Datasets:

hndc

/

AI4Hazard-small

like 0

9. Human Factors in equipment design 107 9.7.2.4 Supporting mental models and natural mapping A key process safety measure in process in dustries is for cont rol room staff to know what is happening out on the plant, and what processes are currently in operation. This means having a representa tion in their minds (or a “mental model”) of the plant function, so that they can start and stop processes as needed and intervene if something goes wrong. The “correctness” of the mental model will influence the correctness and appropriateness of the operator’s actions. One of the problems at the Buncefield si te (see Section 9.3) was that the control room staff did not have a clear mental mo del of which tanks were being filled or of their status. If they had, they might have realized that tank 912 was being filled to an unsafe level, and they could have stopped the filling process. A key aid to helping provide an accurate mental model are “mimics” that display the plant status. These mimics should be designed so that they give correct information that is easy to interpret. Wh en the operator makes changes or inputs to the system, they should be provided with accurate and timely feedback in response to their actions. A key aspect of supporting this mental model is, as far as possible, to “map” to the arrangements in the real world. This is called “natural mapping”. A familiar example of natural mapping is the relationship between a cooking stove and its controls. Commonly used ma pping means that it is not always obvious which control maps to each cooking ring. Their relation to one another is usually indicated by a diagram next to th e controls that shows which control works for each cooking ring. With good natural mapping this would not be necessary, as the controls would be mapped in such a way that no additional guidance or information would be needed. It should be obvious how the cooking rings are operated. Figure 9-6 illustrates a way to naturally map the controls to the burner without the need for a diagram. This is called “knowledge in the world” rather than “knowledge in the head”. This means it is not necessary to know or remember any additional information to understand how an object should be operated.

| 257 Essential Practices for Creating, Strengthening , and Sustaining Process Safety Culture , First Edition. CCPS . © 2018 AIChE . Published 2018 by John Wiley & Sons, Inc.APPEN DIX A: ECHO STRATEGIES WHITE PAPER Doing Well by Doing Good: Sustainable Financial Perform ance Through Global Culture Leadership and Operational Excellence This paper may be viewed or downloaded from www.aiche.org/ ccps/publications/Guidelines-culture

130 Human Factors Handbook 11.4 Step 2: Identify required competency Operational Level – Job and Task analysis Job and task analysis explore the required job or task competency in detail and provide inputs for defining performance standards. “Task analysis for training design is a process of analyzing the kind of skills and knowledge that you expect the learners to know how to perform” [52, p. 3]. A task analysis is a systematic breakdown of a job into its component parts. A task analysis is conducted by collecting information from subject matter experts, ideally, using a walk-through/talk-through process. Where the subject matter expert demonstrates how the task is carried out. This can be supplements with interviews, focus groups and additional observations. A task analysis comprises the following roles: • Defines tasks and subtasks. • Specifies knowledge, skills, and attitu des required by each team member to complete the tasks. • Determines learning goals, and objectives. • Determines learning activities, and training requirements. A task analysis provides a breakdown of task steps and decision points to be taught in training of procedural tasks. A distinction can be made between knowledge and skills an individual or team needs (i.e., “what individuals need to know”) and knowledge which can be accessed from elsewhere (such as from operating procedures and manuals). For example, it may not be feasible to remember length y operating procedures (e.g., 20 or more pages long). Therefore, the focus of the tr aining should be on providing where to find the relevant procedure, how to understand the procedure and how to use it. Task analysis indicates which type of knowledge (information content) the training/learning opportunity should provide, such as: • Memory-based information – information that trainees need to remember (e.g., what are the visual signs of excessive rust). • Resource access and their application – information that teaches trainees: o Where to access specific knowledge (e.g., procedure). o When and how to apply this knowledge (e.g., in which circumstances/situations to use the procedure, and how to use it appropriately). See Chapters 5, 1, 7 and 8 for more information on the design and role of manuals and procedures.

A.3 Index of Publicly Evaluated Incidents | 203 Section 1. RBPS Elements (Continued) Emergency Management—Primary Findings A7 C1, C16, C17, C24, C30, C40, C45, C48, C49, C61, C62, C74 J44, J107, J109, J122, J244 S1, S3, S8, S16 Emergency Management—Secondary Findings A2, A10 C7, C23, C29, C34, C35, C36, C38, C69, C71, C77 D33 HB4, HB7 J24, J88, J149, J150, J164, J172, J227, J229, J233, J239, J245, J271 S4, S11, S12, Incident Investigation—Primary Findings C20, C34, C52 J74, J83, J86, J245, J264 Incident Investigation—Secondary Findings C14, C22, C25, C37, C39, C50 D7 J16 Metrics Not indexed. Readers seeking to improve metrics can make more rapid progress by following references A.1–A.3 (API 2016; CCPS 2009; CCPS 2018) Audits Not indexed. Readers seeking to improve auditing can make more rapid progress by following reference A.4 (CCPS 2011) Management Review and Continuous Improvement—Primary Findings C16, C27, C44, C57, C61, C70 D19 J164, J166, J169, J205, J260 S9, S15 Management Review and Continuous Improvement—Secondary Findings A10 C8, C11, C26, C38 D25 HA10 J27, J39, J42, J43, J58, J79, J82, J143, J168, J206, J207, J229, J237 S16

Table A.5 IST Checklist Location/Siting/Transportation Questions 5 LOCATION / SITING / TRANSPORTATION Questions: 5.1 Can the plant be located to minimize the need for transportation of hazardous materials? (e.g., co-located with supplier/customer, on- site production of hazardous raw materials) 5.2 Can hazardous process units be located to eliminate or minimize: • Adverse effects from adjacent hazardous installations • Off-site impacts • On-site impacts on employees and other plant facilities including control rooms, fire protection systems, emergency response and communication facilities, and maintenance and administrative facilities 5.3 Can a multi-step process, where the steps are done at separate sites, be divided up differently to eliminate the need to transport hazardous materials? 5.4 Can materials be transported: • In a less hazardous form (e.g., refrigerated liquid vs. pressurized) • In a safer transport method (e.g., via pipeline, top- vs. bottom-unloaded, rail vs. truck) • Along a safer route (e.g., avoiding high risk areas such as high population areas, tunnels, or high-accident-rate sections of roadway)? 454

1.2 Why Should We Learn from Incidents? | 7 • human factors • isolation • job hazard analysis/job safety analysis • line and equipment opening • lockout/tagout • monitoring for flammables/toxic gases in work area • non-routine work • permits to work. Very few process safety incidents occur for unforeseen reasons. In nearly every incident that has occurred, at least one, and usually more than one, important process barrier has failed. This means gaps existed in the PSMS and/or standards, or in the execution of the PSMS and/or standards. Faithful and professional execution of the PSMS and adherence to standards should control and manage all barriers. When an incident has occurred, a proper incident investigation should identify the gaps in the PSMS and standards that allowed some barriers to fail. Closing these gaps will prevent the same incident from happening again. More importantly, closing these gaps makes progress toward eliminating all incidents that the existence of these gaps would allow. 1.2.2 Acting on Learning from High-Potential Near-misses A near-miss is sometimes defined as an incident that, by chance, had no or only minimal consequences on people, the process, and the environment. As mentioned above, most process hazards require more than one barrier to control them. What if only one of these barriers fail, but the other(s) function? What if a confused operator reaches to open a valve and realizes just in time that it was the wrong valve? These, too, are near-misses. In all the following cases, a near-miss has occurred: • any time your process has gone outside the safe operating window • when testing or inspection reveals that a barrier has failed or has a latent defect • when an incorrect action was stopped just in time. Just like actual incidents, near-misses happen because of gaps or weaknesses in PSMS, standards, and process design. Because near-misses point out these gaps without harming people or your process, Near-misses are a gift — on a silver platter.

14.7 Implement | 187 building. This would minimize the possibility of inadvertently mixing ammonium nitrate with combustible or incompatible materials. The new building would be located in the remote corner of the site, as far as possible from the residential housing and equipped with a blow-out panel. To prevent new construction from coming close to the ammonium nitrate storage building, they proposed to buy the adjacent land; the company could use it to test new fertilizer blends on different crops. The cost of these changes would be significant, but much less than if the entire company had to move. In the meantime, Mei and Andrew were working on the stakeholder outreach plans. They devised a plan that called for quarterly meetings with the first responders to review the hazards at the facility. They wanted to make sure the meetings were engaging and memorable. So, they proposed to hold a series of competitions for police officers and firefighters from the various districts to see which group could best handle a simulated emergency. For the police, the challenges would revolve around how to communicate with the public in the event of a situation and what actions to take: Calmly and effectively evacuate the residents to a designated shelter, or call for shelter-in- place? For the firefighters, the competitions would test whether they responded to fires or smoke with an understanding of the chemicals in the facility. As an added twist, they would also be challenged to minimize the environmental impact by using the least amount of water possible. Mei also wrote into the plan a recommendation that the company sponsor an Agriculture Day, open to the public. The event would allow members of the public to get to know the company, plus safety could be discussed in an open environment. “The more informed the public is about what we do and what to do if an accident happens, the better,” Mei said. Wai-Kee added that the employees involved in organizing Agriculture Day would be regularly reminded of safety when blending the fertilizers. 14.7 Implement Chen was able to get a meeting with the party committee secretary, the governor of Shandong Province, and the mayor of Qingdao to make the case for not relocating the company. The government representatives were impressed by the plan, particularly the company’s willingness to build a new building to house only the ammonium nitrate, but they needed more assurances that the public would be safe. In the end, the government allowed the Qiezi Fertilizer Company to stay at its location provided it would limit the amount of ammonium nitrate in

96 Guidelines for Revalidating a Process Hazard Analysis Example 2 – Required Elements Missing from the PHA: The prior PHA, conducted using the HAZOP methodology, was reviewed in accordance with the guidance given in Chapter 4, and it was determined that the PHA did not provide a “qualitative evaluati on of a range of the possible safety and health effects of failure of controls” as required by a specific company’s PHA procedure. The revalidation team could review the HAZOP worksheets in detail and add ( Redo ) consequence rankings (or other qualitative descriptors for the range of effects) to the existing scen arios as appropriate. Remaining changes and incidents would be Updated as appropriate. Example 3 – New LOPA Requirements: Since the prior PHA, a company now requ ires a LOPA to be performed on all scenarios that have the potential for a fata lity. If the PHA is a suitable candidate based on the discussions in Chapters 2, 3, and 4, an Update can be performed on the prior PHA (since it is of good quality, changes have been well managed, incidents are minimal, etc.). Then a new, supplemental LOPA can be performed on the required scenarios extracted from the Updated core HAZOP. Note that the term Redo , as used herein, includes any analysis being done from first principles, such as the LOPA in this example. Example 4 – Hazards Were Previously Overlooked: Assume that hydrogen sulfide (H 2S) release scenarios are both credible and significant in the facility under consider ation, and they were not addressed or documented in the prior PHA. If the H 2S was present in only a relatively small portion of the process, the team could choose to Redo the affected nodes/sections and Update the rest of the PHA. Example 5 – Issues Associated with Only One Part of the Process: A PHA of a unit consists of three major process operations (Reaction, Separation, and Compression). Since the prior PHA, several major incidents and changes have occurred, but only in the Reaction section of the process. The revalidation approach could be to Redo the PHA for all nodes associated with Reaction but Update the Separation and Compression portions.

82 GUIDELINES FOR MANAGING ABNORMAL SITUATIONS Example Incident 3.20 – Physical Property Differences A hydrodesulfurizing unit included an amine scrubber to remove hydrogen sulfide from a hydroc arbon vapor stream. The amine scrubber circulation tank was fitted with a differential pressure level gauge that was calibrated for amine so lution that had a specific gravity of about 1.03. Because of an upstream problem with hydrocarbon separation, liquid hydrocarbons entered the scrubber and the circulation tank. The specific gravity of the hydrocarbons was about 0.6, leading to a significant under-reading of the level in the circulation tank. As a result, the level control system allowed the tank to fill hydraulically to the top, although it was still indicating a normal level. This created additional pressure in the tank that released through safety relief valves. The operators believed that the problem was with the relief valves that were not correctly seating, so they isolated one of them, attempting to reseat the valve. This led to overpressure and major damage to the tank. Lessons learned in relation to abnormal situation management: Understanding abnormal situations: This event occurred because the situation was misunderstood, resulting in the unsafe isolation of the relief valve. The operators had a poor mental model (understanding) of the process, which led them to diagnose the situation incorrectly and the take an incorrect action. Knowledge and Skills: Increased knowledge of upstream process upsets that may affect a process is recommended. Communications: Increased co mmunications between process units can be helpful in preventing or diagnosing abnormal plant upsets. Procedures: A robust management procedure should be in place to ensure relief valves are not isolated without conducting a risk assessment and approval process. Process Monitoring and Control: Where processes vessels can contain liquids of varying density, consider a technology that compensates for density variation, or is independent of it.

202 Human Factors Handbook 17.7 The Human Factors of control of work packages 17.7.1 The role of control of work packages in error management Control of work packages, including Permits to Work (PTWs) and work instructions, are a standard part of process safety management. PTWs can also be called Safe Work Permits (SWPs). They play important roles in error management. These roles include: • Specifying task pre-requisites, such as equipment and people. • Communicating a safe system of work, especially for infrequent, complex, and unfamiliar tasks. • Indicating who is responsible for safety critical actions, Hold or Stop Points and checks – to help spot, capture, and correct errors or unsafe conditions. • Specifying the level of task verification (See section 18.6.5), such as self- verification for low risk tasks versus independent checking for higher risk tasks. • Ensuring task sequencing occurs to avoid task conflicts. For example, stopping hot work in areas near to openings of flammable gas tanks. • Ensuring required isolation and safety actions are carried out before work takes place, for example, before opening gas pipes. • Communicating information about safety critical actions and specific items of equipment, such as statin g which valves to close and which pumps to stop. • Supporting communication between team s when long tasks take longer than one shift. For example, where two or more teams need to work together to coordinate a task through shift handover. • Supporting shared situation awareness of safety critical activities and their interaction across teams. Control of work packages are frequently used multiple times a day. This creates a potential for people to perceive control of work packages to be too detailed or unnecessarily prescriptive. If the control of work package is perceived as being too detailed, this may reduce its acceptance. If the control of work package repeats generic safety requirements this may al so cause people to think they are not needed. (Reproduced from BP [127] )

CASE STUDIES/LESSONS LEARNED 173 features and components of the flight control systems available in the report (French BEA Final Report). 7.1.5 Airbus Pitot Tube History There had been a history of problems with icing of pitot tubes on some of the Airbus aircraft. In September 2007, Airbus issued a service bulletin SB A330-34-3206 (Rev. n°00) (Air france/Airbus Bulletin) which recommended, on an optional basi s, the replacement of the THALES C16195-AA pitot tubes fitted onto a ll the A320/A330/A340s aircraft with a new THALES P/N C16195-BA model. This was said to improve the performance of the probe by limiti ng water ingress during heavy rain and reducing the risk of probe icing. At that time, only the A320s had experienced problems with the pitot tubes, so Air France’s technical teams decided to modify only the A320 fleet. They planned to replace the probes on the A330/A340s only wh en a failure occurred, as, at the time, those aircraft had experi enced no incidents involving inconsistencies in speed data. In November 2008, Airbus changed its position on the probes and stated that the “BA” model did not improve the situation with icing. By April 2009, a total of eight incidents of probe icing on the A340s and one on an A330 had been re ported. Additionally, the probe manufacturer, Thales, had conducted its own tests that indicated the “BA” model was indeed an improvem ent, so Air France decided to replace all “AA” probes with the “BA” immediately. The first batch of “BA” probes arrived at Air France on 26 May 2009 and their first aircraft was modified on 30 May 2009, just two da ys before AF 447 crashed, which had not yet been fitted with the new “BA” probes. 7.1.6 The Incident - Air France AF 447 At 22:29 on 31 May 2009, Airbus A330-203, registered F-GZCP, operating as Air France flight AF-447 took off from the Rio de Janeiro Galeão Airport bound for the Paris Charles de Gaulle Airport. Three flight crew (Captain, First and Second Co-Pilots), nine ca bin crew and 216 passengers were on board. Initially, one of the Co-pilots was flying the aircraft and the Captain was the “Pilot not Flying”. By midnight, the aircraft was flying on autopilot and autothrust at a cruising altitude of 35,000 ft.

8 Emergency Shutdowns 8.1 Introduction This chapter defines shut-downs that occur during emergencies, whether the operations team has deci ded that the process should be shut-down immediately during abnormal operations or if there has been a catastrophic release of hazardous materials or energies. A brief discussion follows on how to respon d to loss of containment incidents of hazardous materials or energies safely (Section 8.3). Guidance and discussions follow on how to antici pate and prepare for these shut- downs (Section 8.4) and on how to start-up safely after the recovery efforts have been completed (Sec tion 8.5). Section 8.6 describes incidents and lessons learned from emergencies, focusing on those pertinent to the activated shut-downs and start-ups afterwards. This chapter concludes in Section 8.7 with a brief overview of how an RBPS program and its elements can be used to manage the risks of the transient operating modes associated with these shut-downs and subsequent start-ups effectively. 8.2 Emergency shutdowns A shut-down activated for an emerge ncy shutdown is defined as “the time when the operations team has to abruptly shut the process down using emergency engineering controls and/or administrative shut-down procedures” (Table 2.2). Th ese shut-downs can occur when: The process cannot be successfully recovered from an abnormal situation, when the op erating conditions are at, or may have exceeded, the safe operating limits, and there has been no loss event; Guidelines for Process Safety During the Transient Operating Mode: Managing Risks during Process Start-ups and Shut-downs . By CCPS. © 2021 the American Institute of Chemical Engineers

THE UPSTREAM INDUSTRY 15 provide more detailed introductions to upstream operations than is possible in this chapter. Also, the SPE Technical Library provides multiple books on many aspects of well design and operation, and offshore production. Table 2-1 provides an indication of the range of drilling rigs and production facilities onshore and offshore and provides links to photographs of these facility types. This is not an exhaustive list, but it covers most facility types. Table 2-1. Types of upstream rigs and facilities Location Characteristics Description Onshore Large integrated treatment facility (see Figure 2-2) These facilities separate oil and gas remove unwanted materials (e.g., produced water), stabilize the crude oil, and export the oil and gas, usually by pipelines. If no gas export pipelines are available, the gas is compressed and reinjected. Onshore Smaller onshore drilling rig (see Figure 2-3) Drilling rig for vertical or directional drilling. Conventional reservoirs may be free flowing. Some wells may need additional treatments such as fracking or enhanced oil recovery to increase production. Offshore – shallow water Jack-up drilling rig (see Figure 2-3) These are usually 3-legged floating hulls with legs that can be jacked-up to allow for free movement and jacked-down to provide a stable platform for well construction. These are often used in 300-400 ft of water depth due to practical limits on leg height. Offshore – shallow water Jacket / Platform (see Figure 2-4 and Figure 2-9) These jackets and platforms support the processing facilities and any necessary accommodation. Accommodation(s) are sometimes located on other structures linked by a bridge. Pipelines export oil and gas to shore. Offshore - deepwater Drillship or semi- submersible rig These are floating drilling rigs (either ship-shaped or semi-submersible) which use GPS-based dynamic positioning to hold position or an anchoring system while conducting well operations.

17. Error management in task pla nning, preparation and control 199 17.6 Long and low demand tasks 17.6.1 Attention spans and task failure Everyone has a limited attention span. Th e timeframe that people can maintain attention and concentration varies among people and between tasks. Some of the factors influencing attention span s are illustrated in Figure 17-5 . Attention and concentration can fail after ab out 15 to 20 minutes or even faster in situations of low task demands, especially if: • People are demotivated. • People are fatigued. • It is during a low point of someone’s circadian rhythm (or the changing energy levels throughout the day). • It is a low demand and repetitive task. Some common examples of low task demands include: • Monitoring an automated high reliab ility process that rarely experiences faults or process upsets. • Long distance tanker driving on motorways. • Standing watch on a tanker offloading operation. These activities are long, do not require significant action and require monitoring of an unchanging situation. One of the risks of losing attention is that a person may not realize that their attention has lapsed. This may reduce the ability to recognize a loss of attention, and therefore reduce the likelihood that they will take corrective action. Motivated people performing engaging ta sks can maintain attention for longer, possibly a few hours, especi ally for diverse tasks. However, a long high demand task may actually exceed peoples’ ability to maintain concentration.

Preparing for PHA Revalidation Meetings 107 Table 6-1 is an example of scope topi cs for a particular PHA revalidation, including specific tasks for the team to complete, Updating some and Redoing others: Table 6-1 Example Scope for a Specific PHA Revalidation PHA Item (and Method) PHA Revalidation Task(s ) for Example Unit HAZOP (both Update and Redo) • MOCs/Changes – Update existing HAZOP • Incidents – Update existing HAZOP • Previous PHA recommendation resolution – Update existing HAZOP • Risk judgments for new risk corporate matrix – Team will re-assess ( Redo ) all risk rankings in the PHA • Identify any new scenarios that should undergo LOPA Facility Siting Checklist (Update)* • Review each question in the previous checklist and validate (or Update ) response Human Factors Checklist (Redo)* • Review each question and develop a new (Redo ) response; team will use the new corporate checklist and disregard the previous PHA checklist Additional Risk Analysis (Update) • Review and Update external events checklist • Review any new or affected high-severity fire, explosion, or toxic release scenarios; if revisions are needed (e.g., to the plant facility siting study due to new release points or hazardous inventory) LOPA (Update) • MOCs/Changes – Update existing LOPA • LOPA Recommendation closure – Update existing LOPA • New independent protection layer failure rate data is available; ensure the previously claimed PFDs are still accurate and Update if necessary • Add new LOPAs scenarios identified during the HAZOP review

134 Guidelines for Revalidating a Process Hazard Analysis on the effectiveness (or ineffectiveness) of safeguards credited in the PHA. The team should also consider the imp lications of recurring or prolonged operational issues along with factors such as aging, degradation, and obsolescence. If a combination of the two approaches is allowed, the revalidation team will often Update whatever is possible and Redo everything else. For example, the revalidation might involve relatively few changes, but the organization’s tolerance for risk is different. In th at case, the team members should be informed of the reason(s) for, and the co ntents of, the change record prepared ahead of the sessions. (See Section 4.2.1.) The study leader would then guide the team through the core analysis much like an Update , but the risk decisions in every node would need to be Redone in light of the organization’s revised risk tolerance. 7.1.2 Revalidation of Comp lementary Analyses The revalidation approach for complement ary analyses is largely independent of the approach chosen for the core analysis. For example, it may be adequate to simply Update a facility siting study (qualitative or quantitative), even though the core study was completely Redone . Conversely, it may be necessary to Redo a human factors analysis while simply Updating the core analysis. Some complementary analyses may be performed by experts, apart from the team re- validating the core study. Thus, a damage mechanism review may be performed by expert metallurgists, none of whom was a member of the core revalidation team. Redoing the Complementary Analyses. Complementary analyses should be evaluated just like core studies in deciding whether they must be Redone . However, few organizations require that complementary analyses be Redone simply because of their age. The Redo Example – Consolidated Facility Siting Study Some facilities have a consolidated, quantitative facility siting study that applies to the entire location. While these facilities would reference the study in the individual unit PHA revalidations, the consolidated study may or may not be Updated along with each revalidation. It is prudent for the facility to collect changes identified in the individual unit revalidations and periodically Update the consolidated facility siting study to ensure that it is accurate and that new buildings, release points, and siting concerns have been considered.

Chapter No.: 1 Title Name: Toghraei c17.indd Comp. by: ISAKIAMMAL Date: 25 Feb 2019 Time: 12:32:47 PM Stage: Proof WorkFlow: <WORKFLOW> Page Number: 359 359 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 17.1 Utility System Components Utilities need to be generated in the plant, or purchased from another plant or company. For example, electricity can be generated in a small power plant inside an oil refinery, or it can be purchased from another external power plant. Steam is usually generated in a plant by a boiler or a steam generator. To provide suitable utilities for each piece of equip- ment, the next step is to have a utility network. A utility network transfers the required utilities from the utility generator to the equipment. For example, if a piece of equipment needs steam, then steam needs to be routed from the boiler to that equipment. Because we may have multiple steam users in a plant, instead of having one pipe that is bringing steam from the boiler to equip-ment, we usually have a network of pipes that routes this steam from the boiler to multiple steam users. Therefore, in each plant, besides a string of equipment, we also have utility generation units and a utility distri-bution network. Sometimes, after a utility is used it will convert to another type of utility, which needs to be recycled back to utility generation units. For example, after steam is “used, ” it will be converted to condensate, and we don’t want to lose this by discarding it into the drain system. Plants like to save some money so we like to recycle this condensate to the boiler to convert it back into steam. So, for the majority of utilities, besides a utility distribution network, we also need a “utility collection network” for the “used” utility (Figure 17.1). For example, when we have cooling water as a cooling medium utility for a system we generally call it “CWS, ” which means “cooling water supply”; after this cooling water has been used in a system it is no longer cold, and this warm cooling water needs to be collected and returned to the system. We name this stream “CWR, ” or “cooling water return. ” Therefore, seeing – S and – R for heat transfer media is very common; these letters refer to supply stream and return stream, respectively.Table  17.1 shows the pairing nature of some utility streams. Not all the utilities have a collection network; for example, fuel gas could be used as a utility in a plant, but this utility generally doesn’t have a “collection system” per se, because the fuel gas will be converted to flue gas and released into the atmosphere after “use. ” 17.2 Developing P&IDs for Utilit y Systems To develop P&IDs for utility systems different steps should be taken. They are: 1) Iden tifying the utility users 2) De ciding on the utility network topology 3) De signing the detail of utility network 4) Pl acing priority on utility users 5) Connec ting each user to the utility network 6) Connec ting the distribution network to the collection network, if any collection network is available. These steps are explained below. 17.2.1 Identifying the Utilit y Users There are plenty of utility users in process plants. There are no universal guidelines that specify the utility users. This is the reason it happens that a P&ID development engineer misses a utility user until late in the project. Table 17.2 shows some typical utility users in process plants. 17.2.2 Utility D istribution and Collection Network Topologies Utilities can be distributed in a plant through one of these arrangements: a tree distribution network (mani-fold distribution) or a loop distribution (grid or mesh distribution).17 Utilities

9. Inherent Safety and Security 9.1 INTRODUCTION This chapter is concerned with the a pplication of inherent safety (IS) concepts as a component in addressing chemical security risk. Inherently safer approaches can reduce or elimin ate hazards integral to the process and thereby provide benefits to facility security. With increasing worldwide concern over terrorism and the increased potential for intentional attacks on chemical fac ilities and assets, those responsible for chemical facilities must approach the security of th eir processes and facilities from a holistic viewpoint. In a security context, an “asset” is a target of potential interest for an intentional attack. As such, an asset may be an individual production uni t, a specific piece of equipment, a particular area of the facility, or even the facility as a whole. Increasingly, chemical engineers are being calle d upon to apply their chemical process expertise and techniques to chemical process security management because it: 1.minimizes the risk of harm to the public, employees, and environment from intentional acts against a process facility and; 2.protects the assets within a process facility by maintaining ongoing integrity of operations and preserves the value of the investment. Inherently safer approaches and conc epts identified throughout this book incorporate security principles that reduce perception of the facility as a target, encourage greate r safety through facility layout or design, expand buffer zones, su bstitute or minimize hazardous chemicals, and apply controls to limit release or mitiga te consequences. Each IS option has the potential to enhance “defense in depth” and may improve process security so long as application does not conflict with operational requirements or create unintended security or safety exposures. 212 (VJEFMJOFT
GPS
*OIFSFOUMZ
4BGFS
$IFNJDBM
1SPDFTTFT
"
-JGF
$ZDMF
"QQSPBDI #Z
$$14 ¥

5IF
"NFSJDBO
*OTUJUVUF
PG
$IFNJDBM
&OHJOFFST

14.2 Seek Learnings | 183 to surrounding structures including apartment blocks and a railway station. Camden, AR, USA, 2019 The brakes of a commercial truck hauling ammonium nitrate in Arkansas caught fire (Rddad 2019). The driver attempted to extinguish the fire, but with no success. The fire heated the content of the truck, which led to the explosion that resulted in the death of the driver. In her search for information, she came across the May 2016 CCPS Beacon (Figure 14.1) that focused on ammonium nitrate incidents and mentioned the CSB report on the 2013 West, TX incident. She downloaded the Chinese version of the Beacon and the CSB video (CSB 2013). She decided to bring it with her when she presented her findings to Wai-Kee and Mei. Figure 14.1 Process Safety Beacon in Chinese about Ammonium Nitrate

CONSEQUENCE ANALYSIS 307 CCPS 2012, Guidelines for Engineering. Design for Process Safety , Center for Chemical Process Safety, John Wiley & Sons, Hoboken, N.J. CCPS 2012, Guidelines for Evaluating Process Plant Build ings for External Explosions, Fires, and Toxic Releases , Center for Chemical Process Safety , John Wiley & Sons, Hoboken, N.J. CCPS 2021, Process Safety in Upstream Oil and Gas , Center for Chemical Process Safety, John Wiley & Sons, Hoboken, N.J. CCPS CHEF, https://www.aiche.org/ccps/resource s/tools/risk-analysis-screening-tool-rast- and-chemical-hazard-engineering- fundamentals-chef/chef-overview. CCPS RAST, https://www.aiche.org/ccps/resources/ tools/risk-analysis-screening-tool-rast-and- chemical-hazard-engineering-f undamentals-chef/rast-overview. Clancey 1972, “Diagnostic Features of Explos ion Damage”, 6th International Meeting on Forensic Sciences, Edinburgh, Scotland. CSB 2003, “Investigation Report Chlorine Releas e DPC Enterprises, L.P.”, Report No. 2002-04-I- MO, U.S. Chemical Safety Hazard and Investigation Board, Washington, D.C., May. DEGADIS, https://www.breeze-software.com/Software/ LFG-Fire-Risk/Product-Tour/DEGADIS- Model/ . DOT 1980, LNG Facilities, Federal Safety Standard s, Federal Register, Vol. 45, No. 29. U.S. Department of Transportation, Washington, D.C. EFFECTS, TNO, https://www.tno.nl/media/10741/effects-brochure.pdf. EN 1363, “Fire Resistance Tests”, Comite Europeen de Normalisation, https://standards.iteh.ai/catalog/standards/cen/ EPA 1970, Workbook on Atmospheric Dispersion Estimates , Environmental Protection Agency, https://nepis.epa.gov/Exe/ZyNET.exe EPA AEGL, https://www.epa.gov/aegl . EPA ALOHA, https://www.epa.gov/cameo/aloha- software. EPA CAMEO, https://www.epa.gov/cameo/w hat-cameo-software-suite. HGSYSTEM, https://19january2017snapshot.epa.gov/s cram/air-quality-dispersion-modeling- alternative-models_.html#hgsystem. Explosion Research Cooperative, https://www.bakerrisk.com/products/research- development/joint-industry-programs/e xplosion-research-cooperative-erc/ Finney 1971, Probit Analysis. 3rd Edition , Cambridge University Press, London, U.K., ISBN 0- 521-080-41-X. FLACS, FLame ACcelleration Simulator, GexCon, https://www.gexcon.com/products- services/flacs-software/.

15.7 Dow Chemical Company, Dow's Fire and Explosion Index Hazard Classification Guide , 7th Edition . New York: American Institute of Chemical Engineers, 1994b. 15.8 Franczyk, T. S., The Catalytic Dehydrogenation of Diethanolamine. In Paper Preprints from the 213th ACS National Meeting, April 13-17, 1997, San Francisco, CA. (342). Washington, DC: American Chemical Society Division of Environmental Chemistry, 1997. 15.9 Heikkila, A.M., Inherent Safety in Process Plant Design An Index-Based Approach. Technical Research Centre of Finland, VTT Publications 384. 129 p, 1999. 15.10 Khan F.I. and Amyotte, P.R., How to make inherent safety practice a reality. The Canadian Journal of Chemical Engineering, 81, 2- 16, 2003. 15.11 Overton, T.A. and King G., Inherently safer technology: An evolutionary approach. Process Safety Progress 25 (2), 116-119, 2006. 15.12 Perry, R.H. and Green, D., Perry's Chemical Engineer's Handbook, 6th Edition. New York: McGraw-Hill, 1984. 15.13 Study, Karen, In Search of an Inherently Safer Process Option . In Beyond Regulatory Compliance, Making Safety Second Nature . Mary Kay O’Connor Process Safety Center 2005 Annual Symposium. College Station, TX: Te xas A&M University, 2005. 15.14 United Stated Environmenta l Protection Agency (US EPA) (January 22, 1996). Off-site Consequence Analysis Guidance (Draft). Washington, D.C.: U.S., codified at Title 40 CFR Part 68.22, 1996..432

Appendix 218 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 % 12345 67 89 1 0 1 1 1 2 1 3 1 4 1 5 1 6 1 7 1 8 1 9 2 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 C7.6.1-1Batch Vacuum Still Standby (Kletz 2009; pp. 287-288)Not Known1 11 C7.6.2-1Distillation Column Standby (CSB 2003a)2002 1 11 111 1 1 1 (C7.6.2) (A.4-1)ExxonMobil Torrance (CSB 2017)2015 1 1 1 1 1 1 1 C7.6.3-1Millard Refrig. Sys. (CSB 2015a)2010 1 1 2 1 1 1 C7.6.4.1-1Fukushima Nucl. Power (CCPS 2019)2011 1 1 1 1 1 1 1 1 1 C7.6.4.2-1Hurricane Georges (Marsh 2018)1998 1 11 C7.6.4.2-2Arkema Crosby (CSB 2018b)2017 1 11 1 (C7.6.4.2) (A.4-1)After lightning strike (Kletz 2009; pp. 549-552)Not Known11111 1 1 1 (C7.6.4.2) (A.4-1)Texaco Pembroke (UK HSE 1997)1994 1 1 1 1 1 1 1 (C7.6.4.2) (A.4-1)Esso Longford (Khan 2017, Chap 2)1998 1 1 1 1 1 1 1 1 1 1 1 C7.6.4.3-1 Hydrological EventsNone Ident. C7.6.4.4-1Ice Storm (CCPS 2018)Not Known1 1Chapter 7 - Table 1.1 Modes 7, 8 Unscheduled ShutdownsNo. of Identified RBPS CausesIncident Elements Identi fied as "weak" (See Figure 10.3) Pillar IV Learn from ExperienceRisk Based Process Safety ElementTransient Operating Mode Pillar I Commit to Process Safety Pillar II Understand Haz. and Risks Pillar III Manage Risk (Standby) Not discussed in Chapter 7 (Start-up) Not discussed in Chapter 7 Table A.2-2 Summary of the in cidents during the transient o perating mode (Continued)

Appendix B - Major accident case studies 393 B.5 DuPont Yerkes chemical plant explosion, 2010 The United States Chemical Safety Board reported that, on November 9th, 2010, an explosion occurred at E.I. DuPont de Nemours and Co. Inc. (DuPont) Yerkes chemical plant in Buffalo, New York [98]. This explosion occurred when a contract welder and foreman were repairing the agitator support atop an atmospheric storage tank containing highly fl ammable vinyl fluoride (a gas). The plant had a Tedlar® process to convert vinyl fluo ride into polyvinyl fluoride (PVF), as shown in Figure B-6. The process includes the following stages: • Vinyl fluoride is pumped from st orage tanks to a reactor to form polyvinyl fluoride slurry in wate r and unreacted vinyl fluoride. • The unreacted vinyl fluoride is transferred from the separator by a compressor and recycles back to the reactor. • After the separator stage, steam is in jected into the polyvinyl fluoride slurry to vaporize any vinyl fluoride. The heated mixture passes through a flash tank where residual vinyl fluo ride is released to the atmosphere. • Non-combustible polyvinyl fluoride fl ows from the flash tank to three insulated slurry tanks and then to a production area. • If the flash tank level is too high, an overflow line goes to tank 2. A liquid trap (seal loop) was designed to stop steam and vinyl fluoride from directly entering tank 2. An equalizer line connected all three tanks. Figure B-6 The polyvinyl fluoride process (reproduced from CSB [98]).

(VJEFMJOFT
GPS
*OWFTUJHBUJOH
1SPDFTT
4BGFUZ
*ODJEFOUT
5IJSE
&EJUJPO By 5IF
"NFSJDBO
*OTUJUVUF
PG
$IFNJDBM
&OHJOFFST Copyright
¥

5IF
"NFSJDBO
*OTUJUVUF
PG
$IFNJDBM
&OHJOFFST 13 2 OVERVIEW OF CHEM ICAL PROCESS INCIDENT CAUSATION For an investigation of a chemical process incident to be effective, the investigation team should apply a systematic approach that identifies the root causes of the incident, as defined in Chapter 1. As a rule, the benefits of this systematic approach result from: • Applying a consistent and effective investigative effort, and • Implementing sound process sa fety management principles. The investigation team should a pply an approach based on basic incident causation concepts. When a system or process fails, it may be difficult to trace the reasons for its failure. Based on available historic incident data, the makeup of a major incident is rarely simple and rarely results from a single root cause. Serious process safety incidents typically involve a complex sequence of occurrences and co nditions that can include, but are not limited to: • equipment faults or faulty design, • latent unsafe conditions, • environmental circ umstances, and • human errors. Understanding the concepts of in cident causation is essential to comprehensively investigate incidents and prevent their recurrence or mitigate their consequences throug h implementation of effective recommendations. Numerous theories and models of incident causation have been developed over the years (Heinrich, 1936; Gibson, 1961; Recht, 1965; Haddon, 1980; Peterson, 1984, etc.). These theories and models may appear at first to be diverse and disparate, bu t they do contain a number of common themes and concepts. As a resu lt of this research, industry best practices in incident investigation have evolved sign ificantly over the last few decades, based upon a number of key incident causatio n theories. This chapter discusses models that illustrate how a process safety incident can develop in a staged manner, often as a result of weaknesses in the management system. It also provi des a brief overview of key causation

EVIDEN CE ANALYSIS & CAUSAL FACTOR DETERM IN ATION 199 understanding to others. This is especially important with complex process incidents, although the di agrams can become rather complex. The technique for developing causal factor charts shares a number of fundamental principles with MES and ST EP. Basic principles for constructing sequence diagrams (Benner, 2000; Hendrick, 1987) are given below. Chart Format • All events are enclosed in rectan gles, and conditions are enclosed in ovals. • All events are connected by solid arrows. • All conditions are connected to other conditions and/or events by dashed arrows . • Each event or condition should be based upon valid evidence or, if presumptive, shown by dotted rectangles or ovals. • The primary sequence of events is depicted in a straight horizontal line (bold arrows are suggested). • Secondary event sequences are pr esented at different levels. • Relative time sequence is from left to right. Criteria for Events Description • Events should describe an action, not a condition. • Events should be described with one noun or verb. • Occurrences should be precisely described. • Events should describe one discrete action. • Events should be quan tified when possible. • Events should range from beginning to end of the accident sequence. • Each event should be deriv ed from the one preceding it. These principles are not mandator y. The most important aspect is that the investigator underst ands the incident, and these principles are meant to facilitate that objective. Some investigators draw causal factor charts differently; for example, some in vestigators do not distinguish between events and conditions. It is permissibl e to deviate from the above principles

Principles of P&ID Development 59 rotate in reverse, the bolts may not be fastened enough, thus leaking, and there can be a junk left inside pipes. A good P&ID should take care of these issues, too. For example, a temporary suction strainer (TSS) can be installed on the suction side of centrifugal pumps to protect the impeller from incoming debris during com-missioning. The strainers should be removed from the pipe later. A general procedure for starting up of a process equip- ment is as follows: ●Venting or draining the system: A system ready to be started can be filled with process fluid (gas or liquid), it can be empty, or it can be filled with air. In any case, the system should be empty at the start. ●Inerting: This step can be skipped in some cases. The goal is to replace any dangerous atmosphere with an inert and readily removable atmosphere. This is done by introducing an inert gas to the piece of equipment. ●Warming up: This step is needed if the operation tem-perature of the piece of equipment is different than the ambient temperature. To do this, a small amount of  process fluid is allowed to into the piece of equipment. ●Partial loading and recirculation: Start‐up operations can be done by recirculation. Figure  5.24 shows the basics of this procedure. If the system is not reversible, the start‐up operation can be more complicated and case specific. If the start‐up procedure goes through the recircula- tion (which is not a rare case), the highest attempt should be made to avoid using a long pipe for the pur - pose of start‐up. It is not a sound decision to spend lots of money for the pipe that is supposed to be used only during the start‐up. As much as possible, the existing pipe arrangement should be used for the purpose of start‐up recirculation, especially when a high‐bore pipe is needed.The tendency of using the piping arrangement, which was implemented for normal operation, for the purpose of start‐up recirculation is so strong that some process engineers forget to think about the start‐up operation during the development of the P&ID and assume that they will find a way to accommodate start‐up somehow. 5.4.5 S hutdown There are two types of plant shutdown: planned and emergency. The goal of a shutdown is stopping the pro duction of a plant or a unit. However, a planned shutdown happens based on a prior decision and prepa-rations, whereas an emergency shutdown occurs because of cir cumstances beyond anyone’s control. The aims of a planned shutdown are upholding oper - ators’ safety, protecting the hardware asset, and mini-mizing or preventing product loss. The restart‐up after a planned shutdown is easy. An emergency shutdown has the same aims; however, minimizing or preventing product loss may not be achievable. One aim of the interlock system is to co mplete an emergency shutdown in a way similar to a planned shutdown. During an emergency shutdown, the main driver of a process plant can be interlock system (SIS) because in such situations, the operator’s actions are not reliable. The reasons for an emergency shutdown are diverse. However, because every process plant operates based on cooperation among the elements of equipment, utility system, and instrumentation and control system, an emergency shutdown can be classified based on the fail-ure of each those elements. The error by operators or other plant personnel is the fourth source of emergency shutdown. Failure or rupture of a tank may lead to an emergency shutdown. Fire in a limited area of a plant may also lead to an emergency shutdown because of failure in equipment, utility network, or instrumentation and control system. One very common emergency shutdown is because of utility failures. For example, an electricity blackout may cause an emergency shutdown. In many cases, an emer - gency shutdown leaves stopped equipment with fluid contained in them. Such a situation is considered poten-tially problematic. The main issue about equipment with fluid in it is setting the fluid inside of the equipment, making it difficult to restart it. A winterization concept should be implemented when developing the P&ID to deal with this issue. Contrary to the word winterization, it is not limited to winter time. Using a winterization tries to prevent the settling of remained fluid in a piece of equipment because of low Unit Normally closed will be open only fo r starting up Figure 5.24 Gener al procedure for starting up a unit.

126 PROCESS SAFETY IN UPSTREAM OIL & GAS Safety critical elements demonstrate four objectives: functionality (does it achieve the desired result), reliability (does it have an acceptably low probability of failure on demand), availability (is it functio nal for an acceptable percentage of time when the facility is expected to be operating), and survivability (the element’s ability to survive the event for which it must respond). The Pryor Trust incident described in Chapter 5 shows the importance of survivability. The drilling rig should have been designed and maintained such that the hydraulic control lines actuating the BOP remained usable for an adequate period of time following a blowout-induced fire so that the BOP could be closed. In fact, the control lines failed quickly and the BOP could not be closed. 6.3.5 Emergency Management Emergency Management is an element of RBPS as well as being included in API RP 75 and in safety case requirements. Gene rally, this element comes into play after there is a loss of containment event, but some aspects are designed to minimize escalation of the incident and others to protect personnel and the asset. This section considers the fire and gas detection system and fire protection equipment as examples in SEMS and the escape routes and evacuation and rescue plan as an example of the safety case regulations. Test ing and drills are critical to maintaining the effectiveness of the system. Fire and Gas Detection System As mentioned previously, HIRA studies underpin almost all of RBPS, including Emergency Management planning. Hazard identification establishes the need for fire and gas detection and risk analysis can be used to establish the number, location, and voting logic (if any) for the detectors. Fire and gas systems that detect a fire and initiate a shutdown are usually designed as a SIS (safety instrumented system) following guidelines such as IEC 61511 or national equivalents (e.g. ANSI / ISA 84). SIS systems have three elements – detectors, logic solver and actuators, which as an instrumented function must deliver the specified reliability. The logic solver is usually a computational module and the actuators are fast acting emergency block valves. A human can take the place of a safe ty instrumented function if they meet the criteria required to determine the probl em (detect), understand what should be done (decide), and execute the task at hand (deflect), such as manually activating an ESD system. Flammable gas detectors along with flame/smoke detectors form the detection part of a fire and gas system. The number and location of detectors can be aided by historical rule sets for spacing, CFD-based scenario modeling, and by other approaches. Scenario analysis from the HIRA identifies the lower bound size of gas leaks that should be identified as no system can alarm all leak events quickly. Basic approaches use simpler consequen ce modeling tools to predict flammable envelopes, and there is increasing use of CFD tools to better account for gas flows around obstacles common in offshore modules.

Evaluating Operating Experience Since the Prior PHA 81 The CCPS book Guidelines for Managing Process Safety Risks During Organizational Change [40] details a procedure to conduct risk assessments to evaluate impact of reduction of personnel during regular and emergency operations. The PHA leader should challenge the PHA team on the validity of the safeguards or protection layer in light of the staffing changes. If the staffing changes only affect a few scenarios or particular safeguards, then the Update approach should be adequate. However, if the staffing changes may have significantly increased the likelihood of human performance gaps causing upsets or delaying the response to many loss scenarios, this may indicate the need for the Redo revalidation approach. For example, the temporary staff reduction associated with a coronavirus lockdown reportedly contributed to a ma jor styrene release in Visakhapatnam, Andhra Pradesh, India, in 2020. Turnover. S t a f f i n g t u r n o v e r m a y a l s o s i g n ificantly affect the revalidation strategy, as mentioned among the human fa ctors topics in Section 3.1.3. It is very common for a group of workers to be hired when a new unit is put in operation. That group participates in the turnover of the unit from the project team to the operations group, and some of them will likely stay with the unit for many years. Most importantly, they know the unit’s original design intent and its history of changes. For PHA revalidation, the key question is whether that same veteran team is responsible for ongoing operations. Perhaps many have retired or are planning to retire. Perhaps some have found job opportunities elsewhere. If the current or prospective operating staff lacks the deep process knowledge that was needed in the previous PHA, then the Redo approach could be a wise choice for the revalidation. In fact, some companies require a Redo in high turnover situations for its training value. They believe that the mental exercise of carefully thinking through po tential loss scenarios will improve the performance of their less experienced work ers, both by avoiding upsets and by more effectively responding to those that do occur. Example – Staffing Reduction A company significantly reduced operator staffing after a state-of-the- art distributed control system was installed. The company believed that this change would improve the overall reliability of operations by automating many of the monitoring, control, and emergency response actions. An Update of the prior PHA may be inappropriate as many fundamental assumptions made by the prior team should be questioned.

86 PROCESS SAFETY FOR ENGINEERS: AN INTRODUCTION Figure 5.2. T2 Laboratories control room (CSB 2009) After the event, the CSB estimated the explosio n was equivalent to 635 kg (1,400 lb) of TNT (CSB 2009). A key outcome of this event was a recommend ation by the Chemical Safety Board to Accreditation Board for Engineering and Technology, Inc. (ABET). The ABET now requires programs of Chemical, Biochemical, Biomolecul ar and Similarly Named Engineering Programs to include “Engineering application of these scie nces to the design, analysis, and control of processes, including the hazards associated with these processes…” (ABET 2015) The hazards include chemical reactivity hazards. Key Points: Process Safety Competency – Does someone on the job understand the importance of process safety and its proper application? We work with many very intelligent people; but that does not mean that they understand how to relate the technica l aspects of the facility to process safety concepts. Without someone on th e site asking the right questions, process safety may be lacking. Hazard Identification and Risk Analysis – What if? It is a very simple and powerful question. It can help to identify a range of hazards and once those hazards are identified, then mitigation and protection measures can be put in place.

233 through policies and procedures, traini ng is required for those who have responsibilities that impa ct the consideration and use of IS in process safety-related activities. Management promotion and adoption of the principles of inherent safety, wherev er possible, is necessary to provide a visible demonstration of leader ship and commitmen t to effective process safety management (Ref 10. 1 Amyotte). However, this is not sufficient by itself to establish th e IS program. Education and awareness of IS concepts and application are also necessary. As a general philosophy, one of the objectives of an IS education and awareness program is that all empl oyees responsible for design or engineering decisions be come aware of inherently safer concepts and “automatically” look for opportunities in their everyday work activities. This is more proactive than any review process. As technologists do their design and operations tasks, they understand the value of IS and implement the concepts where appropriate at all levels of detail, from basic process chemistry selection by re search chemists, to the writing of operating procedures by plant personnel. 10.3.2 IS in Education In recent years, the basic concepts of chemical process safety have increasingly been integrated in to the undergraduate chemical engineering curriculum at many college s and universities. As the profile of inherently safer technology increases, interest in IS is expected to likewise increase. In order to better prepare students for work in the chemical process industri es, this education should be extended to both chemistry and technical management programs. Basic chemistry is a key to ISD, and chemists need to be taught to recognize this and look for IS chemistries when conducting research into new or modified processes. IS has a parallel in current efforts to institute the philosophy and practice of “green chemistry,” which is the design of chemical products and processes in such a way as to reduce or eliminate the use and generation of hazardous substances, to minimi ze or prevent pollution and the creation of hazardous waste. Green ch emistry, which has inherent safety as a parallel benefit, is an over arching philosophy of chemistry now becoming an important part of the chemistry education curriculum. Another parallel to IS that exists in the environmental field is the current emphasis on pollution prevention, which calls for designing

PROCESS SAFETY CULTURE 461 Beyond the Management of Process Safety Process safety culture extends beyond the regulations and process safety management systems. Figure 23.1 illustrates this concept wi th the CCPS Vision 20/20 which describes the characteristics of companies with great proce ss safety performance. Vision 20/20 recognizes not only responsibilities of industry (industria l tenets) involved, but also responsibilities of external stakeholders (societal themes) nece ssary to achieve this great process safety performance. These stakeholders . include regulatory and investigative authorities, labor organizations, communities, research inst itutions, and academia. Industry working collaboratively with these stakeholders en ables great process safety performance. Table 23.1. CCPS Vision 20/20 indust ry tenets and societal themes Vision 20/20 Industry Tenets In a Committed Culture , executives involve themselves personally, managers and supervisors drive excellent execution every day, and all employees maintain a sense of vigilance and vulnerability. Vibrant Management Systems are engrained throughout the organization. Vibrant systems readily adapt to the organizati on’s varying operations and risks. Disciplined Adherence to Standards means using recognized design, operations, and maintenance standards. These standards are followed every time, all the time, and are continually improved. Intentional Competency Development ensures that all employees who impact process safety are fully capable of meeting the technical and behavioral requirements for their jobs. Enhanced Application & Sharing of Lessons Learned communicates critical knowledge in a focused manner that satisfies the thirst for learning. Vision 20/20 Societal Themes Enhanced Stakeholder Knowledge promotes understanding of risk among all stakeholders, including the public, government, and industry leaders. Responsible Collaboration is a cooperative relationship among regulatory and investigative authorities, labor organizati ons, communities, research institutions, universities, and industries. Harmonization of Standards for the safe design, operation, and maintenance of equipment streamlines practices, eliminates redundancy, and cooperatively addresses emerging issues. Meticulous Verification by knowledgeable independent parties helps companies evaluate their process safety programs from an independent perspective.

138 Human Factors Handbook 12.3 Training Needs Analysis Once the competency gaps are identified, it is important to determine learning needs requirements, via a Training Needs Analysis. Training Needs Analysis would benefit from including the knowledge of supervisors, operators, designers and managers. An effective Training Needs Analysis should provide answers to the following questions: 1. What learning is required to bridge the competency gaps? 2. Who/which roles require the training? 3. How will learning opportunity be provided? 4. How soon does the gap need to be filled? Training Needs Analysis identifies training needs based on Competency Gap Analysis findings and determines how th e competency gaps may be filled with learning opportunities. This is the case wh en a person needs further training and development to obtain the knowledge, sk ills, and/or attitudes described in the competency standard. Knowledge requirements will change in line with: • Organizational changes (refer to CCPS Guidelines for the Management of Change for Process Safety [55]) e. g., new roles and responsibilities. • Changes in process or the introduction of new technology. • Operating procedures e.g., a new control system, or a plant modification. • Insights from relevant industry process incidents or near misses. • Inputs from operator insights and experience. • Gaps identified from incidents and near misses. Changes in knowledge requirements will require learning updates, to ensure that the competency of trainees matches the latest standards requirements. Training Needs Analysis is an on-going process of analysis, which determines training needs so that training can be developed to fill any gaps in competency. Training Needs Analysis determines all the training that needs to be completed in a certain period to allow te am members to complete their jobs to the defined competency standard.

350 INVESTIGATING PROCESS SAFETY INCIDENTS 16.5.2 Safety N ewsletter A safety newsletter should convey the learning as concisely as possible. An example of a safety newsletter is the monthly issue of the CCPS Process Safety Beacon (CCPS website), which has been produced for a number of years. A copy of the Apr il 2018 issue is provided in Figure 16.2. Figure 16.2 CCPS Process Safety Beacon

120 Guidelines for Revalidating a Process Hazard Analysis 6.3.3 MOC and PSSR Records As previously noted, one of the basic re asons for PHA revalidation is to address changes that have occurred since the pr ior PHA was conducted. Consequently, one of the most important, and time-consuming, preparatory steps for the revalidation effort is the identification of those various changes. A full understanding of the number and nature of the changes is key to determining the scope of the revalidation effort, and to completing an effective revalidation. If a Redo is being performed, the intent of the Redo is to conduct the PHA as the process is currently configured with up-to-date PSI and procedures. Thus, in theory, the MOC and PSSR documentatio n should not be needed because the changes made to the process since the prior PHA should be reflected in the current information the revalidation te am will analyze and/or reference. However, if the robustness of the MOC program is questionable since the prior PHA, the information on MOCs and P SSRs for permanent changes could be collected. The information could then be examined (or audited) for any discrepancies between the “as found” condition of the process and the P&IDs. The review to ensure that all relevant updates to P&IDs and/or operating procedures are properly considered could be conducted as a separate activity in preparation for (or during) the r evalidation meetings themselves. If an Update is being performed, the MOCs and PSSRs since the prior PHA are critical inputs to the Update. During the PHA revalidation sessions, the team will address each change by updating the affected scenarios. Change documentation is often lengthy and ca n contain more information than is possible to present to the PHA team in a concise and helpful manner. During preparation, this information should be gathered and organized in a table that can be reviewed with the team, with notation of the location where more detail can be found. (See Appendix C.) Ideally, all changes would be carefully controlled and implemented via the MOC and PSSR systems, and the changes recorded in those systems would be easily collected and logged; however, some changes may have slipped through the system. It is important that changes, whether or not they are controlled or previously documented, be identified for consideration during the revalidation effort. In practice, it is extremely di fficult for a revalidation team using the Update approach to compensate for a poor MOC system, despite their best efforts. Table 6-4 lists examples of the types of changes that may be encountered. Experience has shown that there are a number of productive sources of this change information. These will be discusse d in detail in this section, along with suggested strategies for identifying the changes. While these suggestions will

62 PROCESS SAFETY FOR ENGINEERS: AN INTRODUCTION Figure 4.7. Relationship between flammability properties (Crowl 2012) Figure 4.8. Flammability diagram (Crowl 2012) A flammability diagram, Figure 4.8, also sh ows the flammable range. In addition, it illustrates that the introduction of an inert gas to the fuel/oxygen mixture can be used to bring the mixture outside of the flammable range. A description of the flammability diagram and how to use it is provided in Minimize the Risks of Flammable Materials . (Crowl 2012) Flammability properties are availa ble in Safety Data Sheets (SDS), National Fire Protection (NFPA) standards, and the CCPS Chemical Reactivity Worksheet (refer to Chapter 5) as well as other sources (refer to Chapter 7). These are im portant pieces of process safety information (Refer to Section 2.5) Table 4.1 presents flamma bility properties of selected vapors and liquids.

Table 21-2 continued Human performance tool (HPT) Description Usage STOP and seek STOP when unsure PAUSE when unsure This technique promotes awareness of work ers’ knowledge limitations as applied to dealing with specific work situat ions, deviations, or uncertainties. Workers will seek help, usually from supervis ors and/or co-workers, to continue work and to deal with these uncertainties and/or lack of knowledge. Especially when workers operate in knowledge-based modes. Pre-task and post- task briefings Pre-task briefings often follow the S-A-F-E-R pattern: • Summarize the critical steps. • Anticipate errors and error precur sors for each critical step. • Foresee probable and worst-case consequenc es should errors occur at critical steps. • Evaluate controls and contingencies at each step to prevent, catch and recover from errors, and/or to re duce their consequences. • Review previous experience and lessons learned, relevant to specific tasks and their critical steps. Post-task briefings – staff should review jo b environments, identify program gaps, and discuss corrective actions. When complications have occurred, after completing a non- routine or important work activity, or after each high-risk phase of an important project. HPT for teams.

xx Guidelines for Revalidatin g a Process Hazard Analysis This book is a supplement to the CCPS book Guidelines for Hazard Evaluation Procedures [2, p. 8] and is premised on the same concept stated in the book: This book does not contain a complete program for managing the risk of chemical operations, no r does it give specific advice on how to establish a hazard anal ysis program for a facility or an organization. However, it do es provide some insights that should be considered when making risk management decisions and desi gning risk management programs. This book outlines a demonstrated, common-sense approach for resource- effective PHA revalidation. This approach first examines a number of factors, such as PHA requirements, the quality of the prior PHA, and the operating experience gained since the prior PHA (including changes and incidents that have occurred). A revalidation approach is developed based upon this input. The revalidation concept described in this second edition has been updated based on experience and knowledge gained since the first edition was published over 20 years ago. HOW TO USE THIS GUIDELINES BOOK To use this book effectively for PHA r evalidation, it is helpful to follow the chapters in sequence. The book progresses through each key/major step in the process of revalidating a PHA. Chapter 1 – Overview of the PHA Revalidation Process explains the role of PHA and PHA revalidation in understanding and managing risk. It describes two revalidation approaches and the typical analytical tools used in revalidation activities. Chapter 2 – PHA Revalidation Requirements explains the external and internal requirements that must be sa tisfied by a periodic PHA revalidation. Chapter 3 – Evaluating the Prior PHA explains how to identify deficiencies in the prior PHA (with respect to current re quirements) and to evaluate its usability in revalidation activities. Chapter 4 – Evaluating Operating Experience Since the Prior PHA explains how to gather and evaluate information (arising from process changes, equipment changes, procedure changes, organizational changes, new research, incident investigations, etc.) that could materially alter the risk judgments documented in the prior PHA.

128 | 9 REAL Model Scenario: Chemical Reactivity Hazards The team then developed the communication and training associated with the standard. They agreed that in addition to posters which displayed the hold time and temperature limits for the two processes, they would develop a simple simulation to show how fast the reaction would run away if the time and temperature limits were surpassed. They also wrote a report summarizing their search of external incidents, the testing, and the recommendations they derived, to include in the process safety knowledgebase. 9.7 Implement Amelia brought Jason and his colleagues to the leadership team to present their recommendations. They showed the overlay graphs for past hold scenarios and similar graphs for high temperature excursions during normal operations. This got everyone’s attention, and they received no pushback on the need for the standard or the investment in control equipment. Already impressed with the analysis, leadership team members debated the communication plan among themselves, adding a few suggestions for improvement. After the proposal had been addressed to everyone’s satisfaction, Roger, the VP of operations leaned forward and said, “Please proceed with your plan. Is there anything specific you need us to do?” Amelia replied, “When we’ve completed the turnaround and are ready to start up with the new catalyst, we’d like two or three members of the leadership team to participate in the awareness campaign around the new standard and the implications in the new design.” Enthusiastic nods around the table signaled the leadership team’s agreement. “One more thing,” Amelia added. She brought up a new PowerPoint slide summarizing the general findings Jason had extracted from his review of publicly investigated incidents. It read: Recommendations • Add standing agenda item: Review status of all barriers. o Why any barriers are out of service. o Any common reasons for barriers being out of service. • Initiate special program: Identify error-prone situations. o Review progress towards eliminating them.

2.5 Maintain a Sense of Vulner ability |47 industries. Lodal (Ref 2.27) points out that the hazards inherent in chem ical, oil and gas, and related facilities have the same kind of immediacy and severity as other industries that follow the HRO m odel. Lodal also suggest that becom ing high reliability has a positive connotation that can strengthen the sense of vulnerability. A strong can-do attitude am ong facility personnel can also weaken the sense of vulnerability. Most leaders would consider this to be a positive cultural trait and an indicator of a strong esprit de corps, or personnel bonding within the organization. They m ay also point to their team’s successes and their ability to recover from upsets, preventing severe consequences. However, this trait can result in the willingness to take chances that are not consistent with the risks at hand and organization’s risk tolerance. Investigators found that the can-do attitude of NASA contributed to both the Challenger and Columbia incidents. In the earlier M ercury, Gemini, and Apollo programs, engineers and m anagers observed backup systems taking over when primary systems failed; they figured out how to return Apollo 13 safely to earth despite losing many of the back-ups; and most of all, they succeeded in sending men to the moon. Such an attitude contributed to m inim izing the O-ring and foam strike near-m isses as problem s they could work around, despite violations of safety of flight standards (Refs 2.15, 2.17) . In its deliberations regarding the process safety culture at Texas City and four related US refineries, the B aker Panel (Ref 2.5) noted that a strong can-do attitude can also result in overconfidence that encourages bypassing established procedures and practices. A sense that corporate process safety initiatives are at cross purposes with facility process safety initiatives could impair process safety can result in bypassing or ignoring of those corporate procedures. This can further result in

164 INVESTIGATING PROCESS SAFETY INCIDENTS 8.3.4 Photography and Video Photography can be used to capture a great deal of information about the condition of equipment and the relative positions of items following the incident and can be used throughout th e investigation process. The term photography is used in this section in the broadest sense and includes film cameras and a host of digital recordin g devices. Since the earliest days of image reproduction, investigators and documenters have applied this powerful tool in continuously more crea tive ways. Drones are proving to be extremely useful tools to collect video evidence from incident sites, although safety regulations must be follo wed and may disallow their use. Although photography of the scene as soon as possible after the incident should be a high priority for the team, emergency response activities, including treatment of injured personnel, containment of chemical spills, securing unstable equipment, and de-energizing systems always come first. Some hazard reduction activities could take days or weeks; nonetheless, photography may be possible in selected locations as designated by the incident commander. There is an increasing tendency for witnesses to use their mobile telephones to record incidents as they occur. While this is not encouraged (and is often contrary to safety and se curity regulations) evidence on such devices can prove to be invaluable as pa rt of the investigation process. This is discussed further in Chapter 7. It may be appropriate to declare an amnesty from disciplinary action (for use of the device against policy) in order to obtain as much relevant data as possible from personal electronic devices, although advice from legal co unsel should first be sought (See 7.3.4.11). Video footage can appear on social media platforms several weeks after an event, although this data sh ould be treated with skepticism since some details can be doctored or even faked. Incident investigation involves varying levels of photographic expertise. For most minor incidents, the team or a company employee can adequately meet the photographic needs. In cidents that are more serious may require an experienced individual, such as a forensic specialist, who systematically documents the scene, equipment involved, damage, evidence collection, and position data. For specialized photog raphic needs, the services of a professional commercial photographer or other specialists are necessary and are justified.

73 research of McGill University ha s focused on the development of numerous transition-met al-catalyzed reactions, both in air and water. Specifically, Professor Li has developed a novel [3+2] cycloaddition reaction to gene rate 5-membered carbocycles in water; a synthesis of beta-hydroxyl esters in water; a chemoselective alkylation and pinacol coupling reaction mediated by manganese in water; and a novel alkylation of 1,3- dicarbonyl-type compounds in water. His work has enabled rhodium-catalyzed carbonyl addi tion and conjugate addition reactions to be carried out in ai r and water. A highly efficient, zinc-mediated Ullman-type coupling reaction (a metal-catalyzed Nucleophilic Aromatic Subs titution between various nucleophiles with aryl halides), ca talyzed by palladium in water, has also been designed. This re action is conducted at room temperature under an atmosphere of air. In addition, a number of Barbier-Grignard-type reaction s (metal-catalyzed methods for forming carbon–carbon bonds) in water have been developed. These novel synthetic methodol ogies are applicable to the synthesis of a variety of useful chemicals and compounds. Transition-metal catalyzed reacti ons in water and air offer many advantages. Water is readily ava ilable and inexpensive, and is not flammable, explosive, or to xic. Using water as a reaction solvent can save synthesis steps by avoiding protection and deprotection processes that affect overall synthesis efficiency and contribute to solvent emission (i.e., simplification). Product isolation may be facilitated by simple phase separation, rather than energy-intensive and organic-emitting processes involving distillation of organic solvent. The temperature of reactions performed in aqueous media is also easier to control since water h as su ch a h i g h h eat ca p aci ty ( i . e. , m od er ati on ) . ( R ef 4 . 3 8 U. S. EPA – Dow) 4.3 SOLVENTS The replacement of volatile organic solvents with aqueous systems or less hazardous organic materials improves the safety of many processing operations and final product s. In evaluating the hazards of a solvent, or any other process chemical , it is essential to consider the properties of the material at the pr ocessing conditions. For example, a

218 | Appendix: Index of Publicly Evaluated Incidents NPO Association for the Study of Failure (ASF) of Japan Incident Database (For incident reports J1–J163: see www.shippai.org/fkd/en/lisen/cat102.html) Code Investigation J1 Large Explosion at a Fireworks Factory (2003) J2 Explosion Caused Due to Aged Deterioration of Equipment at an Initiative Explosive Factory (2002) J3 Fire Due to Accumulated Material in an Exhaust Gas Duct at a Finishing Section of a Synthetic Rubber Plant (2000) J4 Fire Caused Due to a Flange Loosened from Vibrations at a Synthetic Rubber Plant (2000) J5 Explosion Due to Inadequate Storage of Explosives in the Temporary Storage House (2000) J6 Explosion and Fire of Highly Concentrated Hydroxylamine at a Re- Distillation Unit (2000) J7 Fire of Hexene-1 in the Piping of a Hydrocarbon Vapor Recovery Unit (VRU) at a Tank Lorry Filling Station (2000) J8 Explosion Caused Due to Unexpected Contaminant During Neutralization Treatment in a Wastewater Tank (1999) J9 Leakage from a Crack of a Heat Exchanger Due to Corrosion and Abrasion at a Manufacturing Plant of Crude Copper Phthalocyanine Blue (1999) J10 Leakage and Fire Caused Due to an Abnormal Reaction from Contamination of a Heat Medium to Raw Materials in a Heat Exchanger Type Reactor Having a Corroded Part at an Acrolein Manufacturing Plant (1998) J11 Explosion Caused Due to Air Contamination During Subdivision Work of Trimethylindium (1998) J12 Explosion and Fire Caused Due to Mixing of Waste Acids of Different Concentrations in a Waste Acid Tank (1998) J13 Rupture and Fire of a Measuring Tank Caused Due to an Incompatible Reaction from Transport of a Different Chemical to a Tank Containing a Chemical (1998) J14 Leakage of Toluene Caused Due to Incorrect Disconnection of a Coupling During Draining of Toluene for Cleaning (1998) J15 Explosion of Ethanol Vapor Caused Due to Insufficient Exhaust During Drying Operation of Vitamin Tablets at a Pharmaceutical Factory (1998)

Piping and Instrumentation Diagram Development 74 6.2.3 Pipe Off‐Page Connector T he third pipe identifier is pipe off‐page connector. Wherever a pipe is introduced or removed from a P&ID sheet, it should carry an off‐page connector with a set of information specified by the project. Pipes on P&IDs are shown as lines. Because a plant is a large set of interconnected equipment and generally does not fit onto one sheet, there are some pipes that go from one P&ID sheet to another. Off‐page connectors are arrow‐shaped symbols that appear at the edge of P&ID sheets and show the continuity of each pipe. Off‐page connectors are determined per project and are dif - ferent within each company. Figures  6.8 and 6.9 show samples of an off‐page connector for an outgoing pipe and an incoming pipe, respectively. 6.3 Pipe Tag Anatomy A tag number bears many information about a pipe. The anatomy of a pipe tag varies from company to company; however, the following information can usually be found in a pipe tag: 6.3.1 Ar ea or Project Number Possibly the first component of a pipe tag is a number or letter that shows the area where a pipe is located. If a  pipe goes from one area to another area, generally the area that is mentioned in the pipe tag is its origin. The area designation could be like 06, 21, or AB. The area number could be between one to three figures or letters.6.3.2 Commodit y Acronym The commodity name is basically the fluid that pipe is conveying. The acronym could be between two and three or more letters; however, using the acronym is not an arbitrary choice by the designer. For example, if the con-veying fluid is water, the designer cannot arbitrarily choose to use the acronym of WAT, without consulting a commodity table. Commodity acronyms are specific in each plant and project, hence the variation. For example, fire water can be FW or, in some companies, FWA. Table 6.4 lists those that are universally used among plants. There are also some common rules in choosing an acronym for commodities. For example, H, M, and L at the beginning of utility codes mean high, medium and low. It is also common to see an S or R as the last letter, meaning supply and return, respectively. The two types of acronyms, S‐ending acronyms and R‐ending acronyms are widely used for utility pipes, which will be discussed in more detail. 6.3.3 Pipe M aterial Specification Code In essence, this code is an acronym that specifies the construction material and thickness of the pipe of inter - est. It is simply called a pipe specification (spec) or pipe class. Each pipe spec is typically a string of two to four letters or numbers or a combination of them. Each com-pany has its own pipe tag anatomy per pipe class. For example, a pipe class could be A2S. Companies try to develop pipe class anatomies that are more meaningful for the designer or reader of the P&ID. Best Not common6/uni2033-PE-BC-1043-1.5/uni2033 HC-GT 6/uni2033-PE-BC-1043-1.5/uni2033 HC-GT6/uni2033-PE-BC-1043-1.5/uni2033 HC-GT Not good Figure 6.7 Good and bad e xamples of showing pipe tags. Destination P&ID no. Destination equipmentStream nameEdge of P&ID sheet Sour gas PID-600-1390 To treater Figure 6.8 A leaving off‐page c onnector for pipe. Source P&ID no.Stream name Source equipmentTreated gas PID-500-1256 From treaterEdge of P&ID sheet Figure 6.9 An inc oming off‐page connector for pipe.

PREPARING THE FINAL REPORT 297 leader should discuss investigation status and any recommendations that affect startup (if formulated) with management so that management can make an informed decision about the timing of a process restart. As the investigation proceeds new issues may arise, open items may be resolved, and, recommendations modified accordingly. Interim report documents should be updated or an notated as necessary. Each report issued by the incident investigation team should be retained and its distribution documented. The team leader should coordinate all such interim reporting activity. Someone sh ould serve as the appointed liaison between the incident investigation team, management, and external organizations. This is often the team leader, but others with special training may also fill this function. A single communications channel is especially helpful when team members must deal with external regulatory agencies. 13.3 W RITING THE REPORT The written incident investigation report is the vehicle for documenting and communicating the investigation results. Process safety incident investigations cover a wide variety of topics, but unless a report is well laid out, the impact of its presentation is not as effective as it could be. A quality report can be extremely useful, leadin g to process safety improvements, and extending the impact of the team’s investigation. Likewise, a poorly prepared report may fail to convey important information, negating weeks or months of productive investigation. A mechanism for capturing and do cumenting the results of the investigation should be an integral part of the management system for process incident investigation. Guidelines for Technical Management of Chemical Process Safety (CCPS, 1989) states that, “The lessons learned from an incident investigation are limited in usefulness unless they are reported in an appropriate manner .” The American Chemistry Council recognizes this need by including it as on e of the twenty-two mana gement practices in the Responsible Care ®, Process Safety Code of Management Practices (ACC, 1990). The written report should convey the findings and recommendations of the investigation clearly and succinctly. It is helpful to identify the audience before drafting a report and to ensure that the report writers understand the needs of the entire audience. For example, the audience may include varying levels of management, operators, maintenance workers, engineers, future PHA teams, other sites,

84 INVESTIGATING PROCESS SAFETY INCIDENTS Table 5.2 Tier 1 Process Safety Ev ent Severity Categories (CCPS, 2018) Note: CCPS provides additional guidance on loss of containment, injury/death of wildlife, etc.

•A database of the hazards associ ated with different types of equipment and unit operations, including the applicability of inherently safer design in each. As innovative solutions to hazards in equipment and process operations are discovered, these could be shared across the company through this database, reducing risk in similar equipment and processes. A summary of design approaches for a number of common types of chemical process equipment is published in CCPS (Ref 16.2 CCPS 1998). 16.5 DEVELOPING TOOLS TO A PPLY INHERENTLY SAFER DESIGN Inherent safety is still a young and evolving area of process risk assessment and decision-making. New tools and analytic approaches are needed to help process risk as sessors and decision-makers identify inherent safety opportunities an d evaluate them against other competing risks and opportunities. 16.5.1 The Broad View and Life Cycle Cost of Alternatives One of the recognized barriers to the successful application of inherently safer design is the lack of appreciation of the benefits that can be derived from viewing a process broadly, rather than narrowly. Employing a cradle-to-grave and a feed-end-to-pr oduct-end view will lead to the development of processes which are as inherently safe as possible. Examples of myopic design could include the following. •A chemist may think only of optimizing the route of synthesis to avoid a runaway reaction hazard and not consider the safety and design implications of flammable reaction by-products. •A designer fails to identify and consider the environmental land use implications of dismantling a process at the end of its useful life. At any given time, individuals are more concerned with a particular unit operation, or with a particular life cycle stage of the process. A broader approach, however, would requ ire an analysis of the impact of the process on: •upstream and downstream operations 435

354 | Appendix F Process Safety Culture Assessment Protocol resident contractors, from certain host facility EHS activities so m uch that the resident contractors lack key safety or process safety information? For exam ple, are resident and other contractors not allowed to attend host facility safety m eetings, participate in host facility HIRAs/PHAs, or sim ilar activities? 87. Is shift turnover a form al process? Is there a procedure or checklist for shift turnover? Is it logged? M aintain a Sense of Vulnerability 88. Could a serious incident occur today, given the effectiveness of the current operating and process safety practices? When was the last serious close call or near m iss? 89. Is there belief that com pliance activities are guaranteed to prevent major incidents? 90. Are process safety policies, practices, and procedures institutionalized? Does the success of the PSMS rely primarily on the individual knowledge level, initiative, and decisions of those personnel who are assigned various responsibilities for process safety program elem ents and their activities? 91. Are lessons from related industry disasters routinely discussed at all levels in the organization? Has action been taken where similar deficiencies have been identified in the organization’s operations? 92. Do hazard/risk analyses include an evaluation of credible m ajor events? Are the frequencies of process safety events routinely determ ined to be unlikely and thus not credible? 93. Have proposed safety im provements been routinely rejected as not necessary because “nothing like this has ever happened here?” 94. Do risk analyses routinely eliminate proposed safeguards under the banner of “double jeopardy?” 95. Are critical alarms treated as operating indicators, or as near m iss events when they are activated?

191 15 CONCLUSION “The person who refuses to learn deserves extinction.” —Hillel the Elder, Rabbi, Sage, and Scholar As we discussed in Chapter 1, and as documented in many CCPS publications, the leadership and technical discipline of process safety provides significant business benefits (CCPS 2019a). These benefits include increased revenue, lower costs, improved productivity, the removal of obstacles to growth, protecting the company image, and enhanced shareholder value—all while improving leadership overall. If these benefits did not exist, however, we would still have the moral obligation and societal imperative to eliminate process safety incidents. Intellectually, we have known for a long time that any root cause gap in our PSMS, standards, culture, or policies can lead to a range of incidents. We know that by eliminating any given gap, we can prevent a wide range of incidents. We also know that properly investigating incidents and near-misses is one of the most impactful ways of finding these gaps. Yet incidents based on the same root cause gaps continue to repeat— both inside individual companies and among companies and industry sectors. We continually fail to convert what we know intellectually into practice. This outcome can be attributed to five basic factors: 1.Failure to perform complete incident investigations. This results in failure to discover key findings and to implement recommendations that address all root causes. 2.Failure to learn from others’ incidents. By doing so, the company lose the opportunity to improve without having to suffer an incident. Driving Continuous Process Safety Improvement From Investigated Incidents By CCPS and EI © 2021 the American Institute of Chemical Engineers

372 process. The regulations adopted the definition of “inherently safer technology” from the firs t edition of this book. “Inherently safer technology” means the principles or techniques incorporated in a newly designed and constructed covered process to minimize or eliminate the potential for an Extraordinarily Hazardous Substance (EHS) accident that includes, but is not limited to, the following: 1.reducing the amount of EHS material that potentially may be released; 2.substituting less hazardous materials; 3.using EHSs in the least hazardous process conditions or form; and 4.designing equipment and processes to minimize the potential for equipment failure and human error. In explaining the 2003 regulations, the New Jersey Department of Environmental Protection (NJDEP) called for consideration of the full range of inherent safety options, recognizing “that every (inherently safer technology) technique will not be suitable for every process and will rely on the owner or operator’s evaluation of which techniques, if any, will be appropriate for a particular newly designed and constructed process.” NJDEP also limited application of inherent safety as follows: “newly designed and constructed processes,” concluding that “the most reasonable and effective time for an owner or operator to evaluate inherently safer technology is during the design phase of a process, before it has been built.” These policy positions – describing “inherently safer technology” in terms of principles or techniques, reta ining the full rang e of options, and applying only during new design or construction – are critical. They demonstrate that, as late as 2003, New Jersey recognized inherent safety as a philosophy to be incorporated in the design process, rather than an endpoint to be reached. In May 2008, NJDEP proposed amendments to the TCPA rules that expanded the program in two ways: (1) all TCPA facilities—not just those with the potential for offsite conseq uences - must conduct IST reviews

Evaluating the Prior PHA 63 3.3.4 Continuous Improvement Some companies depend on having a fresh set of eyes periodically look at their hazard analysis worksheets to both id entify (1) previously undetected hazards or unrecognized loss scenarios and (2) opportunities to improve reliability, availability, throughput, yield, and product quality. Although many companies focus primarily on hazards leading to inju ries or major loss events, if desired, the PHA can identify operational impr ovement and equipment reliability opportunities as well as EHS-related hazards. 3.3.5 PHA Documentation Software Changes Most PHAs are documented using electron ic tools. The software falls generally into one of three categories: • Commercially available, general-pu rpose software (e.g., Microsoft® Office products such as Excel® or Word®) • Commercially available PH A documentation software • Proprietary or company-specific software designed for docu- menting PHAs Converting from one software package to another can be challenging (though some software packages offer impo rt functions). If the software is no longer available, no longer provides the required functionality, or is no longer licensed or used by the company (such as when a company standardizes on one software and drops the licenses for others ), the revalidation team faces a choice. It can either (1) recreate the prior analysis in the software now being used and then proceed with an Update , or (2) recreate the analysis directly as a Redo . The most beneficial approach largely depends on the cost and availability of a person or method to re-enter old data, ver sus devoting those resources to a Redo that might yield additional benefits. 3.3.6 Time Since the Previous Redo Many companies limit the number of times a PHA can be simply Updated , and they require a Redo every other or every third revalidation cycle (i.e., every 10 or 15 years), regardless of the quality of the pr ior PHA. One reason is that they want to help ensure new information and perspective are introduced to the hazard analysis process. Another reason is that some changes may not be captured by the MOC process, particularly subtle or creeping changes, and if the PHA is always Updated , any associated hazards may never be recognized. For example,

121 9 REAL MODEL SCENARIO: CHEMICAL REACTIVITY HAZARDS “Learn as if you were to live forever” —Mahatma Gandhi, Indian Civil Rights Leader The Boudin Basic Chemical Company operates a large integrated manufacturing facility along the US Gulf Coast, its only site. Boudin produces a wide range of products from petrochemical feedstocks and distributes them all over the Americas. As a small company, Boudin has only a few engineering and R&D personnel and licenses its process technology. However, the company tries to operate as professionally as larger companies, embracing Risk Based Process Safety (CCPS 2007), Responsible Care®, and a state environmental stewardship program. With limited resources, it’s challenging to get everything done. Boudin had no room to build additional capacity, and senior leadership was continuing to insist on limiting operations to a single site. All capacity expansion had to come from debottlenecking, with which the company had had great success previously. An upcoming project had been approved to replace the catalyst in four units, with the expectation that process throughput would increase by 25%. 9.1 Focus At the annual process safety management review meeting, the chief process safety engineer, Amelia, presented the company’s process safety performance to the senior leadership team. She discussed the results of a recent audit, performance metrics, and progress toward corporate goals. Members of the The individuals and company in this chapter are completely fictional. Driving Continuous Process Safety Improvement From Investigated Incidents By CCPS and EI © 2021 the American Institute of Chemical Engineers

KEY RELEVANCE TO PROCESS PLANT OPERATIONS 45 Example Incident 3.5 – Buncefield Explosion, 2005 Buncefield is a fuel storage terminal located about 3 miles (5 km) from the center of Hemel Hempstead, Un ited Kingdom, and is connected by three pipelines to oil refineries, other fuel terminals, and airports. The site was permitted to store up to 194,000 metric tonnes of fuel, including aviation fuel (Jet A1), pe trol (gasoline), diesel, and kerosene. From Saturday, December 10 to Sunday, December 11, 2005, Tank 912 was being filled with gasoline from one of the pipelines. At about 5:20 am on Sunday morning, the tank began to overflow. Unobserved by the operators, the tank continued to overflow, forming a vapor cloud, until a huge explosion occurred at 6:01 am, followed by a series of explosions and fires that ignited 20 large fuel storage tanks nearby. The explosion resulted when the va por cloud ignited, possibly from the electrical switchgear associat ed with the firewater pumphouse. The vapor cloud was produced when over 250 m3 (66,000 gallons US) of gasoline overflowed from the tank vents at the tank filling rate of above 550 m3/h. (145,200 gallons/h). The magnitude of the explosion registered 2.4 on the Richter scale. No fatalities resulted, although th ere were 43 non-serious injuries occurred. Properties nearby were destroyed, and damage extended to some 5 miles (8 km) away from the site. Environmental consequences from the smoke (vis ible from France), loss of containment from secondary containment (bunds) and the site effluent systems due to the quantity of firewater used and the type of foam. Total economic costs were estimated at £1 billion (IChemE Report 2008).

80 GUIDELINES FOR MANAGING ABNORMAL SITUATIONS Example Incident 3.18 - Unre liable Interface Detector The site had a procedure for a zero-l evel case, but implementing it was delayed because of uncertainty in th e instrument’s reading, and a lack of full knowledge on the part of the control panel operator about other clues that could be used to ve rify the actual state of the level. This demonstrates the need for no t just having a procedure but assigning the authority for implementing the procedure to the right people and training the control panel operator on backup detection of performance indicators. Lessons learned in relation to abnormal situation management: Understanding abnormal situations: The situation was generally understood, but because of unreliable level devices, the response to the situation became an uncertainty. Organizational Roles and Responsibilities: This organization had three layers of decision-making involved, which contributed to lagging response. Assigning authority to the front-line personnel can improve response time. “Stop Work Authority” needs to be defined and understood. Knowledge and Skills: In this example incident, an increased knowledge and training on ot her process parameters could reduce the time to diagnose the situation. (cont. )

INTRODUCTION 7 Chapter 2—Overview of Chemical Process Incident Causation This chapter discusses the basics of det ermining incident causation, general types of incidents, and the linkage betw een causation theories, root causes, and management systems. Understanding incident se quence models, barrier analysis, and failure modes can greatly assist investigators in dissecting the anatomy of process incidents. Chapter 3—An Overview of Investigation M ethodologies This chapter provides an overview of investigation methodologies, associated tools, and techniques th at come together to form a modern structured investigate approach. An overvi ew of the historical transition is provided along with description of methodologies and tools most commonly used by CCPS members. Chapter 4—Designing an Incident Investigation M anagement System This chapter provides an overview of a management system for investigating process safety incidents. It opens with a review of responsibilities from management through the workforce and presents the important features that a management system can addre ss to be effective. It examines systematic approaches that start with notification, team structure, functional and agency integration, document control, team objectives, etc. The learning objective is to define a management system that supports incident investigation teams, root cause determinations, effective recommendation implementation, follow- up, and continuous improvement. Chapter 5—Initial Notification, Classification, and Investigation of Process Safety Incidents Timely reporting of incidents en ables management to take prompt preventative or corrective measures to mitigate consequences. Many major process safety incidents were preceded by precursor occurrences (typically referred to as near-misses) that migh t have gone unrecognized or ignored because “nothing bad” actually happened. The lessons learned from any incident can be extremely valuable. Howe ver, this benefit is only realized when incidents are recognized, reported, and investigated. This chapter describes important consi derations for internal reporting of incidents, the process of classifying incidents into categories, and means for determining appropriate levels of investigation to be conducted.

APPENDIX B – EXAM PLE PROTOCOL 365 Mark the chain wheel (Mark #3) relati ve to the alignment mark on the housing, and phot ograph the mark. Restore System State The valve will be left in the final position from Step 6. Test Personnel and Observers One third party forensic engineer will oversee the execution of the protocol. Interested parties will be given notice and the option to attend. Operations personnel will attempt to close the valve per the protocol. A radiograph contract or will perform the radiograph. Documentation Execution of the protocol will be videotaped, except when the radiation source is exposed. Still photogra phs will be taken to document as- found condition prior to making an y changes. The ra diograph images will be the radiograph documentation. Radiation Safety Guideline The radiation source will be chosen by the radiograph contractor as appropriate for the valve being radi ographed. Because of the health hazard and potential exposure danger to radiographs, the site radiograph safety procedures will be followed.

16. Task planning and error assessment 183 Having identified potential failures, task plans are developed to support successful performance. These task plans may take the form of written work instructions, or verbal task briefi ngs, such as “Tool Box Talks”. The task plans are communicated as part of “Tool Box” talks, Tail Gate briefing, start of production shift briefings and other forms of operational briefings. This concept can be articulated as an acronym, such as “SAFER”: Summarize the task Anticipate high risk situations Foresee potential errors and mistakes Evaluate task plans Review task preparations 16.4.2 Error management and error-likely situations Error-likely situations include circumstances beyond the immediate control of the team that may impact task performance and/or the margins of safety. These can be termed “error traps”, where circumstances and conditions can cause failure. Error-likely situations can be thought of as being: • Anticipated – something could happen. • Unanticipated – something that may be more unexpected or unlikely, but could still happen. • Latent – not obvious, may be hidden wi thin systems or ways of working, but could be identified through a safety analysis. Typical error-likely situations are shown in Figure 16-1. The presence of these conditions may make the task high risk. An example of an error-likely situatio n could be a strong wind unexpectedly developing, while carrying out maintenance using a heavy lift crane.See Chapter 18 for more information on how to spot, capture, and correct errors during operations.

F.2 Culture Assessment Protocol |347 or advocating certain operations or conditions required to prove that those operations or conditions are safe? 27. Are the collection and analysis of process safety m etrics treated as adversarial or punitive activities? 28. Are managers less strict about adherence to procedures when work falls behind schedule? 29. Does the tension between production and safety result in a slow and gradual degradation in safety margins? 30. Are shortcuts encouraged and rewarded to meet production or other goals? 31. Are rewards and incentives heavily weighted towards production outcom es? 32. Has the organization included inherently safer technologies considerations in its process safety program ? 33. Have critical safe work practices (SWP) been designated as “Life Saving Rules” or “Cardinal Rules” (or similar designation for inviolable rules) for their application, with no tolerance for not doing them right every time? Im plementing and m aintaining a “Life Saving Rules” program requires that m anagem ent enforce the rules consistently. 34. Does the im perative for process safety include understanding and accom modating, and sometimes influencing/advancing the related cultures of outside organizations that interact with a facility and affect its PSM S? These outsiders include contractors, regulators, unions, corporate staff, boards of directors, interest groups, com m unity groups, and others. Provide Strong Leadership 35. Does m anagem ent have a firm understanding of risk and process safety in general, and accepts the identification of high-risk levels? Does organization senior m anagement understand the technical aspects of process safety and how process safety requirem ents are interpreted for the site/com pany?

4 Human Factors Handbook Human Factors also provides a set of principles and concepts that can be used to guide day-to-day decisions. The decisions focus on how best to support successful human performance. This appr oach helps people to understand tasks from the perspective of the person doing the work and provides ideas on how to support people to perform better. It advo cates an orientation (a way of thinking) towards making improvements that support human performance and the prevention of error. It recognizes people’s capabilities and commitment, and it aims to maximize people’s roles in safe and productive operations, and to build their ability to cope mentally and emotiona lly with stressful and demanding tasks, i.e., psychological resilience. A short video that presents a Human Factors view for successfully addressing human performance, titled Being Human , is available as a resource for “understanding and accepting why, as people, we do what we do, why we do it, and the way we do it.” [7] Human Factors covers a very wide range of topics including, training, work planning, and fatigue. Many of these to pics come under existing management systems, such as the operation of rotating shift schedule systems, and training systems. Human Factors provides knowledge, tools, and insights that can be integrated into an organization’s exis ting systems of work and operational management, safety assessments, incident investigations, and day-to-day operational decision-making. In this book, the terms ‘incident’ and ‘accident’ will be used interchangeably. 1.2 Purpose of this handbook 1.2.1 Purpose and scope This handbook provides practical advice and examples of good practice that can be applied to design, process operations, start-ups and shut-downs, maintenance, and emergency response. It is a comprehensive but simple to understand handbook aimed at people responsible for the process operations. The handbook: • Provides examples of practical application, principles, and tools. It also provides an understanding of the fundamentals of Human Factors, so the reader can develop their own approach.

eductor in a circulating heat exchanger loop (Figure 11.6). For the purpose of this example, the desi red concentration was about 30 wt.%, and the normal operating temperatur e was about 100 ºF (38 ºC) or less. Figure 15.5: Comparison of centerline vapor cloud concentration 404

8 Guidelines for Revalidating a Process Hazard Analysis 1.3 GENERAL RISK ASSESSMENT PRINCIPLES Developing an understanding of risk requires addressing three specific questions: 1. What can go wrong? (What are the hazards?) 2. What is the [credible] potential impact? (How bad could it be?) 3. How likely is the event to occur? (How often might it happen?) The PHA team systematically addresse s each of these questions for the range of credible loss scenarios. Th is gives the team a more complete understanding of the risks associated with a process and a basis for judging which, if any, exceed the organization’s ri sk tolerance. If the risk is not tolerable or ALARP, this risk understanding typi cally leads to recommendations to better control the risk, which are addressed and resolved by facility management. 1.3.1 Risk and Risk Tolerance The overall risk of a loss scenario is often represented as a function of the consequence Severity and Likelihood , where Severity is “How bad could it be?” In PHAs, severity generally refers to the unmitigated consequences – the maximum credible loss if none of the safeguards worked. Likelihood is “How often might the event occur?” Likelihood may refer to either the frequency or probability of an event. However, in PHA risk calculations, the likelihood generally needed is the mitigated frequenc y of a loss scenario (the frequency of a cause adjusted downward to take credit for applicable safeguards), and it is usually defined as the number of events per year. To promote consistent risk judgme nts, PHA procedures rely upon management defining and communicating it s risk tolerance to the team before the PHA is initiated. One simple risk tolerance criterion could be an organizational requirement to comply with RAGAGEPs. When PHA teams identify a nonconformance, they recommend corrective action. Many organizations go a step further and develop a risk matrix to communicate their tolerance for any loss scenario falling within a specific ra nge of severity and frequency. PHA teams apply this risk matrix to scenarios identifi ed during the PHA, and if a risk ranking is too high, recommendations are made to reduce risk to tolerable levels. An example risk matrix is shown in Figure 1-2, and each of its 25 cells is assigned to one of three risk levels – Lo w, Medium, or High. To be useful in practice, this example matrix woul d have to be accompanied by the organization’s definition of each s everity and frequency category, and those

108 INVESTIGATING PROCESS SAFETY INCIDENTS 6.7 SUM M ARY Following incident notification, the next step in the investigative process is assembling a team. It is important to select the appropriate size and composition of the team commensurate with the actual (and potential) incident severity, the complexity of th e incident, and other factors such as legal implications. For simple incidents, the team could be small. For more complex events, a larger team may be required, with spec ialists and incident support personnel included on an as-needed basis. Team leaders should have the necessary level of expertis e and experience for the particular incident and need to be effe ctive at delegating parts of the investigation to specific individuals, when required, while facilitating collaboration amongst the team members. Team members should be selected based on their skills, capability to work well with others, ability to be objective and unbiased, and to coordinate the wide ran ge of activities. Training is important to develop and maintain an understanding of the investigation management system and the specific activities to be perf ormed by various participants in the process. Training can include instru ction on the managem ent system, roles, and responsibilities as well as specific training on the tasks that each individual would be expected to perfor m in support of the investigation. Once a team is formed, the next step is for the team to develop a plan to gather data. The next chapters discuss considerations for witness interviews and evidence collection.

EMERGENCY MANAGEMENT 425 Communication is critical to an effective em ergency response. It involves the responders and many external stakeholders during a time when events may be confusing and moving quickly. Communication equipment warrants cons ideration as it should be available when needed and not be made ineffective by the emergency itself. Other Incidents This chapter began with a description of the West Fertilizer Explosion. Other incidents relevant to emergency response include the following. Grandcamp Freighter Explosion, Texas City, Texas, U.S., 1947 Sandoz Storehouse Fire, Basel, Switzerland, 1986 Gulf Oil Refinery Fire, Philadelphia, Pennsylvania, U.S., 1975 Deepwater Horizon Well Blowout, Gulf of Mexico, U.S., 2010 Exercises List 3 RBPS elements evident in the West Fertilizer explosion summarized at the beginning of this chapter. Describe their shortcomings as related to this accident. Considering the West Fertilizer explosion, what actions could have been taken to reduce the risk of this incident? Name three types of emergencies that a chemical facility should prepare for? What are key steps a company should take to prepare for an emergency? Why is it important to have an Emergency Response Plan? Name two leading indicator metrics that wo uld help you judge the ‘health’ of the emergency response preparedness in a given facility? When planning a drill what external stakeh olders might you invite to participate? Considering process safety emergencies, ho w might you decide what emergencies to include in the emergency response plan? References CCPS Glossary, “CCPS Process Safety Glossary ”, Center for Chemical Process Safety, https://www.aiche.org/ccps/resources/glossary . CCPS 2003, Guidelines for Analyzing and Managing the Security Vulnerabilities of Fixed Chemical Sites , Center for Chemical Process Safety , John Wiley & Sons, Hoboken, N.J. CCPS 2019, “Monograph Assessment of an d planning for natural hazards”, https://www.aiche.org/sites/default /files/html/536181/NaturalDisaster- CCPSmonograph.html. CCPS 2020, Monograph Risk Based Process Safety During Disruptive Times, https://www.aiche.org/sites/default/files/ht ml/544906/RBPS-during-COVID-19-and-Similar- Disruptive-times.html. CSB 2013, “West Fertilizer Company Fire and Explosion, Chemical Safety and Hazard Investigation Board”, Investigation Report, Re port 2013-02-I-TX, U.S. Chemical Safety and Hazard Investigation Board, Washington, D.C.

APPENDIX E – QUICK CHECKLIST FOR INVESTIGATORS 401 may have recent data at more fr equent intervals for a limited time period then start to average over a longer period. – Some evidence such as burn char patterns, surface fractures, or volatile chemicals spills can degrade as a result of weather conditions (rain, wind, or sunlight) Ensure that the investigation m eets regulatory requirements. For example, OSHA has specific requirements fo r the incident investigation teams. OSHA 1910.119 (m) (3) states: An incident investigation team shall be established and consist of at least one person knowledgeable in the process involved, including a contract employee if the incident involved work of the contractor, and other persons with appropriate knowledge and experience to thoroughly investigate and analyze the incident. Requirements in other jurisdictions may differ. Establish roles and expectations for the investigation team. Roles and expectations need to be defin ed early so that there are no misunderstandings. – What expectations do local management and corporate management have for the investigation team for timing, interim reports, final reports, and defining requirements for startup of units or equipment? – What resources are available and just as important, what resources are not available? Interviews need to be done promptly. Memories fade with time and are influenced by discussi ons with other witnesses. - Interviewing techniques are important. – Plan the interview. Do not do it haphazardly. – Interview one person at time and in a private comfortable setting. Use only one or two interviewers. – Set the interviewee at ease. One method is by asking questions about activities prior to the incident. – Be sensitive to the inte rviewee’s emotional state. – Do not express opinions. – Do not lead the interviewee. Ask questions that allow the interviewee to describe the in cident in their own words. Questions should be neutral, unbiased, and non-leading. – Do not interrupt the interviewee. – Use a plot plan to better understand « the location of interviewee « the location of people and activities the interviewee saw « movement of the interviewee

10 • Risk Based Process Safety Considerations 196 10.3.1 Effects of a weak commit to process safety pillar “It is simply not possible to talk about process safety without considering the impact of the safety culture and the leadership at a company…” J. A. Klein [21, p. 31] The transient operating mode incidents occurred, in part, to a weak leadership-driven process safety culture, the first RBPS element in the first RBPS pillar (Figure 10.1). The process safety competency element, also in pillar I, was within the top eight elements identified, as well. Strong leadership supports development of the different process safety competencies across their organization. These competencies are then applied with steadfast operational discipline. Some additional reflections on process safety culture and leadership, Element 1: A safety culture is the “normal way things are done…reflecting expected organizational values, beliefs, and behaviors…” [21, p. xxi]. This culture is committed to process safety, sets the priorities for safe process safety performance, and provides sufficient resources in personnel, technologies, and equipment to maintain safe process safety performance. It is the collective need to prevent major incidents and to do the right thing [99]. Leaders at all levels in an organization that has a committed, positive process safety culture will nurture everyone, identify the hazards, evaluate the risks, manage the processes, and learn from both good and bad experiences. A strong process safety culture and leadership instills safe operations during all of the transient operating modes, as well during normal, abnormal, and emergency operations. Process safety culture was noted in more transient operating mode incident reports after

References 371 [110] “National Commission on the BP Deepwa ter Horizon Oil Spill and Offshore Drilling, “Report to the President”,” 2011. [111] M. Christou and M. Konstantinidou, “S afety of offshore oil and gas operations: Lessons from past accident analysis.,” Pub lications Office of the European Union., Luxemburg, 2012. [112] Offshore Engineer, “The Future of Offshore Energy and Technolog,” onedigital, https://www.oedigit al.com, Undated. [113] E. Smith, R. Roels and S. King, “Guidanc e on learning from incidents, accidents and events.,” in Proceedings of Hazards 25 Conference , www.icheme.org, 2015. [114] F. K. Bitar, D. Chadwick-Jones, M. Nazaruk and C. Boodha i, “From individual behaviour to system weaknesses: The re-design of the Just Culture process in an international energy company. A case study.,” Journal of Loss Prevention in the Process Industries, vol. 55, pp. 267-282, 2018. [115] D. R. Edwards, “Tripod Be ta: Guidance on Using Tripod Beta in the Investigation and Analysis of Incidents, Accidents and Business Losses.,” Energy Institute., London, 2017. [116] United Kindgon Health and Safety Exec utive, “Human Failure Types,” HSE Books, https://www.hse.gov.uk, Undated. [117] S. Dekker, “Restorative Just Culture Checklist,” h ttp://sidneydekker.com, 2018. [118] A. Swain and a. H.E.Guttmann, Han dbook of Human Reliability Analysis with emphasis on nuclear power plant applications, https://www.nrc.gov: Nuclear Regulatory Commission, 1983. [119] J. Reason, “Managing the risks of orga nisational accidents,” Ashgate, Singapore, 1997. [120] Energy Institute: Hearts and Minds, “Making compliance easier (formerly Managing rule breaking),” Energy Institute, https://heartsandminds.energyinst.org, Undated. [121] U.S. Department of Energy, “Human Performance Improvement Handbook Volume 1: Concepts and Principles,” U.S. Department of Energy, https://www.standar ds.doe.gov, 2009. [122] M. Naderpour, J. Lu and G. Zhang, “The explosion at Institute: Modeling and analyzing the situation awareness factor.,” Accident Analysis & Prevention, no. 73, pp. 209-224., 2014a. [123] International Association for Oil and Ga s Producers (IOGP), “Introducing behavioural markers of non-technical skills in oil and gas operations - Report 503,” International Association for Oil and Gas Producers, https://www.iogp.org, 2018.

2 | 1 Introduction Working on behalf of the chemical engineering profession, the American Institute of Chemical Engineers (AIChE) began to share findings and recommendations from process safety incidents via the Ammonia Plant Safety (Williams 2005) and Loss Prevention Symposia in the 1950s and 1960s (Freeman 2016). AIChE’s Design Institute for Emergency Relief Systems (DIERS) began publishing guidelines for multiphase relief systems in the 1970s (AIChE 2020a). Until the mid-1980s, institutional lessons learned came in the form of technology innovations, new or revised standards and codes, or back-up systems. This began to change with the formation of AIChE’s Center for Chemical Process Safety (CCPS) in 1985. CCPS began the process of formally leveraging incident findings and successful practices into “Guidelines” and “Concepts” (Berger 2009). In 1988 CCPS codified the first Process Safety Management System (PSMS). The CCPS 12 Elements (CCPS 1989) provided the first organized common framework to comprehensively manage all the standards, technologies, and practices needed to control a company’s process safety hazards. The original framework has evolved into today’s 20 elements of Risk Based Process Safety (RBPS), which are organized in four pillars: Commit to Process Safety, Understand Hazards and Risk, Manage Risk, and Learn from Experience (CCPS 2007). Regulations around the world also began to emerge in the 1980s, most notably the Sevesso Directive in the European Union, the Process Safety Management (PSM) regulation in the USA, and the Control of Major Accident Hazards (COMAH) in the UK. Most national and regional process safety regulations are based on one or a combination of these original regulations. Unfortunately, incidents continue to happen despite 200 years of continuous development of technology, standards, publications, and management systems. They continue to happen despite the great number of recommendations from incident investigations conducted by every operating company in this industry. And nearly every incident that occurs in an industry, a company, or a plant has root causes that resemble the causes of previous incidents. 1.1 The Focus of this Book CCPS (CCPS 2019a) and others have written guidelines addressing the general process of incident investigations. These books focus heavily on the process of investigation, the determination of root causes and causal factors, and the

73 | 6.1 Focus standard, or policy. Therefore, Tier 3 metrics clearly identify areas that require improvement. Examples of Tier 3 metrics include: • trips of Safety Instrumented Systems (SISs) • activations of interlocks • manual activation of emergency shut-down • exceedance of safe operating limits • any leak or release not included in Tier 1 or Tier 2. Near-misses covered by Tier 3 metrics should be investigated as if an incident had occurred. Audit Findings Process safety audits for regulatory compliance are generally conducted every three years; for risk management purposes, audit frequency may range from one to five years, depending on risk. Audits conducted by leading companies typically go beyond compliance and may include (CCPS 2011a): • more frequent audits of processes that have higher risk and/or higher potential consequences • auditing processes not covered by regulations that nonetheless have appreciable risk • benchmarking against industry-leading practices. All audits capture deviations from the PSMS as findings; however, some will also capture recommendations for improving the PSMS, standards, and policies. Findings and recommendations can inform overall improvement efforts, especially those that recur through multiple sites. Although repeat findings often reflect deficiencies in how site management carries out process safety responsibilities, they may also reflect an inherent gap in site or corporate knowledge that could be informed by examining internal and external incidents. Internal Incident and Near-Miss Investigations Investigation of internal incidents and high potential near-misses should automatically trigger a review of both relevant external incidents and similar internal incidents. While this review could occur separately from the investigation, in many cases it can be beneficial to review other incidents

80 Although the focus is on reducing occupational (and household) exposure to hazardous chemicals as well as environmental protection, many of these tools are applicable to chemical processes and process safety as well. The Ontario Toxics Reduction Program is intended to promote green chemistry and engineering and has de veloped a valuable reference tool for evaluating chemical substituti on using the framework shown in Figure 4.1. This tool is intended to be general for use in industries and companies of various sizes. It should be noted that several processes for assessing the use of alternative chemicals have been deve loped, however there is currently no standard method established. De fining metrics for selection of the most attractive alternative is a major challenge; this is particularly important given the tradeoffs that ar e often necessary when substituting a hazardous chemical with a safer alternative. (Ref 4.23 Ontario) The U.S. Environmental Protection Agency (EPA) has also developed a process for evaluating safer chemical alternatives, known as Design for the Environment (DfE) Alte rnatives Assessments. (R ef 4.35 U.S. EPA) This effort is focused primarily on product safety and consists of the following steps: 1.Determine the feasibility of an alternatives assessment 2.Collect information on chemical alternatives 3.Convene stakeholders 4.Identify viable alternatives 5.Conduct the hazard assessment 6.Apply economic and life cycle context 7.Apply the results in decision -making for safer chemical substitutes

158 | 12 REAL Model Scenario: Overfilling Bayamón, PR, USA, 2009 An overfill of a 19,000 M3 (5 million-gallon) atmospheric storage tank with gasoline caused a vapor cloud that ignited causing multiple tank explosions and fires. 17 of 48 tanks were burned and it took 3 days before the fire was under control. 3 people were injured. Pamela had also mentioned her concern over climate change, so Alexandre did another search on process safety incidents caused by natural disasters. He discovered a CCPS Beacon on this topic that covered wildfires near Fort McMurray, Alberta, Canada; Hurricane Harvey along the Texas, USA, coast; and the earthquake and subsequent tsunami in Fukushima, Japan (Figure 12.1). Alexandre decided that of the three, the incident caused by Hurricane Harvey was the most relevant. He searched the CSB database and found an incident investigation in Crosby, TX. Figure 12.1. An issue of the CCPS Beacon focused on natural disasters. See Appendix index entry C15

138 GUIDELINES FOR MANAGING ABNORMAL SITUATIONS Operator training simulators can be an important tool in reinforcing a mental model and building "muscl e memory" responses to key fast acting issues. Emerging technologies are allowing drills to be developed with a more immersive experience for staff including virtual reality “gaming” type simulations. These have the added benefit of simulating some of the stress and key time-related aspects of incidents that are not present in most paper-based drills Another focus of training is for fiel d operators to stay aware of their surroundings by using their natural senses such as sight, hearing, or smell. This can be important in detecting and recognizing abnormal conditions related to field equipment. Some examples are: Equipment vibrations Cavitating pumps Missing plugs or caps Unexpected ice or frost on equipment or piping Smell of burning rubber or insulation Squeaking belts Unusual odors Open electrical boxes or fittings Corrosion Leaks and drips Layers of combustible dust Missing or damaged pipe hangers or supports These are just a few examples of abno rmal situations that should be investigated quickly. Other examples can be created and customized for the specific process that personne l are being trained to operate. Note: To read more about training and knowledge, please refer to Chapter 4, Section 4.3.

220 | Appendix: Index of Publicly Evaluated Incidents NPO Association for the Study of Failure (ASF) of Japan Incident Database (Continued) (For incident reports J1–J163: see www.shippai.org/fkd/en/lisen/cat102.html) Code Investigation J33 Leakage and Fire Damage Caused Due to Stress Corrosion Cracking of a Heat Exchanger at a Glycerin Concentration Plant (1996) J34 Explosion Due to Sparks of an Electric Grinder During Repairing a Wastewater Treatment Vessel with Neutralization (1996) J35 Explosion Caused Due to a Catalytic Effect of Contaminant in the Reactor at a Resin Intermediate Manufacturing Plant (1996) J36 Fire of a High-Pressure Thermal Gravimeter at a University (1996) J37 Explosion and Fire at a Manufacturing Plant for Manufacturing a RIM Raw Material Liquid (1996) J38 Explosion in the Workroom to Fill Explosives with Press (1996) J39 Fire of Leaked Hydrogen Due to Misuse of a Gasket at a Solvent Manufacturing Plant (1996) J40 Explosion of Benzene Used as Washing Oil During Cleaning a Fuel Oil Tank (1996) J41 Explosion in a Storage Tank Caused Due to Decomposition of a Polymerization Catalyst (1995) J42 Explosion of a Methanol Recovery Residue at an Organic Peroxide Manufacturing Plant (1995) J43 Explosion and Fire of Tetrahydrofuran During Air-Pressure Transfer to a Tank from a Drum Can (1995) J44 Ignition of Leaked Gas Caused Due to a Runaway Reaction During a Power Failure at an Expanded Polystyrene Manufacturing Plant (1995) J45 Explosion Caused Due to a Change of a Draw-Off Method of Coolant in a Jacket of a Reactor from Air Pressure to Steam Pressure at a Medical Intermediate Manufacturing Plant (1995) J46 Fire Caused Due to Back Flow of Dimethylamine During Repairing Work of Piping from a Reactor to a Scrubber (1995) J47 Explosion During Taking Out Used Desiccant from a Drum Can (1995) J48 Rupture, Explosion, and Fire Caused Due to Pressure Rise During Cleaning Operation for a Distillation Kettle (1995) J49 Explosion During Cleaning by Brushing a Tank for Dangerous Materials at a Paint Manufacturing Factory (1995)

10 • Risk Based Process Safety Considerations 198 efforts in Pillar II, the elements in Pillar III could not be designed, implemented, or sustained effecti vely. Thus, these risk management programs and systems were inad equate when responding to the transient operating mode shut-downs caused abnormal or emergency operations. Since administrative controls are less effective than engineering controls, it should come as no surprise that inadequate Operating Procedures during the transient op erating mode rank as the weakest of all the elements (refer to the hierarchy of controls [21, p. Figure 3.6]; and Figure 10.2). Often, this was due, in part , to unwritten, quick, and inadequately reviewed decisions and actions taking during start-up upsets. The start-up procedures simp ly cannot address all unknowns, with the responses (or lack ther eof) sometimes proving fatal. The established normal start-up and shut -down procedures should guide decisions for safely addressing unexpected issues as they arise (see the Appendix for more unexpected shut-down-related guidance). Unexpected start-up issues occurri ng after a project, maintenance, or facility shutdown were often caused by an unexpected start-up condition of the equipment. At other times, abnormal operation upsets may have caused an unusual equipment idling condition or an emergency shut-down Again, incide nts during start-ups occurred because the equipment was not in its normal start-up condition. The lack of understanding the state of th e equipment could have been due, in part, to inadequate handovers . Refer to Chapter 3 for a more detailed discussion on preventing inadequate procedural handovers. The second most noted Pillar III el ement contributing to transient operating mode incidents was the Asset Integrity and Reliability element. If the equipment failed catastrophically during normal operations, operations had to implement an immediate emergency shut-down. Also, when the safeguardi ng equipment failed to operate as expected during the abnormal or emergency operation, the

DEVELOPING EFFECTIVE RECOM M ENDATIONS 289 In summary, there are various te chniques available to help the investigation team identify the layers of protection (barriers) that have failed. The recommendations should address the root causes of the failure of the management systems that have been in sufficiently robust to maintain the barriers effectively. 12.3.3 Commendation/ Disciplinary Action When an investigation reveals an empl oyee action worthy of commendation, the incident investig ation team should acknowl edge the individuals directly, but not name them in the formal investigation report. Cool and rational actions in the midst of an emergency often limit the consequences of an incident, but naming an individual pu blically could bring undesired attention or could be culturally inappropriate. Disciplinary action is outside the scope of the investigation team’s remit. Even the perceived threat of disciplinar y action has a detrimental effect on an investigation and may discourage cooperation during interviews. In the event that disciplinary issues should be considered, this should be part of a separate management process, involving a different team and in line with the company’s internal disciplinary procedure. 12.3.4 The “Further Action Required” Recommendation Another special case is the recommenda tion for further work; for example, the investigation team may recommend re-evaluation of an existing safeguard, evaluation of a new safeguar d, or consideration of an inherently safer design. This does not mean that the investigation team has failed to complete its task. It is common for an investigation team to generate a recommendation to confirm whether an existing physical system or administrative measure (such as written procedures or training program) provides adequate protection . It could be that spec ialists, who had not been available to the investigation team, are needed to conduct further work or engage additional expertise outside the main investigation process. It is not appropriate for an investigation team to attempt to engineer a solution in an area in which they are not qualified. In these instances, the team should specify, for example, what action is ne cessary if the safeguard is found to be inadequate. If the team only specifies a vague action such as “review the start-up procedu re,” then the implemented acti on may or may not meet the team’s intentions.

35 3 EVALUATING THE PRIOR PHA The primary objective of a PHA revalidat ion is to ensure the PHA accurately describes and evaluates current risks. To accomplish that, the team should consider the current hazards, current causes, current consequences, and current risk controls. The purpose of eva luating the prior PHA is to assess how closely that analysis matches the current reality, which is the key determinant of the best revalidation approach. Presumin g regulations or policies do not dictate the use of the Redo approach as discussed in Chapter 2, any decision to set aside previous efforts and completely Redo the prior PHA with a clean sheet of paper and a new team, risks losing all the effort and experience that were embodied therein. For example, tec hnology-specific hazards that had been learned by the operating company might have prompt ed some of the loss scenarios documented in the prior PHA. That potent ial loss must be balanced against the possibility of relying too much on outdated information that is not relevant to the unit’s current risk profile along with any errors or omissions of previous teams. Thus, given the spectrum of revalidation approaches ranging from a complete Redo to a focused Update , several factors should be considered when selecting an optimum approach. The domina nt considerations generally fall into the following categories: • Changes in PHA requirements • Methodology used for the prior PHA • Inputs for the prior PHA • Overall quality and completeness of the prior PHA • Changes made to the process since the prior PHA • Consistency of PHA risk judgments with operating experience As illustrated in Figure 3-1, changes in PHA requirements were addressed in Chapter 2. This chapter focuses on evaluating the acceptability of the prior PHA. Evaluating what has transpired since th e last PHA and assessing the consistency of risk judgments with operating experien ce (including learnings from incidents) are addressed in Chapter 4.

58 Guidelines for Revalidating a Process Hazard Analysis • Claiming multiple safeguards that are not independent. Claimed safeguards provide no significan t risk reduction unless they are largely independent of each other [37]. However, when reviewing the prior PHA, it is sufficient to recognize alleged safeguard(s) that could be incapacitated by the same failure or error would typically fail the independence criterion. The choice of Redoing or Updating the PHA depends on the prevalence of this error in the prior team’s risk judgments. • True statements leading to illogical conclusions. Consider the following true statements: (1) High flow in the outlet line from a supply tank could cause a low level in the supply tank. (2) Low level in the supply tank could cause low flow in its outlet line. Possible illogical conclusions: (1) High flow in the outlet line causes low flow in the outlet line, (2) Opening the level control valve too much causes low flow in the outlet line. If the prior PHA contains circular reasoning of that sort, the Redo approach may be necessary. 3.2.4 Failure to Document Hazards PHAs are generally required to identify the hazards of the process, and those are usually evident in the consequence descriptions. In some cases, the consequences of a deviation, question, or failure mode are simply listed as “release of process material.” The prior PHA team may have known exactly what that meant. Subsequent PHA revalidation teams will likely struggle to leap from this generic statement to how people may be harmed (or what other undesirable things might happen). Process hazards are properties of the materials and Example – Safeguard Independence If the prior PHA lists a safeguard “TE-123 with high temperature alarm” and a second safeguard “TE-123 with high temperature shutdown,” these two safeguards are not independent. If TE-123 fails, both safeguards will fail. Causes are Not Hazards It is impossible for a PHA to list every conceivable cause related to a particular hazard. For example, a PHA might identify “pump stopping” as the cause for a low flow hazard and list “spare pump” as the safeguard. The prior PHA did not need to identify and evaluate every possible reason for the pump stopping (e.g., broken shaft, shorted motor, seized bearing) if the listed safeguard (spare pump) is equally effective for all those cases.

xxx | Executive Summary safety culture may help strengthen other parts of the overall culture. Leaders at any level of the organization will benefit from the guidance provided in this book. Senior executives will likely be drawn most to the first 3 chapters and the beginning of chapter 5, while the remainder of the book contains more detailed guidance useful at the implementation level. However, all readers will find useful information throughout the book. Afte r definin g proce ss safe ty culture , this book outlin es 10 core principle s of proce ss safe ty culture: Establish an Imperative for Process Safety Provide Strong Leadership Foster M utual Trust Ensure Open and Frank Comm unications Maintain a Sense of Vulnerability Understand and Act Upon Hazards/Risks Empower Individuals to Successfully Fulfill their Process Safety Responsibilities Defer to Expertise Combat the Norm alization of Deviance Learn to Assess and Advance the Culture The book then shows how these core principles strengthen process safety m anagement systems (PSM Ss), which implemented together can lead to success. The role of process safety culture in m etrics, compensation, and other related activities is addressed. Lastly, the book discusses how to make process safety culture sustainable. Appendices include more detailed descriptions of several concepts presented in the book, such as organizational culture, hum an behavior, and high reliability organizations, along with • • • • • • • • • •

A.3 Index of Publicly Evaluated Incidents | 201 Section 1. RBPS Elements (Continued) Asset Integrity and Reliability—Primary Findings (Continued) J121, J122, J128, J130, J132, J144, J145, J146, J148, J153, J156, J160, J166, J168, J169, J170, J171, J172, J173, J175, J176, J179, J182, J184, J186, J193, J194, J195, J199, J202, J203, J205, J207, J210, J212, J214, J215, J218, J226, J230, J231, J232, J233, J238, J239, J241, J249, J250, J254, J256, J265 S10, S12, S13, S16, S17 Asset Integrity and Reliability—Secondary Findings A2, A5, A11 C2, C9, C31, C32, C39, C40, C41, C47, C64, C73 HA1, HA2, HA5, HA8 J2, J57, J78, J88, J99, J104, J110, J112, J113, J125, J131, J134, J139, J167, J181, J183, J189, J196, J209, J213, J227, J235, J240, J255, J258, J262, J269 S9 Contractor Management—Primary Findings C22, C22, C57, C71 J46, J71, J75, J189 Contractor Management—Secondary Findings A1 C11, C46, C58, C70, C77 J47, J76, J172, J179, J190, J194, J202 S3 Training and Performance Assurance—Primary Findings C10, C43, C45, C47, C60, C77 J37, J42, J46, J61, J70, J76, J83, J85, J97, J108, J120, J121, J129, J131, J133, J134, J143, J145, J149, J152, J174, J180, J181, J185, J187, J190, J195, J196, J197, J206, J216, J221, J244, J247, J248, J254, J256, J261, J270, J271 S7, S15 Training and Performance Assurance—Secondary Findings A1 C11, C12, C19, C20, C22, C23, C24, C36, C38, C44, C61, C69, C70 D9, D20 J19, J20, J21, J25, J26, J28, J38, J47, J48, J49, J50, J51, J52, J55, J58, J63, J65, J122, J123, J125, J128, J132, J148, J154, J157, J162, J171, J177, J178, J186, J208, J209, J210, J211, J217, J236, J257 S1, S14

Conducting PHA Revalidation Meetings 143 Once the study leader resolves any issu es that may arise (e.g., information access and exchange with remote team participants), the revalidation can commence. However, before the actual meetings begin, some companies allocate time and resources for teams to tour the unit together. Such a tour would be of greatest value to team me mbers who do not routinely work in the unit and who did not have an opportunity to visit the unit before the kickoff meeting. If not explicitly budgeted and planned, such a tour takes time away from the team’s scheduled meetings. Practi cal considerations of access permits, escorts, PPE, weather, noise, disruption of other unit activities, etc. must also be considered. Thus, the team must make a judgment as to whether the cost of a group tour is worth the benefit for the team as a whole, whether other arrangements should be made for partic ular individuals, or whether tours will be conducted as needed during the revalidation sessions. Sometimes a virtual tour is possible with three-dimensiona l photography or animation. Regardless, a tour will help ensure a common fam iliarity with the process and a common frame of reference with respect to facilit y layout and equipment location. It will also help identify and correct any misconceptions that team members might have as to the physical scope of the review. Many revalidation teams have found the benefits of a field unit tour are well worth the time invested. 7.2.3 Meeting Productivity Meeting time is one of the team’s most precious resources, and one of the facilitator’s main goals is to optimize th e use of that time. For example, there is a significant amount of preparatory work involved in locating, collecting, and preparing information for use in the r evalidation sessions. As discussed in Chapter 6, it is incumbent that the study leader, or designee(s), do the needed preparatory work ahead of the meeting so that this information can be accessed and used as efficiently as possible. Piping and instrumentation diagram (P&ID) revisions can be cross-referenced to PHA nodes, and to MOC and pre-startup safety review (PSSR) documentation, befo re the sessions begin. Sessions will be more productive, and less onerous for pa rticipants, if documents are organized so they can be found quickly and efficiently when they are needed. A unique issue in a revalidation meeting, particularly one using the Update approach, is the use of the previous PHA. The monotony of the leader simply reading it to the team or flashing it up on a screen can easily lead to the team becoming disengaged and result in an incomplete or otherwise inaccurate analysis. On the other hand, forcing the team to regenerate accurate information already in the existing do cumentation is of little value. The facilitator’s challenge is to keep the te am mentally engaged while maximizing the team’s efficiency. Often this can be accomplished with simple changes in

229 consider risk/risk tradeoffs that may occur. They also must be weighed against the benefits of the operation, as well as alternatives, such as more traditional security measures, to manage exposures, and whether overall risk is sufficiently managed. 9.9 REFERENCES 9.1 American Chemistry Council (ACC), Responsible Care , (www.responsiblecaretoolkit.com/security.asp) 9.2 American National Standard s Institute/American Petroleum Institute (ANSI/API); ANSI/API Standard 780 “Security Risk Assessment Methodology for the Petroleum and Petrochemical Industries, First Edition; 2013 9.3 American Petroleum Institute (API), new.api.org/policy/otherissues/upload/SecurityGuideEd3.pdf 9.4 Center for Chemical Process Safety (CCPS), Guidelines for Managing and Analyzing the Security Vulnerabilities of Fixed Chemical Sites, American Institute of Chemical Engineers, 2002. 9.5 Chemical Facility Anti-Terro rism Standards (CFATS), 6 CFR Part 27, Interim Final Rule published April 9, 2007 (72 Fed. Reg. 17696). 9.6 Kletz, T.A. (1998). Process Plants: A Handbook for Inherently Safer Design. Philadelphia, PA: Taylor & Francis. 9.7 Organization for the Prohibition of Chemical Weapons (OPCW), Convention on the Pr ohibition of The Development, Production, Stockpiling and Use of Chemical Weapons and On Their Destruction, (www.opcw.org/) 9.8 United States Department of Homeland Security (DHS), Energy Sector Specific Plan , 2010.

Table A.2 IST Checklist Minimize Questions 2 MINIMIZE Questions: 2.1 Inventory Reduction 2.1.1 Can hazardous raw materials inventory be reduced? • Just-in-time deliveries based on production needs • Supplier management including strategic alliance • On-site generation of hazardous material (including in situ) from less hazardous raw materials • Hazardous raw material inventory management system based on production forecast 2.1.2 Can (hazardous) in-process storage and inventory be reduced? • Direct coupling of process elements • Eliminating or reducing size of in-process storage • Designing process equipment involving hazardous material with the smallest feasible invento ry (see also Section 2.2) 2.1.3 Can hazardous finished product inventory be reduced? • Improving production scheduling/sales forecasting • Improving communication with transporters/material handlers • Hazardous finished product inventory management system based on sales forecast 2.2 Process Intensification Considerations 2.2.1 Can alternate equipment with reduced hazardous material inventory requirement be used? • Centrifugal extractors in place of extraction columns • Flash dryers in place of tray dryers • Continuous reactors in place of batch • Plug flow or loop reactors in place of continuous stirred tank reactors • Continuous in-line mixers (e.g., static mixer) in place of mixing vessels or reactors • Intensive mixers to minimize size of mixing vessel of reactor • High heat-transfer reactors (e.g., microreactor, HEX reactor) • Spinning-disk reactor (especially for high heat-flux or viscous liquids) 445

4.3 Maintenance of Barriers / Barrier Integrity | 47 In Jaipur, Rajasthan, India, in 2009, two separate human barrier failures led to 11 people dead and more than 150 injured, many offsite. An operator was tasked to change the position of a spectacle blind to direct the transfer of gasoline to a desired location. The operator failed to block in the blind using two valves provided for this purpose. Upon swinging the blind, gasoline spewed into the dike. Unfortunately, the dike drain valve had been left open after a manual draining of rainwater. Gasoline vapors ignited and exploded and, unconfined to the dike, the fire spread widely. The probability of human barrier failure increases with fatigue and stress. Sleep deprivation and distraction due to personal issues can lower a worker’s decision-making ability and reduce reaction time. These situations are hard to guard against, so facility design should include one or more automated barriers in case human reactions are not correct or fast enough. Ensuring that such barriers are in operational order is paramount. This was apparent in the case of the 2010 Macondo oil well incident (CSB 2016a). Located 50 miles off the coast of Louisiana in the Gulf of Mexico, the oil well had multiple barriers that were supposed to prevent uncontrolled release of oil and gas. At the time, the crew onboard the rig was working on temporary well-abandonment activities. However, several barriers failed: • A cement plug that had been put in earlier to keep the oil and gas below the sea floor was not installed effectively due to time pressure. The data from integrity testing of this barrier was misinterpreted. The cement barrier was incorrectly deemed effective. • When the drilling mud was removed, hydrocarbons flowed past the blowout preventer (BOP) valve. There was no human detection (and thus no intervention) for nearly an hour. Once the crew did detect the release, they manually closed the BOP. By that time, however, the hydrocarbons had reached the surface and found an ignition source. • An automated emergency response system worked as designed to shear the drill pipe and effectively seal the well. Pressure conditions in the well caused the drill pipe to buckle, however, rendering this barrier ineffective. The failure of all these barriers resulted in 11 fatalities, 17 injuries, approximately 4 million barrels of hydrocarbons released into the environment, and an economic impact on the company of at least USD 65 billion (Maritime Executive 2018). See Appendix index entry C46 See Appendix index entry S2

5 • Facility Shutdowns 88 hydrocarbons into the process area. The hydrocarbons pooled and quickly formed a vapor cloud that ignited an d exploded. Incident impact : There were fifteen fatalities, 180 injuries, a community shelter -in-place order affecting 43,000 people, and significant property damage, including residences alm ost 1.2 km (0.75 miles) away. Risk management system weaknesses : LL1) At the time of the incident, leadership at every level had emphasized personal safety bu t had not emphasized, reviewed, audited, or measured its process safety programs as it managed its process hazards and risks. The leadersh ip at the Texas City refinery at the time of the incident had not established the essential positive, trusting, and open environment with its workforce, as well. In addition, at the time of the incident, leadership had not provided sufficient training or resources (including co ntractors), especially with process safety competencies and capabilities; had allowed high overtime rates for operations and maintenance personnel; ensured that personnel effectively follow the safe work pr ocedures (i.e., permit to work; job safety analyses) , and did not ensure that proper start- up procedures were in place —and being used—before operations resumed. Relevant RBPS Elements Process Safety Culture Process Safety Competency Workforce Involvement Operating Procedures Safe Work Practices Contractor Management Training and Performance Assurance Operational Readiness Measurement and Metrics

30 PROCESS SAFETY IN UPSTREAM OIL & GAS identification study or PHA (CCPS, 2008a), and the need for any additional barriers for specific incident sequences may be further defined by a risk assessment, usually LOPA (CCPS, 2011) or by a full QRA (CCPS, 1999). Some confusion also exists regarding the term ‘safeguard’. It is us ed as a collective term for any measure reducing risk whether or not it meets the criterion for an IPL. A simple phrase to remember this is “not all safeguards ar e IPLs, but all IPLs are safeguards”. Some formal definitions that assist in understanding the use of the term “barrier” are provided as follows. The CCPS Glossary defines barrier and Independent Protection Layer as: Barrier: Anything used to control, prevent, or impede (interrupt) energy flows. Includes engineering (physical, equipment design) and administrative (procedure s and work processes). Independent protection layer (IPL): “A device, system, or action that is capable of preventing a po stulated accident sequence from proceeding to a defined, undesirable endpoint. An IPL is independent of the event that initiated the accident sequence and independent of any other IPLs. IPLs are normally identified during layer of protection analyses.” The CCPS (2018c) Bow Ties in Risk Management defines barrier as follows. A control measure or grouping of controls that on its own can prevent a threat developing into a top event (prevention barrier) or can mitigate the consequences of a top event once it has occurred (mitigation barrier). A barrier must be effective, independent, and auditable. The term barrier is also used widely in upstream, but with different interpretations as described in the following definitions. API 100-1 (2015b) for onshore wells provides the following barrier definition. “a component or practice that contributes to the total system reliability by preventing liquid or gas flow when properly installed”. (Note: in context this refers to unplanned flows). NORSOK D-010 provides the following definitions. Well barrier: Envelope of one or several dependent barrier elements preventing fluids or gases from flowing unintentionally from the formation into another formation or to surface. Well barrier element: Object that alon e cannot prevent flow from one side to the other side of itself. IOGP 415 defines barrier as follows. A risk control that seeks to prevent unintended events from occurring or prevent escalation of events into incidents with harmful consequences. Hardware and human barriers are put in place to prevent a specific threat

EMERGENCY MANAGEMENT 419 An emergency can cause harm to people or the environment or damage to property. It can also prompt a facility’s license to operate wi thin the community to be questioned. Effective emergency management can save lives, protect property and the environment, and reassure stakeholders that a facility is well managed. Em ergency management is part of the U.S. OSHA PSM and U.S. EPA RMP regulations and are covered by regulations in other countries. Managing human performance is challengin g during an emergency and having a well- practiced plan can improve this performance. Emergency management strives to protect workers, neighbors, and emergency responders. It also focuses on communication so that both those involved in the response and external to it are aware of what is happening. Many types of emergencies can impact process safety either directly, such as a fire, or indirectly by lessening the ability of those support ing the safe operation of a facility to do so. Process safety incident (refer to Chapter 9) Natural disasters such as floods, hurricanes, and tornedos (as seen in Figure 20.6) Incident at a neighboring property Pandemic Intentional attack or sabotage Emergency management encompasses activities that occur at the facility and in the community before, during, and after an emergency including the following. 1. Emergency response planning for potential emergencies 2. Emergency response resources to execute the plan 3. Emergency response exercises for practici ng and continuously improving the plan 4. Emergency response training or informin g employees, contractors, neighbors, and local authorities on what to do, how they will be notified, and how to report an emergency 5. Emergency response communications with st akeholders in the event an incident does occur

8 INVESTIGATING PROCESS SAFETY INCIDENTS Chapter 6—Building and Leading an Incident Investigation Team Personnel with proper training, skills, and experience are critical to the successful outcome of an incident investigation. This chapter describes team composition as a function of incident type, complexity, and severity, and includes suggested training topics. It also provides team leaders with a high- level overview of the basic team ac tivities typically required in the course of conducting an investigation. Chapter 7 - W itness M anagement This chapter discusses techniques fo r identifying witnes ses and effective interviewing techniques designed to obtain reliable information from them. Witnesses often hold the most intimate knowledge of condit ions at the time of the incident, actions taken pre-inci dent and post-incident, process design and operations, etc. Effective manageme nt of witnesses is a crucial element of the investigation process. Issues related to witness interactions and interviewing techniques are covered in detail. Chapter 8—Evidence Identificati on, Collection, and M anagement Facts are the fuel that an investigation needs to reach a successful conclusion. This chapter addresses the methods an d practical considerations of data- gathering and archiving ac tivities. It describes plan development; priority establishment; different types and sour ces of data; data-gathering tools, techniques, and preservation; docume ntation requiremen ts; photography and video techniques; sugg ested supplies; etc. Chapter 9 - Evidence Analysis and Causal Factor Determination This chapter provides practical gu idelines for analyzing evidence, proving/disproving hypotheses, and devel oping causal factors. The use of a scientific methodology to sort out fact s from collected data is explained, and techniques are offered for use during th is iterative and overlapping process. Identifying causal factors is an inte rmediate step towards determining root causes, and implementing recommendati ons based on root causes should inherently address the causal factors as well. Chapter 10—Determining Root Causes—Structured Approaches This chapter addresses methods and tools used successfully to identify multiple root causes. Process safety incidents are almost always the result of more than one root cause. This chapte r provides a structured approach for determining root causes. It details some powerful, widely used and proven tools and techniques available to incident investigation teams, including

17. Error management in task pla nning, preparation and control 207 Table 17-4: An isolation incident: relying on experience An example of not following an isolation procedure What happened? A pig trap was not depressurized before attempting to remove it from a pipe. When the trap door was opened, a sudden release of high-pressure gas caused the door to be blown 30 feet (10 mete rs) across the deck, through two handrails and overboard. A worker was injured (the Injured Person - IP) and treated for facial lacerations. Why did it happen? The worker did NOT: • Open depressurization valves to confirm there was “zero energy”. • Check the local pressure indicator (PI). The worker was relying on years of experience instead of following a safety critical procedure. The worker was not positioned sufficiently outside of the line of fire and got injured but avoided fatality. What did they learn? Check the pressure indicator and ensure ‘zero energy’ before opening a pig trap door. Breaking containment is a high-risk activity. When opening a pig trap door, ensure that staff members are positioned outside of the direct ‘line of fire’ in order to prevent fatality. It is vital to understand and follow safety critical procedures. Compliance with the work manage ment system is mandatory. (adapted from Energy Institute toolbox [69])

198 | 5 Aligning Culture with PSMS Elements MOC process, and updating the procedures should be a requirement to close out all M OCs. Sometimes procedures may be changed for other reasons, such as new process knowledge or improvement in clarity. Such changes should also be tracked and controlled. To verify consistent use of procedures, they should be reviewed periodically for accuracy and to verify that they are being followed. Some regulations and standards specify that procedures be reviewed every 1-3 years. In the absence of such requirements, facilities should set review cycle based on risk; review higher risk processes more frequently. These m easures help prevent normalization of deviance . Having operators or m aintenance personnel review and com ment on the procedures they use is another way to reinforce PS culture. When m anagement acknowledges the value of these reviews, it fosters open communication and builds mutual trust. CCPS (Ref 5.6) describes several ways of formatting procedures, each with advantages and disadvantages. Com panies should select the style(s) that work best for them . Once selected, operating procedures should follow the standard style. Operators should be closely involved, both in selecting the standard style and in writing. This helps ensure that procedures are understood and followed. In appropriate cases, experienced operators could write operating procedures to be checked by engineers. Experienced operators may com e to know procedures so well they’ve m emorized them. In good cultures, operators follow the procedures even if they know them very well. There are two reasons: If operators read procedures as they operate the process, they are less likely to normalize deviance or have an error of mem ory. When procedures change, memory will no longer be correct. • •

205 8.14 Center for Chemical Process Safety (CCPS 2008). Guidelines for Chemical Transpor tation Risk Analysis, American Institute of Chemical Engineers, 2008. 8.15 Center for Chemical Proce ss Safety (CCPS), Guidelines for Combustible Dust Analysis, American Institute of Chemical Engineers, 2017. 8.16 Center for Chemical Pr ocess Safety, Guidelines for Engineering Design for Process Safety , American Institute of Chemical Engineers, 2012. 8.17 Center for Chemical Process Safety (CCPS 2010), Guidelines for Evaluating the Characteristics of Vapor Cloud Explosions, Flash Fires, and BLEVEs, 2nd Ed., American Institute of Chemical Engineers, 2010. 8.18 Center for Chemical Process Safety (CCPS 1996b). Guidelines for Evaluating Proce ss Plant Buildings for External Explosions and Fires, American Inst itute of Chemical Engineers, 1996. 8.19 Center for Chemical Process Safety (CCPS 2008a), Guidelines for Hazard Evaluation Procedures, 3rd Ed., American Institute of Chemical Engineers, 2008. 8.20 Center for Chemical Proce ss Safety (CCPS 2003a), Essential Practices for Managing Chemical Reac tivity Hazards, American Institute of Chemical Engineers, 2003. 8.21 Center for Chemical Process Safety (CCPS 2016a). Guidelines for Safe Automation of Chemical Processes, American Institute of Chemical Engineers, 2016. 8.22 Center for Chemical Process Safety (CCPS 2004), Guidelines for Safe Handling of Po wders and Bulk Solids, American Institute of Chemical Engineers, 2004. 8.23 Center for Chemical Process Safety (CCPS 2016b), Guidelines for Safe and Reliable In strumented Protective Systems. American Institute of Chemical Engineers, 2016.

170 exist at this stage. And, many opportunities exist to apply the simplification strategy during detailed design. Plant designs should be based on a risk assessment that considers the process and the site in detail, as well as all of the principles of inherently safer operation. Earlier de cisions may limit the options during the detailed design stage, but inhere ntly safer principles can still be applied. The detailed design step is the last step at which changes can be made at moderate cost because most of the equipment is purchased after this detailed design is approved. Once the equipment has been purchased and fabricated, and the fac ility is constructed, the cost of modification increases substantially. 8.5.1 Process Design Basis To reduce the potential for large releases of hazardous materials: If not already accomplished, mi nimize or eliminate the in- process inventory of hazardous materials to the lowest amount necessary consistent with the minimum operational needs. This includes inventory in the proce ss equipment, piping, as well as in vessels and tanks. Eliminatio n of intermediate storage tanks will likely require improvements in the reliability of the upstream and downstream equipment in or der to preclude unit or plant shutdowns caused by running out of material. Review secondary containments, impoundments, and spacing for tanks storing flammable mate rials. A sump inside a dike facilitates the collection of sm all spills. Sump drains, or pumps can direct material to a safe and environmentally acceptable place. See NFPA 30, the flammable liquids code (Ref 8.60 NFPA 2018). Review the layout to minimize the length of piping containing hazardous materials and to eliminate unneeded piping dead legs. In batch operations, minimize pre-charging the most energetic chemical. Consider adding energe tic material in a “semi-batch” mode. That is, add most of the ingredients initially, followed by the addition of the energetic ma terial under flow control. A Safety Instrumented System (SIS) is used to isolate the feed to

2 • Defining the Transition Times 17 2.3 Responses to deviatio ns during operations The definitions for the ten transient operating modes introduced in Table 1.1 are provided in Table 2.2, listed in order on when they are introduced and then discussed in detail in this guideline’s chapters: Part I Normal Operations 1. Shut-down (normal) 2. Start-up (normal) 3. Shut-down for a process shutdown 4. Start-up after a process shutdown 5. Shut-down for a facility shutdown 6. Start-up after a facility shutdown Part II Abnormal and Em ergency Operations 7. Shut-down for an unscheduled shutdown 8. Start-up after an unscheduled shutdown 9. Shut-down for an emergency shutdown 10. Start-up after an emergency shutdown Each of these transient operating mod es will be discussed in Part I and Part II.

96 | 7 Keeping Learning Fresh Many companies organize their process safety learning bulletins on intranet sites. Figure 7.3 shows an example. Figure 7.3 Lessons learned home page (Source: Repsol. Reproduced with permission.) Conveying messages verbally is another way to reach people with strong verbal-linguistic intelligence. Verbal communications may have increased impact when a supervisor or other trusted influencer is the one sharing information about an incident and its findings and recommendations. The challenge with verbal communication is the possibility that the message can evolve with time. For this reason, verbal communication of key process safety messages should be standardized and monitored for consistency. Because not every employee is oriented to verbal-linguistic learning, many companies seek to distill lessons learned into concise written communications that are easier to process. For example, one company effectively communicates essential learnings about how to prevent distraction and encourage situational awareness using simple pocket cards and matching posters (Anonymous). Figure 7.4 shows an example of a pocket card. The pocket card goes into the badge holder, so the employee sees it whenever they use [Company name and logo] •Eyes on the task •Mind on the task •Out of the line of fire •Do it safely or don’t do it Figure 7.4Simple pocket card

470 PROCESS SAFETY FOR ENGINEERS: AN INTRODUCTION Table A.2. Ethylene Chemical Properties Property Ethylene Chemical Formula C 2H4 MW 0.028 kg/gmole γ= C p/Cv 1.22 Flash Point -213 ⁰F Boiling Point -154.7 ⁰F Autoignition Temperature 842 ⁰F Lower Explosive Limit 2.75 %vol Upper Explosive Limit 28.6 %vol UN dangerous goods hazard class Division 2.1 – Flammable gas Specific Gravity Vapor 0.569 @ -154.7 ⁰F Pressure of ethylene inside the buffer tank 300 psig Temperature of ethylene inside the buffer tank 78 ⁰F MIE 0.08 mJ Heat of Combustion 1322.6 kJ/mol Figure A.2. Ethylene Buffer Tank Use the information in Table A-2 and Figure A-2 to answer the following questions and build your report for your boss. 1. Ethylene is best described as: corrosive, flammable, combustible, or toxic? 2. The information related to hazards associat ed with the equipment, technology and chemicals in the process is called: Process Safety Information (PSI), Process Hazard Analysis (PHA), Management of Change (MOC), or Hazard Identification and Risk Analysis (HIRA).

Principles of P&ID Development 51 For the warm lime softener, we specified the normal temperature of 80 °C; it is always operated on this tem- perature. However, nothing is static, and the tempera-ture will fluctuate with time. This may be because of a change in the ambient temperature, feed temperature, or feed composition. To maintain that temperature, a basic process control system (BPCS) needs to be installed for that piece of equipment. This system will attempt to reg-ulate the temperature for the purpose of process smooth-ness, for example, by control valves. That is why we specify a band around the desired tem- perature of 80 °C for the BPCS to operate in, say, from 75 to 85 °C. If a situation arises where the BPCS cannot con-trol the temperature, and it rises or drops to a level out - side this specified band, an alarm is activated. The alarm is meant to alert the operator to the danger and is an indication to take remedial action. If, for some reason, the operator fails to respond in a timely manner or the action taken does not mitigate the issue, the Safety Instrumented System (SIS) will auto-matically come on and try to fix the situation. The SIS is not an adjusting, regulatory action like BPCS but rather a direct action, which may involve opening a valve, shut - ting down, or starting up a pump. It is a drastic action, intended to protect equipment from damage and to keep operators safe. In this example, the SIS kicks on when the temperature goes above 100 °C or below 60 °C. If the SIS does not work to bring the process back under control, the temperature (the process parameter) moves to the final level. This is called a relief system, which is a purely mechanical system to protect the plant. Often, it may involve a safety valve that is activated to protect the plant, environment, operators, and other plant personnel. In this example, the safety valves can be set on 120 and −29 °C . Relief systems will be discussed in more detail in Chapter 12. This allocation of responsi-bilities among the BPCS, alarm system, and SIS is shown in Figure 5.12. Now we can define the “bands” in regard to different parameter levels. The band of high level to high‐high level and low level to low‐low level could trigger an alarm to warn the oper - ator for the upset. When the parameter goes beyond these bands and passes into high‐high (or low‐low), a trip may be activated because it is in a severe upset band. As you can see from Figure  5.13, the BPCS regulates and controls the process parameters between the high and low points. Therefore, the “playing court” for the BPCS is the band between the low and high levels. BPCS controls try to keep a parameter within low level and high level band. Once the parameter goes past these limits, this becomes a mild upset, and the alarm is activated. If the operator is unable to rectify the situation in response to the alarm, the parameter may progress to the high‐high level or the low‐low level. At these points, the interlock system, or SIS, will be activated. Finally, if the SIS is not able to bring the process under control, a severe upset ensues. This is where the relief action is effected, usually through a pressure relief valve, to avoid an industrial accident. It is interesting to note that the BPCS works within a band, but an alarm system and the SIS are triggered at a certain point or points. Table 5.3 summarizes these concepts for the example of temperature in warm lime softener. It is important to know that when the plant moves from BPCS to SIS, the actions of the SIS are not known as “control” functions. SIS works to prevent or mitigate the hazard. The next section discusses the relationship between different parameter levels and hazards. 5.3.4 Par ameter Levels versus Safety Generally speaking, when parameters are swinging within the band of high level to low level, the operation is safe. In a better terms, when parameters are in that band, Parameter Relief SlS Alarm BPCS Time Figure 5.12 Pr ocess guard layers. High structural integ rity levelM echanical re lief action Mechanical re lief actio nSlS action SlS actionAlarm AlarmBPCS actionsHigh–high le vel High le vel Normal le vel Low le vel Low–low le vel Low structural integ rity level Figure 5.13 Ac tion levels or bands for process guards.

xxxii GUIDELINES FOR MANAGING ABNORMAL SITUATIONS April Lovingood Eastman Marcus Miller TapRoot Assem Saleh Petro Rabigh Nicholas Sands DuPont Bryant Sartor AdvanSix Bruce Spencer Marathon Petroleum Todd Stauffer exida Andrew Trenchard Honeywell (ASM Consortium representative) Matthew Walters Exponent Stephanie Wardle Cenovus Energy CCPS wishes to acknowledge th e many contributions of the BakerRisk staff who contributed to this book, especially the writers listed here. Larry O. Bowler Michael P. Broadribb Michael D. Moosemiller Duane L. Rehmeyer Roger C. Stokes Editorial contributions in editing, layout, and assembly of the book were provided by Phyllis Whitea ker, BakerRisk document editor.

200 | 5 Aligning Culture with PSMS Elements As discussed above for MOC and OR, individuals engaged in the permitting process may be under pressure to com plete their approvals quickly, so the desired work can get started. Leaders should m ake it clear that work perm it writers should get the tim e they need to com plete the job safety analysis properly, and workers should be given the time to perform their work safely. The safety of the job should be valued m ore than the act of com pleting the perm it. Good cultural practices for safe work practices include: Permit approvers should be well trained in hazard recognition and control of hazardous energy. Work perm its should only be signed in the field, only after perform ing a thorough job safety analysis. Perm its should never be signed in the office. Closure of the permit should also occur at the job site. Work should be scheduled as m uch as possible, to avoid everyone wanting a permit on Monday m orning. If many people are sim ultaneously trying to obtain permits, this should not affect the thoroughness of the job safety analysis. Physical work should never begin before the permit is issued, nor should the permit be written and approved after the work is com plete. Equipment should not be taken out of service without approval of the process operators. Permits should be accurate, reflecting the actual work to be done and the correct names of the workers. Each of the workers should personally attach their locks. The scope of each permit should be clearly and precisely stated. If the scope of work needs to change, a new or m odified perm it is needed. In a strong process safety culture, every permitting exercise should be taken with fresh eyes. Every job is a little bit different, and field conditions m ay not be as expected. Every permit review • • • • • • • •

INVESTIGATION M ANAGEM ENT SYSTEM 67 The management system sh ould describe minimum initial training and refresher training. High quality training for potential team leaders, members, and supporting personnel helps ensure success. The level of detail contained in the management system may vary. For example, it may provide a brief summary and then refer to a traini ng management system document or position curricula for the detailed training information. A summary of training topics for each group is provided below: M anagement This group needs to be familiar with the concepts, policies, and extent of commitment from executive managemen t; specific assignments of responsibility and resource commitmen ts associated with process safety incident investigation; the employer’s incident investigation management system; and report content, including what constitutes clear actionable recommendations. Site - M anagement “Management” topics above should be supplemented with basic investigation concepts, investigation methodologies, causation concepts, fact-finding vs. fault-finding philosophy, general internal legal protocols, and media relations/communications po licies and practical exercises. All Employees This group includes operators, mechanics, first- line supervisors, auxiliary staff groups such as tec hnicians and engineers, and middle-level management. These are employees that are in a position to first notice an incident and may provide support ac tivities vital to the success of an investigation team. They should be trained on how to differentiate an accident from a near-miss. They should also be educated on the requirements of employer ’s investigation management system, with a focus on what to do once an incident is identified, and the site’s incident reporting procedure. Investigation Team M embers (Including Team Leader) This group is intended to be a designated pool of specifically trained investigators to be called into service as needed. Additional training focuses on the support functions of an investigation, particularly on how to effectively gather and preserve data. For instance, team members would be trained on how to preserve evidence, interview p eers, develop test plans, and develop sampling procedures. Depending on their role in the investigation, some team members may need training in data analysis and the use of specific

52 are eliminated. Figure 3.2 shows the conventional design, and Figure 3.3 shows the reactive distilla tion design. For this process, Siirola (Ref 3.19 Siirola) reports significant reductions in both capital investment and operating cost for the reactive distillation process. Figure 3.2: Conventional process for methyl acetate (based on Ref 3.19 Siirola).