closed just to support the process, it can be specified as normally open (NO) or normally closed (NC). If the decision for the valve’s regular position is based on equipment protection, then the valve would be locked open (LO) or locked closed (LC). If a blocking valve needs to be closed to protect a plant from a hazard, it is specified as car seal closed (CSC). If it needs to be open for safety reasons, then this valve is specified as car Table 7.10 Regular and failur e position of valves.
Throttling Blocking
Manual
NC
Automatic
FCFC
NCTable 7.9 Two main types of valves and their tagging.
Blocking Throttling
Name of valve Tag Name of valve Tag
Time Sequence valve KV – –
Event Switching valve XV or UV Control valve FV (flow control valve)
Human operator Shutoff valve XV Hand control valve HCV (hand control valve) |
Appendix 222
Process Safety Culture
Compliance with
Standards
Process Safety
Competency
Workforce Involvement
Stakeholder
Outreach
Process Knowledge
Management
Hazard Identification
and Risk Analysis
Operating
Procedures
Safe Work
Practices
Asset Integrity and
Reliability
Contractor
Management
Training and Perform.
Assurance
Management of Change
Operational Readiness
Conduct of Operations
Emergency
Management
Incident
Investigation
Measurement and
Metrics
Auditing
Management Review
and Contin. Improv.
5 3 % 4 7 % 1 2 3 4 5 6 7 8 9 1 01 11 21 31 41 51 6 1 71 81 92 0
5% 3% 6% 1% 2% 9% 15% 10% 3% 12% 1% 3% 8% 2% 3% 8% 4% 1% 1% 2%
Year 35 31 16 12 20 3 6 30 52 35 10 40 5 12 27 8 11 29 15 4 3 7
Three Mile Island Nuclear
Plant111 1 1 1
Chernobyl Nuclear Plant,
Russia111 1 1
PEMEX Terminal
Mexico City1 1 1 1 1
Tupras Refinery
Turkey Earthquake1 11 1
Texaco Oil Refinery
Milford Haven111 1 1 1 1
Elf Refinery BLEVE
Feyzin, France1 11 1 1
BP Grangemouth
UK111 1 1 1 1 1 1
Esso Longford
Australia11 11 1 1 1 1
Flixborough
UK1 11 1 1
Avon Refinery
Tosco CA11 1 1 1 1 1 1
BP Texas City 1 1 1 1 1 1 1 1 1Start-up or Shut-down transient mode incidents from:
CCPS 2009 (Incidents the Define Process Safety)
Pillar IV
Learn from ExperienceIncident
Elements Identi fied as "weak" (See Figure 10.3)
No. of Identified RBPS Causes
Risk Based Process
Safety ElementTransient
Operating
Mode
Pillar I
Commit to Process Safety
Pillar II
Understand Haz. and Risks
Pillar III
Manage Risk
Table A.2-2 Summary of the inci dents during the transient operating mode (Continued) |
15 Fatigue and staffing levels
15.1 Learning objectives of this Chapter
By the end of this chapter, the reader should understand:
• How fatigue impacts human performance and may cause erro r.
• How to manage fatigue.
• How to match staffing levels to workloads.
15.2 A fatigue-related accident
15.2.1 What happened?
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.
Several key operationa l staff had worked 12-
hour shifts for several days in a row.
The Day Board Operator was likely suffering
both acute sleep loss and cumulative sleep deficit
after working 12 hour shifts 29 days in a row. He
usually slept for five or six hours per day.
The Night Lead Operator had worked 33
consecutive days. The Day Lead Operator had
worked 37 consecutive days.
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.
15.2.2 Why did this happen?
The CSB investigation [14] attributed the excessive levels of working to
understaffing. The company did not have a fatigue risk management policy. A seven-day rotation with 12-hour shifts was operated. The shift pattern did not
cater for temporary peak workloads such as plant turnarounds.
“…the CSB concludes
that fatigue of the
operations personnel
contributed tooverfilling the tower.”
(CSB, 2007, [14] p.289) Human Factors Handbook For Process Plant Operations: Improving Process Safety and System
Performance CCPS.
© 2022 CCPS. Published 2022 The American Institute of Chemical Engineers. |
Appendix A - Human error concepts 377
Figure A-1 Energy Institute human performance principles
Error is normal
Blame fixes nothing
Context drives behaviour
Learning is vital
How you respond matters
We all make mistakes
We can predict or prevent , and have to
manage error-likely situations
Actions are rarely malicious , but well-
meaning behaviors intended to get the
job done
Organizations influence their systems
and people: the social context drives
behavior
Majority of errors associated with
incidents stem from hidden
organizational conditions
Understanding how and why errors occur
can help us prevent them
How leaders respond to failure matters;
we need to learn from mistakes
Our people are the experts of their job
and the key to solutions
People who feel valued are more en gaged
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 pe rformance 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]).
The human performance principles (Reproduced with permission [120] ) |
2. Human performance and error 17
2.4 Key learning points from this Chapter
Key learning points include:
• Errors and mistakes are caused by:
o The environment in which people work; and
o The level or type of support offered to people.
• High standards of human performance are achieved by identifying the
demands of tasks and providing the support needed.
• Understanding diverse performance in fluencing factors is important to
improving human performance.
• Understanding human errors and mistakes helps to identify how to
reduce likelihood of occurrence, thereby supporting successful human
performance.
|
EVIDEN CE IDEN TIFICATION , COLLECTION & M ANAGEM ENT 161
Table 8.5 Example Data Collection Fo rm for Recording Physical Evidence
Once the evidence has been identified, stabilized, labelled and
documented, a chain of custody proced ure should be used for situations
where the evidence needs to be moved. This may be necessary for many
reasons including:
Evidence preservation (security/ protecting from the weather)
Transportation to a test facility
Cutting or collecting samples
Chain of custody is an important issue for investigators to address for
physical data. This is not only a concern from a legal and regulatory
perspective, but it is also a good prac tice that ensures ea ch item collected is
retained, preserved, evaluated, and te sted as intended. Large accidents can
have hundreds, if not thousands, of ev idence items and keeping track of all
items is a critical task. Some data may be of interest to multiple groups. This
varied interest requires a clearl y understood and well-communicated
method for data identification whic h can be controlled by the use of
protocols.
It can be helpful to establish a secure “evidence room”, which should
have restricted access and is under th e control of one in dividual from the
investigation team. The initial documentation of evidence should include the details shown below:
Item identification (number and text)
Condition
Date and time placed in evidence room
Person delivering the item (including signature)
|
37
damage or other
cause.
3. Active A reaction capable of generating 150 psig (1034 KPa) pressure
in case of a runaway,
carried out in a 15 psig (103 KPa) design reactor with a 5 psig (34 KPa) high pressure SIS interlock to stop reactant feeds, together with a properly sized 15 psig (103 KPa) rupture disk discharging to an effluent treatment system. The SIS interlock could fail to stop the reaction in time, and
the rupture disk
could be plugged or improperly installed, resulting in reactor failure in case of a runaway reaction. The effluent treatment system could fail to prevent a hazardous release.
4. Procedural The same reactor described in Example 3 above, but without the 5 psig (34 KPa) high pressure SIS interlock. Instead, the operator is instructed to monitor the reactor pressure and stop the reactant feeds if the pressure exceeds 5 psig (34 KPa). There is a potential for human error, the operator failing to monitor the reactor pressure, or failing to stop the reactant feeds in time to prevent a runaway reaction.
Note: These examples apply specific ally to categorization of risk
management strategies with respect to high-pressure haza rds caused by a |
W ITNESS M ANAGEM ENT 123
role, primarily serving as a note taker. Th is division of tasks allows the primary
interviewer to concentrate on listeni ng and asking questions. Having a
secondary interviewer also speeds up th e interview because less time is spent
waiting for notes to be completed. This approach also prevents the witness
from feeling intimidated or becoming defensive, which can occur when
multiple interviewers start asking questions.
For follow-up interviews and general information gathering (fact-finding
type meetings), the interviewer to interviewee ratio is less critical. Later in the
investigation, it may be acceptable to have multiple witnesses present as
details and inconsistencies are resolv ed. A group interview can come across
as more open, honest, and less covert. A team atmosphere can be created.
The team will have to make this judgment based on the specifics of the occurrence and the workplace atmosphere.
7.3.4.3 Avoid Influencing the W itness
There is sometimes a tenden cy for the witness to relay what he thinks the
interviewer is expecting (wanting or waiting) to hear. There is also a
corresponding possibility for the interviewer to “lead the witness” by
inadvertently sending various response si gnals or asking leading questions.
Sometimes the interviewer is not even aw are that they are leading or steering
the discussion. Leading qu estions contain some hint of the answer in the
question. For example, consider the qu estion “After you check the pressure
you then adjust the inlet valve, right?” This wording implies to the witness
that the correct action is that the inlet valve was adju sted, although the
witness may believe otherwise. The witn ess may answer yes, just to satisfy
the interviewer.
Interviewers can also influence resp onses by repeatedly asking about the
same issue or topic. For example, if the interviewer always asks multiple questions about a procedure, the witne ss will start to relate all his answers
to the procedure because he realizes that this is important to the interviewer.
Such questions might include:
“Is that consistent with the procedure?”
“What does the procedure say next?” and
“Is that in the procedure?”
Questions asked by the interviewer sh ould be carefully worded to be as
neutral, unbiased, and non-leading as possible. A common, core group of
questions, such as those in Table 7.1 be low, should be asked of all witnesses
to provide a control sample and to obtain confirmation of key information. |
120 PROCESS SAFETY IN UPSTREAM OIL & GAS
systems survivability analysis. Some companies also develop voluntary safety cases
even if not required by local regulations.
Hazard Identification
The HIRA tools used here are similar to those discussed earlier – What-If, What-If
Checklist, HAZOP and qualitative or quantitative risk assessment. There are specific
industry guides that assist offshore hazard id entification, rather than starting the
exercise from first principles only.
A basic set of design guidance is found in API RP 14C (2017). It contains
multiple design safety features for all parts of the offshore production facility. The
approach does not focus on identification of hazards; rather it provides suggested
designs to enhance safety of that part of the facility. By checking designs against
this reference, this provides an indirect form of hazard identification as any
omissions of safety equipment can be investigated further.
An international standard ISO 17776 (2016) provides greater detail for the
hazard identification process. The standard provides a hierarchy of identification
methods from judgement to fully structured methods. No one method is always
suitable for all facilities, so the operator should consider the complexity of the
facility and the inherent hazards (hazardous i nventories, people at risk, etc.) to make
an appropriate selection. Once hazards are identified then risk evaluation is
recommended to determine if additional risk reduction measures are required. A
schematic for these approaches is shown in Figure 6-3. The standard provides a
suggested risk ranking matrix that has b een found to work e ffectively (reproduced
earlier in Figure 4-4).
ISO 17776 introduces concepts for barrier-based thinking for offshore. It
provides multiple tables with lists of h azards for all phases of offshore operations.
It also breaks out details for each hazard category and uses a standard numbering
system which assists the team to check that they have addressed all hazards and for
others to verify the quality of the assessment.
An additional, more detailed standard for hazard identification, is API RP 14J
(2001). This is a hazard identification guide that provides useful background on a
wide range of hazards in all units of an offshore facility. The 14J RP reviews several
identification methods, including checklists, What-If analysis, HAZOP, FMEA, and
Fault Tree Analysis. Some example chec klists are provided. A discussion is
provided on good layout principles – mainly separating the higher hazards from
occupied areas. It refers readers to CCPS Hazard Evaluation Procedures Guide
(CCPS, 2008a) for details on methodology selection and execution.
Fire and Blast Studies
Fire and blast studies are a part of the RBPS element Hazard Identification and Risk
Analysis , and usually follow from potential scen arios identified in HIRA studies. It
is the aim of fire and blast studies to ensure sufficient barriers protect the facility
from escalation and to assure that refuges and evacuation areas are protected from
these events. |
36 | 3 Obstacles to Learning
They can result in the release of harmful materials. Ultimately, they cause
reputational damage and break the bond of trust, whether between employee
and employer or between the company and the public.
No company wants to have an incident memorialized in a blockbuster
Hollywood movie the way that BP did with the Macondo Deepwater Horizon
blowout (Berg 2016), an enduring reminder of the consequences of not
learning from incidents. On a lesser scale, but still no better, no one should
want to make it into National Geographic’s Seconds from Disaster series, which
covered the 2005 Texas City, TX, USA, incident; Bhopal, India; Fukushima
Daiichi, Japan; and Chernobyl, Russia (National Geographic 2020). A quick
search on YouTube using the keywords “process safety incident” yields a slew
of CSB animations of incidents, and a keyword search on “Pemex Reynosa
Explosion” shows us what happened in real time, the horrifying result of not
learning from past incidents (Reynosa 2012).
To maintain a solid reputation as a good neighbor, companies must
maintain a culture of trust and open, frank communications and commit to
understanding and acting on hazards and risks well before they become an
incident or serious near-miss. These interactions must continue over time, or
the trust and communication will diminish. If organizations lose the public’s
trust, it will be difficult to regain and can lead to resistance to change (such as
expansions) in the future (CCPS 2019b).
3.5 References
3.1 Berg, P. (director) (2016). Deepwater Horizon. United States: Summit
Entertainment (Lionsgate).
3.2 CCPS (2014). Vision 20/20 Process Safety: The Journey Continues. New
York: AIChE.
3.3 CCPS (2019a). Guidelines for Investigating Process Safety Incidents, 3rd Ed.
Hoboken, NJ: AIChE/Wiley.
3.4 CCPS (2019b). Process Safety Leadership from the Boardroom to the
Frontline. Hoboken, NJ: AIChE/Wiley.
3.5 CSB (2007). Valero Refinery Propane Fire. CSB Report No. REPORT NO.
2007-05-I-TX.
3.6 Ebbinghaus, H. (1885). Memory: A Contribution to Experimental
Psychology. New York: Dover.
3.7 Grush, L. (2020). Meet the first NASA astronauts SpaceX will launch into
orbit. www.theverge.com/2020/5/20/21254315/spacex-crew-dragon- |
55 5
LEARNING MODELS
“A donkey that carries a lot of books is not necessarily learned.”
—Danish Proverb
How does a company study and learn? It doesn’t. Individuals study and learn.
Individual employees can teach their coworkers and spread what they have
learned. But companies can only be said to have learned when they make a
permanent change in the way they operate. This corporate change is the goal
of evaluating internal and external incidents.
Many individual learning models exist, as do many corporate change
models. We discuss a representative sampling of each in this chapter.
However, ultimately, individual learning and corporate change together are
not tied together in any model found while developing this book. We therefore
sought to adapt the good features and practices in existing models to a new
model that meets this book’s objective. In this chapter, we discuss the features
and practices of a range of models, extract the useful ones, and assemble
these together to form a new model for driving continuous process safety
improvement from investigated incidents.
5.1 Learning Model Requirements
An effective learning model to help drive continuous process safety from
investigated incidents should do the following:
• Empower leaders, frontline personnel, and process safety professionals to
study past incidents, both external and internal.
• Focus on areas of improvement that can have a significant impact on
corporate process safety performance.
• Direct lessons learned into appropriate corporate systems, e.g. the PSMS,
standards, and policies.
• Motivate adherence to these corporate systems over time.
• Address the obstacles to learning discussed in Chapter 3. Driving Continuous Process Safety Improvement From Investigated Incidents By CCPS and EI
© 2021 the American Institute of Chemical Engineers |
Overview of the PHA Revalidation Process 19
1.7 THE ROLE OF A PHA REVALIDATION PROCEDURE
The RBPS book identifies maintaining a dependable practice as a key principle
to address when developing, evaluating, or improving any management system
for risk [3, pp. 45-46]. Because many fac ilities include PHA revalidations within
their HIRA program, the PHA revalidation procedure (or practice) should be
clearly documented. The PHA revalidation procedure should explicitly state the
benefits, value, and expectations of PHA revalidations, and lay out the
requirements for performing consistent, high-quality PHA studies.
PHA revalidation procedures should:
• Establish responsibilities for revalidation roles and team members
• Establish revalidation schedules and cycles to ensure the analyses
are conducted in a timely fashion
• Help users determine the appropriate methods and techniques for
the study by identifying the preferred core methodology,
complementary analyses, and supplemental risk assessments and
by providing guidance on when and how each should be used
• Ensure the conduct and documentation of the revalidation
complies with pertinent company and regulatory requirements
• Provide standard outlines of the content and format of Redo- and
Update -style revalidation reports, or a desired combination thereof,
so revalidations and their results are documented in a similar
manner throughout a facility or company
While some organizations may need to address additional considerations
(e.g., due to company-specific or re gulatory requirements), a basic PHA
revalidation procedure covers the activiti es discussed in detail in this book.
Figure 1-4 shows the PHA revalidation process, including several key decision
points in determining the appropriate r evalidation approach or approaches. This
flowchart, along with this entire book, ca n be used as a guide to help develop a
PHA revalidation procedure. |
Evaluating Operating Experience Since the Prior PHA 85
Example 3 – Increased Rigor:
Background. A company’s internal audit found that misapplication of LOPA was
an issue at several process units within the facility. Conditional modifiers and
enabling conditions were overused, and therefore insufficient IPLs were
installed.
Review of Operating Experience. As part of this review, a decision is made to Redo
all the LOPA scenarios for the affected process unit. Since other aspects of the
PHA qualify for an Update , only the LOPA will be a Redo . In this case, the full
history is crucial information for the PHA revalidation team and should be
communicated prior to initiating the revalidation.
Example 4 – Transient Operations:
Background. A company has experienced multiple losses of off-site electrical
power due to wildfires in the region. The local electricity supplier changed its
policy and now proactively de-energizes power lines to reduce their risk of
causing or contributing to fires. As a result, the company now must manage unit
shutdown, restart, and standby operations much more frequently than in years
past.
Review of Operating Experience. In this case, the disruptions caused by external
factors are crucial information for cons ideration by the PHA revalidation team.
As part of this review, a decision is made to Redo all the scenarios involving loss
of power to the affected process unit. In addition, the analyses of some transient
operations (startup, shutdown, and hot standby) will be Redone . All other aspects
of the PHA qualify for an Update .
4.4 PRINCIPLES FOR SUCCESSFUL OPERATING EXPERIENCE
EVALUATION
Actual experience in conducting revalid ations, by PHA practitioners across a
number of companies, has highlighted so me keys for success, as well as some
things that can impede success. While none of the items listed are absolute rules,
they do provide valuable guidance.
Successful Practices:
• Allowing adequate time prior to th e revalidation start date to gather
and review operating experience since the prior PHA |
5.2 Risk Management-Related Element Grouping |173
Asset Integrity (Element 10)
CCPS uses the term “Asset integrity” to m ore com pletely
describe the process safety elem ent commonly called
“Mechanical integrity.” However, mechanical integrity as defined
in regulations commonly applies only to inspection, testing, and
preventive maintenance activities related to a specific set of
equipment specified by a regulation. A com pany with a strong
process safety culture should gravitate more towards the holistic
asset integrity approach rather than rely on the more limited
scope of mechanical integrity. The US OSHA PSM regulation,
which has also been adopted in whole or in concept in many other
countries, provides a useful starting point, as do other country-
specific regulatory approaches. The specification of recognized
and generally accepted good engineering practices (RAGAGEP)
can be a very useful tool to help com panies get started.
Asset integrity addresses all equipment used in hazardous
processes. Asset integrity also involves design activities such as
m aterial of construction choices and the design of the process
and layout for m aintainability. More inform ation specifically
describing asset integrity can be found in Reference 5.5.
The goal of asset integrity is to ensure that: Piping, vessels, and equipm ent safely contain the process,
Instrumentation and control elements function as
required; and
Interlocks, relief systems, and safety instrumented
systems perform their function when called on.
B y doing so, the facility can help assure that the frequencies of
equipment failures in the facility are no greater than what was
assumed in the risk assessment. This helps keep the facility’s risk
within the com pany’s risk criteria. When asset integrity is
functioning well, inspections and testing will periodically reveal
that equipment or components must be replaced. When a
com ponent is critically deficient, strong leadership should be •
•
• |
260 PROCESS SAFETY FOR ENGINEERS: AN INTRODUCTION
factors and to LOPA. This edition will assist the reader in delivering optimum safety and
efficiency of performance of the HAZOP team. (Crawley)
Hazard identification study documentation software. Many software packages are
available to facilitate the documentation of hazard identification studies. A selection of these
are listed below.
Spreadsheet software. Although simple, spreadsheet software such as Excel can be
used to document these studies.
Sphera PHA-Pro. This is a popular vendor of haza rd identification software that
facilitates the leading and document ation of these studies. (Sphera)
Primatech PHAWorks. This software facilitates the documentation of hazard
identification study worksheets. It also includes generic checklists. (Primatech)
AE Solutions AEShield. This software facilitates the documentation of hazard
identification studies and the linkage of those studies to subsequent studies such as
LOPA (discussed in Chapter 14) and to SIL verification (discussed in Chapter 15). (AE
Solutions)
Summary
Hazard identification is fundamental to risk ma nagement. If the hazard is not identified, it
cannot be managed. Process hazards analyses are used to identify hazards during the
engineering design of a project and the operatio n of a facility. Many different process hazard
analysis methods are available, and it is important to select the appropriate method for the
purpose and for the engineering design data availa ble. It is also important to have a competent
study leader and a team with the appropriat e technical expertise for the equipment being
studied. Hazard identification analyses are typically qualitative although they may risk-rank
scenarios for the purpose of prioritizing re commendations. Hazard Identification studies
frequently are the source for data that are used in risk assessment or other process safety
analysis. Hazard identification is only the first step. Having identified the hazard, actions should
be taken to prevent its occurrence or minimize the impact. Having a process in place to ensure
that recommendations or actions are considered, changes are made, and validating that the
change addressed the hazard identified are all key steps in a successful hazard identification
program.
Other Incidents
This chapter began with a description of the Esso Longford gas plant explosion. Other incidents
involving hazard identification include the following.
T2 Laboratories Reactive Chemicals Expl osion, Jacksonville, Florida, U.S,. 2007
Celanese Pampa Explosion, Texas, U.S., 1987
Hickson Welsh Jet Fire, Yorkshire, U.K., 1992
Port Neal AN Explosion, Sioux City, Iowa, U.S., 1994
Georgia Pacific Hydrogen Sulfide, Pennington, Alabama, U.S., 2002
Hayes Lammerz Dust Explosion, Indiana, U.S., 2003 |
9 • Other Transition Time Considerations 183
be labelled to warn the personnel of unsafe atmospheres inside the
equipment.
If there is an area where mothballed equipment items will be
stored for possible use later, there will be similar asset integrity management challenges, including re taining the specific equipment
design and inspection documentation for the equipment located in the equipment “boneyard.” If the moth balled equipment or process unit
does start-up later, special recommiss ioning procedures should be
prepared to reverse any preventive procedures that were used when preparing the mothballed equipment.
It is important that all ITPM tasks are completed and verified
before restart. In addition, these in spection, testing, and preventive
maintenance tasks should be re-e ntered in the facility’s ITPM
programming schedule to ensure ongo ing asset integrity. The facility
or equipment should also be subject to an operational readiness review prior to restart [14]. At some point, the mothballed and boneyard equipment will eventu ally progress to the permanent
decommissioning stage (see Section 9.8).
9.7 Incidents and lessons learned, mothballing
Details of some mothballing-related incidents are included in this
section. The incident summary is provided in the Appendix.
9.7.1 Incidents when recommissioning mothballed equipment
C9.7.1 -1 – Starting up a Mothballed Vessel [20, p. 455]
Cause of incident occurring during the start -up: Unremoved residue
reacted with the fresh ingredients.
Incident impact : The reaction caused an in crease in temperature and
a gas release into the working area.
|
76 INVESTIGATING PROCESS SAFETY INCIDENTS
The team leader sometimes makes a br ief orientation visit and considers
numerous factors in developing an investigation plan including the magnitude of
potential outside interest in the investigation. Outside interest in the
investigation includes three aspects:
1. Legal issues,
2. Contractual issues, such as insurance coverage, and
3. Regulatory issues.
Clarify and confirm priorities
Rescue and medical treatment
Secure incident to mitigate further consequences
Environmental concerns
Evidence preservation/Secure the site
Evidence collection (includi ng interviewing witnesses)
Regulatory notification protocols
Legal counsel considerations
Plan for witness interviews
Team leader selection
Team member selection, training, and organization
Initial orientation tour/visit
Initial photography
Plan for evidence identification, preservation, and collection including special
handling of time sensitive material such as query control system logs
Plan for documentation
Plan for coordination and commu nication with other functions
Identify and plan for procurement of team supplies and equipment
Plan for any special or refresher training needed by team
Establish checkpoints, timetables, and schedule of progress
Figure 4.2 Checklist for Developing an Incident Investigation Plan
|
Selecting an Appropriate PHA Revalidation Approach 95
(changed), and Redo is applied to those portions being added or replaced (done
again from the beginning).
Regardless of the approach or approach es used to revalidate the PHA, the
goal of the effort is to ensure a complete, up-to-date, and thoroughly
documented PHA. Correcting gaps and deficiencies and revising for
changes/incidents creates a valid foundation upon which later revalidations can
build. Thus, the decision as to whether to Redo just particular portions or the
entire PHA would be based, in part, upon how pervasive the gaps or deficiencies
were throughout the report.
Several examples demonstrate how Update and Redo approaches can be
used within the same revalidation:
Example 1 – Changes to Risk Ranking System:
Since the prior PHA, the company risk ranking system has changed and now
requires the use of a different risk matrix. Otherwise, the Update approach
would be appropriate. The revalidation team could choose to delete all risk
rankings in the existing HAZOP prior to the meeting but leave everything else.
The revalidation team would then (1) Update the PHA for changes and incidents
as appropriate and (2) Redo the LOPA calculations of all risk rankings in the
HAZOP that still warrant a LOPA.
Figure 5-2 Combining Update and Redo in the Same Revalidation |
APPLICATION OF PROCESS SAFETY TO WELLS 61
Figure 4-3. Example offshore BOP
(derived from Shell, 2015)
Further elements of the BOP system are the choke and kill lines. These pipe
connections are used to terminate a bl owout event but may also be used for
measurements to determine if an influx is occurring. API 16 C (2015c) provides
specifications for choke and kill equipment.
Subsea BOPs include electro-hydraulic pods which are a combination of
computer and hydraulic control valves. The rig sends control signals to the computer
which activates the subsea hydraulic valves and uses the hydraulic energy stored
subsea to activate the rams in the BOP. There are two automatic systems legally
required in the US: one for a disconnect of the riser and one for loss of signal from
the rig. Other than these two events, BOP actuation is manual. Subsea BOPs are
controlled with dual redundant pods – often termed blue and yellow. These respond
to signals from the driller on the surface and direct hydraulic fluid to operate the
multiple BOP systems. Remotely Operated Vehicles (ROVs) are used to address
|
14
All five categories can contribute to the overall safety of a process.
Ideally, the steps of analyzing, re ducing, and managing risk will be
considered in a hierarchical manner as shown in Figure 2.1 (Ref 2.21
Kletz 2010).
Inherent safety uses the properti es of a material or process to
eliminate or reduce the hazard and ar e preferable to passive, active, and
procedural measures. Passive protective layers are preferable to active
layers, and passive and active measures are preferable to procedural ones, which involve direct human inte rface. The fundamental difference
between inherent safety and the other four categories is that inherent safety seeks to remove or at least reduce the hazard at the source, as
opposed to accepting the hazard and attempting to control it in some
manner. If implementing inherently safer approaches alone to meet
project risk goals is feasible, othe r layers of protection - and their
associated costs in time, capital, and ongoing expenses - may not be
required.
Figure 2.1 Hierarchy of Controls (adapted from Ref 2.4 Amyotte 2006) |
Utilities
367
plants the fuel gas specification requires the addition of
an odorizing compound to the natural gas to make the detection of leakage easy for operators.
17.9 Inert Gas
“Inertness” could be a confusing adjective here. “Inert” here means a gas that would not be involved in any reac -
tion with the liquid contents, nor with the equipment body material. As a fundamental requirement, an inert gas should not be toxic to humans and should not be flammable.
The best and the most expensive inert gas is nitrogen
gas, but this gas is usually not economically viable for use in processing plants. One common application of nitro-gen gas as an inert gas is in the food industry. The other expensive gas used as an inert gas is carbon dioxide.
Another inert gas that is very common in the oil industry
is natural gas. The main composition of natural gas is meth-ane, which is a very stable chemical and can be considered as an inert gas. However, one concern that may be raised is the flammability of methane or natural gas. However, this concern is not valid. There is a chance of combustion only when all elements of the fire triangle are available (air, flam-mable material, and heat). When the oxygen content is minimal, the mixture is too rich to inflame.
17.9.1
Blanket G
as
In some atmospheric tanks, there is a need to provide a
specific atmosphere at the top, empty space of the tank above the liquid level. This can be done by blanketing (or padding).
Leaving that top space without any specific atmosphere
would provide an opportunity to have a space full of air (oxygen) and vapors from the liquid. This could be dan-gerous for the liquid or for the tank. In the majority of cases, the existence of oxygen in that space would promote corrosion inside the tank, which is not a good thing. In some other cases, leaving that space with oxygen and flammable vapors from the liquid would create a flamma-ble mix, ready to be ignited and cause the tank to explode.
In still other cases, the conventional air atmosphere
can degrade the liquid. For example, some petrochemical plants have tanks full of liquid monomer, which is stored to be sent to the polymerization unit; however, the exist -
ence of oxygen in the top space of the tank would trigger the polymerization reaction of the monomer in the tank, rather than in the polymerization unit. A polymerization reaction in a tank is not good because it is not designed to handle this. In this case, again we need to create a spe-cific atmosphere instead of a conventional air atmos -
phere in the top space of the tank. Providing a specific atmosphere inside the top space of the tank can be done by introducing a specific gas into that space.In the hydrocarbon industry, natural gas in the most
common blanketing gas. The only concern regarding the application of natural gas as a blanketing gas is some solu-bility of methane into the liquid. If dissolved methane in the stored liquid is acceptable, then natural gas can be used.
If in a plant there is only one tank requiring blanketing,
then there is no need for a “blanketing gas distribution network. ” However, if there is more than one tank that needs blanketing gas, a blanketing gas distribution net -
work should be installed in the plant, and this can be considered as another utility network.
17.9.2
Pur
ging Gas
Inert gas is used where it is needed to push out a gas or
vapor from the piping or equipment. One famous purging gas is nitrogen.
A gas or vapor would be the target for removing gas
from a space for different reasons. Air or a hydrocarbon gas may need to be pushed out of a system to prevent creation of a flammable mixture. This operation is named purging. The operation may be needed after the plant shutdown or before the plant start‐up.
Providing inert gas for burners and fired heaters is very
common.
While steam cannot be used for blanketing it is com-
mon to use it for purging purposes.
17.10 Vapor Collection Network
A vapor collection network is designed to collect vapors out of process units that are not categorized as “emer -
gency” vapors. The most famous non‐emergency vapor producers are tanks. Tanks generate vapors because of their normal breathing.
A vapor collection network is a set of pipe routes
to direct the vapors to a vapor recovery unit (VRU) (Figure 17.8).
To Atm.
To VRU
VRU
Figure 17.8 Vapor c ollection network. |
Provisions for Ease of Maintenance
137
It could be questioned why we don’t do this in the first
place; gravity draining doesn’t need any energy while the
first solution needs equipment and also some external energy. The answer is that gravity draining is generally a very slow process. Removing fluids from equipment using their gravity usually takes a long time. Because of this, for emptying we prefer to use an existing pump in the first place and then rely on gravity draining.
To remove liquids, first of all the equipment or equip-
ment internals should be oriented to naturally direct the liquid toward some drain valves to drain liquids and fully empty the piece of equipment.
Drain valves are usually small valves that the designer
should put in different locations on piping and on equipment to provide drainability for the system. The drain valves are a type of on/off valve and could be gate type or ball type.
To remove gas and vapors from equipment, containers,
and piping, we similarly need to put in some vent vales. Vent valves again are small valves in the form of gate or ball valves.
Figure 8.10 shows drain and vent valves on different
equipment.
The responsibility of the P&ID developer is to decide
on the location, number, and size of drain and vent valves in the system.
If there is some solid or semi‐solid material inside of a
piece of equipment, emptying is more difficult than for fluids. The P&ID developer needs to put adequate number of nozzles in the piece of equipment to be able to empty the equipment by gravity. In such cases the piece of equipment should have a high height to diameter ratio, otherwise the solids won’t be removed from it by gravity. In difficult cases, machines (like a vacuum cleaner) may need to be used for removing solids or semi‐solids from the equipment.
8.8.2.1 Location and Number of Drain/Vent Valves
The required number of drain/vent valves is decided based on several points.
The first point is that draining and venting of a piece of
equipment or system should be done in a reasonable time. This gives a rule of thumb about the number of drain and vent valves that are needed to drain a specific portion of a process plant. For example, it is not acceptable that a small tank (less than 300 m
3) be drained within three days!
The second point about placing drain and vent valves is
that each drain or vent valve has a specific coverage area (vents can cover a bigger portion). This means a single drain valve on a piece of pipe cannot empty all the pipe from point A to point B. At least one drain valve should be placed for each piece of pipe between two isolation valves, or isolation/check valve, or isolation valve/safety valve, or isolation valve and concentric reducer, or enlargers.
Figure 8.11 shows the spans on pipes that need drain
valves.
However, additional vent and drains may be needed if
the pipe route goes up and down. The general rule is that at each summit point we need to put a vent valve, and in each valley point we need to put a drain valve. However, these high point vents and low point drains cannot be “seen” on a P&ID. They only can be recognized from a
Tank VesselH XP ump casing
Figure 8.10 Ven ts and drains on different equipment.
Recommended:
To atmosphere
Figure 8.11 Requir ements of drain valves on pipes. |
432 PROCESS SAFETY FOR ENGINEERS: AN INTRODUCTION
Operators had three ways to operate the BO P in an emergency mode. The explosions
likely disabled the first method. Later investigation showed that the second method which
should have worked automatically without oper ator action, likely did not function due to
critical control pods on the BOP that were faulty . One had a fault in the solenoid valve, and one
had insufficient battery charge. Lastly a remote op erated vehicle was used to again try to close
the blind shear rams, but by this time (33 hour s later), the drill pipe had buckled in the BOP
and was forced outside of the zone of the blades of the blind shear ram. The failure to stop
flow through the BOP resulted in the prolonged oil spill.
Figure 21.3. The diverter system on a rig
(CSB 2014d)
|
SUSTAINING PROCESS SAFETY PERFORMANCE 451
Incident investigation
Process safety incidents can lead to loss of liv es, money, and a company’s reputation. Incident
investigation is a process for reporting, tracking , and investigating incidents and near misses.
(Refer to Chapter 9 for information on near misses and classification of process safety
incidents.) This includes a formal process for conducting incident investigations including
staffing, performing, documenting, and tracking in vestigations of process safety incidents. It
also includes the trending of incident and inci dent investigation data to identify recurring
incidents. The purpose of incident investigatio n is to identify and eliminate the causes of
incidents to prevent their recurrence and sustai n or improve process safety performance. The
incident investigation process also mana ges the resolution and documentation of
recommendations and action generated by the investigations.
Incident investigation provides a way of learni ng from incidents that occur over the life of
a facility or business and communicating the le ssons learned to both employees and other
stakeholders. Depending upon the depth of the anal ysis, this feedback can apply to the specific
incident under investigation or a group of incide nts sharing similar root causes at one or more
facilities.
Incident investigation should not be used to assign blame. Assigning blame serves to stop
the investigation short of identifying the root caus es. The failure to identify root causes results
in ineffective recommendations being implemente d. A more effective approach is to pursue
the investigation to the root causes and develop recommendations that address the
management causes of the incidents.
Root Cause - A fundamental, underlying, system-related reason why an
incident occurred that identifies a correctable failure(s) in management
systems. Typically, more than one root cause can be found for every
process safety incident. (CCPS Glossary)
Incident investigation begins after the emergenc y is contained and the site is stabilized. At
this point, preserve evidence in the field an d electronic data or images that could be
overwritten. Investigation personnel entering the scene should be qualified to enter a
potentially hazardous area as there may be addi tional hazards present following an incident.
The incident investigation team should also interview witnesses and those present during the
incident as practical after the incident as memories fade quickly.
Several types of incident investigation method s are available. The method used will usually
depend on the perceived severity of the incide nt or near miss. These methods can range from
simple brainstorming to creating logic trees. On e simple technique is the ”5 Whys” method as
shown in Figure 22.3. (Serrat 2009)
Using the 5 Whys method to illustrate the point of not pursing an investigation until the
root causes are understood, a recommendation migh t have been to replace the pipe or adjust
the inspection frequency. But this would not have addressed the root cause that the MOC
procedure should be improved to clarify its app licability to changes resulting from budget cuts.
By doing this, a broader range of potential inci dents, those impacted by budget cuts, can be
addressed. |
Piping and Instrumentation Diagram Development
24
no connection to the rest of elements, they can be
drawn on another P&ID sheet.
9) No vi
siting stream is allowed. A visiting stream is a
stream that has no connection with other items on the P&ID. See Figure 4.5.
10)
Do not
try to present a P&ID in a way that follows
geographical directions (e.g. north). P&IDs are drawings independent of these.
The rest of this section outlines the other visual issues
in drawing P&IDs and the general practices in dealing with these issues.
4.2.1
Line Cr
ossing Over
Lines (in different forms) are symbols for pipes or
signals. Crossing lines should be avoided or kept to a minimum, as well as changing the direction of lines. However, sometimes both are inevitable,and it is common to see lines crossing each other in P&IDs.
It is important to know that crossing lines in a P&ID
does not reflect the reality of pipe routes in field. Crossing
lines in the P&IDs is not acceptable aesthetically, but it is
Yes
vs.
No
Figure 4.4 Equipmen t should be fairly distributed horizontally on
the P&ID.
Figure 4.5 Showing visiting str eams is not a good practice. |
KEY RELEVANCE TO PROCESS PLANT OPERATIONS 75
Example Incident 3.15 – Batch Reaction Alarms Ignored (cont.)
Lessons learned in relation to abnormal situation management:
Process Monitoring and Control:
o Alarm systems and associated procedures must align.
Alarms should be meaningful and relevant to the
operator in all modes where they receive them.
o The suppression or isolation of nuisance alarms must be
carried out under a management procedure and “bad
actors” should be dealt with as detailed in Section 5.3.2.
Procedures and Design: Operating teams should be involved in
designing such systems and associated procedures.
|