Datasets:

hndc

/

AI4Hazard-small

like 0

PROJECT DESIGN BASICS 165 Figure 10.4. Typical P&ID symbols (Patel) The PFD and P&ID primarily describe the process. They do not include civil engineering or construction details.

4 • Process Shutdowns 65 Since these additional procedur es may not be performed very often, it is essential that everyone involved in a shut-down for a process shutdown understands what the different steps are, has the operational discipline to follow th ese steps, and can quickly recognize and respond properly when things are not going as planned. Many of the US CSB incidents show that unreviewed and unapproved changes in the field have led to severe incidents when executing nonroutine shut-downs for project-related ac tivities. When people do not adequately assess and address the hazards, or when people make changes to the established, approve d plan without understanding how their changes increased the proces s safety risk, consequences can include severe injuries, fatalities, environmental harm, and property damage. Due to the harm which occu rs, project plans should be thoroughly reviewed and approved by every group involved in the planning and execution the project- related shutdown, especially those in operations, maintenance, and engineering. Effective handover protocols and sy stems should be in place, as well, to ensure that those working on the equipment know what hazards have been or have NOT been addressed before the work commences. There may be special clean-out or isolation procedures required for the project that are not done during a normal shut-down transition. There may be special ha zards that are introduced to make the equipment safer to work on that should be carefully monitored during the shutdown-related work, such as displacing toxic gases or flammable vapors with an inert, as phyxiating gas (i.e., Nitrogen). A robust Management of Change (MOC ) system can help ensure that everything is ready, that all the equipment is prepared and in a known state by everyone before beginning the scheduled work [33]. Ensuring and verifying that everything is ready will help reduce the miscues that have led to significant incidents.

40 | 4 Examples of Failure to Learn You may not have heard of most of the incidents in this index, but you should have heard of the major disasters described in this chapter and Chapter 8 (see Table 8.1 for a summary of landmark incidents with their associated prominent findings and causal factors). You might expect that the most well-known incidents had unusual causes, but it’s clear that many had the same causes as incidents that preceded them. Each of these incidents has been studied deeply, with the findings published, distributed broadly, and pored over by safety professionals around the world. Nonetheless, the causes of the most prominent incidents are repeated in subsequent incidents. In this chapter, we examine a few of these repeating failures in more depth. 4.1 Process Safety Culture Process safety culture is described by CCPS (CCPS 2018) as: …the combination of group values and behaviors that determine the manner in which process safety is managed. It’s how we conduct ourselves when we think no one is watching. An effective and sound process safety culture is a foundational element in CCPS’s Risk Based Process Safety (RBPS) pillar of Committing to Process Safety. This culture must be established at all levels, from the boardroom to the front line. Everyone in the organization has a role in process safety culture and must perform their role reliably and with professionalism. A weak process safety culture often undermines successful execution of the other PSMS elements, increasing the probability of an incident. One commonality among the landmark incidents covered in Chapter 8 is that they all had a weak process safety culture. Although these incidents have been well documented, companies are not learning from them, suggesting a breakdown in process safety culture.] In the 2014 LaPorte, TX, USA, incident, the primary focus of the company’s safety culture program was personal safety. This helped the company to reduce its US OSHA total recordable injury rate. But according to the CSB report on this incident (CSB 2019), the company never evaluated its process safety culture. The report stated: Had its efforts included a focus on perceptions of process safety as well as personal safety in its Safety Perception Survey, or had it performed a separate process safety culture assessment with the intent of improving See Appendix index entry C26

82 | 6 Implementing the REAL Model team or obtain investigation notes. You may even be referred to a contact person at the company so you can ask them directly. Finally, whether it’s better, worse, or only different, your plant almost certainly has different design features, preventive and mitigative barriers, and PSMS, standards, and policies than the company that had the incident. So, think about how the incident scenario might have played out if it happened in your facility. What warning signs might have been present pre-event? Would your systems have prevented the incident? If so, how do you know they would have functioned reliably? 6.5 Internalize The next step is for the effort to expand beyond the individual evaluator to include a small team of diverse individuals who provide relevant expertise. In this step, the group evaluates the list of deeper learnings and internalizes them to the company by developing formal recommendations. This step is important because the recommendations developed for the external incident may not work for your company given its expertise, resources, or technologies. Alternately, your company may have access to better solutions. And finally, a solution that comes from a company’s experience and culture is most likely to be accepted and become institutionalized. Depending on the subject matter, the team members will include those who conducted the initial review as well as others with expertise in relevant areas, such as: • process technology • engineering • corporate and public standards • HSE policy • process economics and finance • manufacturing and operations • procurement • human resources • transportation and logistics.

General Procedures 413 In this method the brain and eyes should work closely together to do a type of “finding the hidden object” game! In this method the checker scans the P&ID to find missing items and text, or illegibility. 19.5 Required Quality of P&IDs at Each S tage of Development Now the question is what should be the quality of P&IDs at each step of a design project. For example do all the drain valves need to be shown even in the IFA revision of a P&ID set? The answer to the above question is clearly no. To expand more on the answer, it should be noted that it is expected that more details are depicted on the P&ID when we are approaching the end of project. At the beginning of the project, for example on the IFR version of a P&ID, only large items are shown and no detail can be found. On the last revision of a P&ID, the IFC version, all the details should be depicted. Each company has its own “standard” for quality of P&IDs in each stage of development. However, Table 19.4 can be used as a guideline where there is no standard available. Table 19.4 Qualit y of P&IDs at each step of a design project. IFR IFA IFD IFC 1.00 Equipment 1.01 Positioning (necessity, existence) Majority of them All Complete 1.02 Type Majority of them All Complete 1.03 Tag Majority of them All Complete 1.04 Call‐out: capacity Not available except for long lead itemsAll Fine‐tuned with vendor dataComplete 1.05 Call‐out: other numerical specificationsX Some of them Majority of them Complete 1.06 Required number of them and sparing policyMajority of them All Complete 1.07 Materials of construction Majority of them All Fine‐tuned with vendor dataComplete 1.08 Diver – type Majority of them All Complete(after fine‐tuning) 1.09 Diver – power Estimation Majority of them Complete 1.10 Critical elevations Majority of them All Complete(after fine‐tuning) 1.11 Utility positioning for equipment (requirement of utilities)Some of them Majority of them Complete 1.12 Utility branch sizing for equipment Some of them Majority of them Complete 1.13 Equipment isolation arrangement Majority of them Complete 2.00 Packaged units 2.01 Positioning (necessity, existence) Majority of them All Fine‐tuned with vendor dataComplete 2.02 Type of components Estimation Fine‐tuned with vendor dataComplete 2.03 Tag Majority of them All Fine‐tuned with vendor dataComplete 2.04 Call‐out: capacity All Fine‐tuned with vendor dataComplete 2.05 Call‐out: other numerical specificationsSome of them Complete 2.06 Required number of them and sparing policyMajority of them All Fine‐tuned with vendor dataComplete (Continued)

Piping and Instrumentation Diagram Development 392 Even though, depending on the type of equipment in each area, the utility streams in each utility station could be different, some companies decide to install just a standard US in which there are a fixed number and type of utilities in all of them. They have made the decision to do this so as to not confuse operators; they will know whether a specific US does or doesn’t have, for example, utility steam. Utility stream pipes are generally 2″ or smaller than that. The detail of each utility stream pipe up to the point for usage by the operator could be different depending on the type of utility stream. However the arrangement of isolation valve, pressure regulator, check valve, and hose connection is very common (Figure 18.13). The isolation valve is needed to isolate the downstream if there is an issue. The isolation system could be decided to be more complicated than a single isolation valve – e.g. double block and bleed – in some cases. A regulator could be needed to adjust the pressure to a pressure that is not harmful for operators. A check valve can also be placed at the last point of a utility pipe before connection to process to make sure con-tamination of utility fluid by process fluid is prevented. It is important to note that, as all the USs are connected to a utility network, the pressure of utilities at the edge of each US may fluctuate during the usage of the utility. For example, if three USs that are located very close to each other using utility water at the same time, the utility water pressure will definitely be less than the pressure of utility water if only one US is functioning. Because operators don’t want to see any pressure fluctuations in any utility steam, they will generally use a pressure regulator to adjust the pressure in the utility network to a fairly con-stant and non‐harmful pressure for the operator. 18.5 Off‐Line Monitoring Programs We need off‐line monitoring programs wherever we need to check the “process properties” of a stream but there is no process analyzer available. The word “available” here is used for the situations where there is no process analyzer in the market to measure the process parameter of inter - est, or where the cost of the available process analyzer is beyond the plant budget, or the criticality of the parame-ter is so low that doesn’t justify the purchase of a process analyzer. In such situations if (and this is an important if) the process parameter is sluggish enough, an off‐line monitoring program can be used. If the process parameter is very agile, which means the process parameter changes a lot and very quickly, off‐line monitoring programs don’t work. 18.5.1 The P rogram Component An off‐line monitoring program is defined as: taking a sample, protocol to transfer the sample to the lab, suitable testing procedure to measure the process parameter, and sending the result to the plant operator to take the appro-priate action. Therefore, an off‐line monitoring program works in a similar way to a control loop and the difference is that in an off‐line monitoring program some automatic actions are replaced with human actions. This concept is depicted in Figure 18.14. Figure 18.14 Off‐line monit oring programs.us USMain header Subheader Figure 18.12 A utility net work connected to a utility station. Utility network Figure 18.13 Detail of utilit y streams in utility stations.

364 PROCESS SAFETY FOR ENGINEERS: AN INTRODUCTION Situational awareness. Simply put, this is “k nowing what is going on”. It is related to information processing and requires attentio n. Good situational awareness depends on sufficient data and time for the data to be sensed, perceived, and interpreted. (Endsley) Decision making. Decision making can be thought of along a continuum as shown in Figure 16.7. Rational decision making is when a person a pplies reasoning and logic, which takes time, to make the most ideal choice, for example when deciding which car to buy. Quicker decision-making uses biases and shortcuts including the following. Using checklists and reminders can lessen the potent ial negative impacts of these biases and shortcuts. Recency - more recent information is given priority Neglect – information is overlooked Availability - information that is easi er to recall, is more influential Small samples – hypotheses are created based on only one or two experiences Confirmation bias - once a hypothesis or response is decided on, more weight is given to evidence that confirms the hypo thesis, and less given to evidence that conflicts with the hypothesis Figure 16.7. Decision making continuum Intuitive decisions are made without mental processing or using shortcuts. Consider a decision in which someone has a feeling that it should be safe enough to ‘bend the rules’ in this case. For example, if they have motivation to get the task done quickly, they may be more inclined to bend the rules. The decision will appear reckless in hindsight but feel acceptable at the time. Setting limits can protect against inappropriate intuitive decisions. Stress. Stress is the response to unfavorable environmental conditions. If excessive demands are placed on a human, it is possible to exceed the individual’s capacity to meet them. Sources of stress include: Environmental sources of stress such as te mperature, vibration, noise, humidity, glare Life stressors such as social pressures, fi nancial pressures, family arguments, death of a close relative, smoking or drinking to excess, as well as physiological factors such as hunger, thirst, pain, lack of sleep and fatigue Organizational stressors such as poor comm unication, role conflict, workload, lack of career development, pay inequa lity, bureaucratic processes Rational Quicker Intuitive

Appendix 221 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. 53% 47% 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 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 T2 Laboratories, Inc. Runaway reaction1 111 1 1 1 1 Millard Refriger. Serv. Ammonia1 11 1 Hoechst Griesheim Runaway reaction11 11 Arco Channelview Explosion 1 1 1 Port Neal, USA AN Explosion1 11 Hickson & Welsh Jet Fire11 1 1 1 1 Chevron Richmond Refinery fire 11 1 11 Buncefield depot Storage Tank 11 1 1 1 11 Celanese Pampa Explosion11 1 Hayes Lemmerz Dust explosion1 1 1 1 1111 11 1 Macondo Well Deepwater Horizon11 1 1 1 1 1 11 DuPont LaPorte Methyl Mercaptan1 11 1 1 1 DPC Enterprises Chlorine1 11 Gaylord Chemical Nitrogen Tetroxide1 11 1 Fukushima Daiichi Nuclear Plant11 1 1 1Start-up or Shut-down transient operating mode incidents from: CCPS 2019 (More Incidents the Define Process Safety) Pillar IV Learn from ExperienceIncident Elements Identi fied as "weak" (See Figure 10.3) No. of Identified RBPS Causes Risk Based Process Safety ElementTransient Operating Mode Pillar I Commit to Process Safety Pillar II Understand Haz. and Risks Pillar III Manage Risk Table A.2-2 Summary of the in cidents during the transient o perating mode (Continued)

TOOLS AND METHODS FOR MANAGING ABNORMAL SITUATIONS 115 and/or checklists of topics specific to abnormal situations such as the bulleted list in this section. For established chemical processes, with an experienced operating team in place, another approach to discussing potential abnormal situations is through tabletop exer cises or drills. The facilitator can challenge the operating team members to document how they would respond to various upset situations . This type of exercise can be expanded to stimulate further discu ssion about situations where other abnormal conditions could be encount ered. The result of these tabletop drills should be used to improve operating procedures, training of personnel, and installation of a dditional safeguard controls and hardware. In summary, traditional HIRA revi ews typically consider scenarios with failure of a single device or system, whereas an abnormal situation can involve simultaneous failure of mu ltiple devices or systems. For such situations, perhaps a “What-If, HA ZOPstructure can be used to brainstorm abnormal scenarios that should be considered, especially with respect to process alarms, emergency procedures, or emergency training drills. Using a “What-If” appr oach for an established plant that has been in operation for many years could be useful in highlighting events that have occurre d but where lessons have not necessarily been learned, incorporated, or embedde d into the operating procedures, culture, or practices. The HIRA fa cilitator must be familiar with the concept of ASM to direct the risk anal ysis team effectively during this “What-If” exercise. 5.3 PROCESS CONTROL SYSTEMS Table 5.3 provides an overview of so me of the strengths and weaknesses of tools that may be considered when developing strategies for monitoring, diagnosing, and predic ting both process variances and abnormal process upsets.

DETERM INING ROOT CAUSES 249 release ultimately resulted in fish being killed in the local river. The overheating of the temporary water treatment unit occurred when a firewater hose providing cooling water to the temporary water treatment unit ruptured. The plant was provided with an automatic trip that apparently failed to work, as well as an alarm to which the operator did not respond. The sequence of events is shown in Figure 10.24. Figure 10.24 Incident Sequence The investigation team interviewed all contract operators and their supervisor, the temporary water treatment unit vendor’s engineers, plant personnel at the process plant unit, procurement personnel, and operations management. 10.8.2.1 Causal Factor Identification After the interviews and other evidence gathering activities are complete, the causal factors should be iden tified and, if appropriate, a causal factor chart can be developed. Four causal factor s were identified: 1. Contract operator falls asleep 2. Fire hose ruptures 3. Automatic shut -off jumpered 4. Sleeping contract operator can’t hear alarm due to nearby diesel (noise)

28 | 2 Core Principles of Process Safety Those responsible for process safety are fully qualified to do the job. Process safety staff is not placed in the untenable position of having to prove that an operation is unsafe. Those desiring or advocating certain operations or conditions should be required to prove that those operations or conditions are safe. Process safety metrics and audits are used to guide improvement. They are not treated as adversarial or punitive activities. The imperative for process safety extends equally to contractors, labor unions, headquarters staff, and outside m em bers of the B oard of Directors. To the degree possible, the imperative also extends to community m em bers, public interest groups, and regulators (see also section 4.4). The Baker Panel (Ref 2.5) noted that com mercial considerations, including cost control and production, play a role in defining the safety culture of an organization. All organizations that produce goods and services not only face limitations on hum an and financial resources, but also m ust effectively manage the tension that exists between the operational demands relating to production and the demands relating to safety. Reason (Ref 2.6) summ arized this natural tension: “It is clear from in-depth accident analyses that some of the most powerful pushes towards local [culture] traps come from an unsatisfactory resolution of the inevitable conflict that exists (at least in the short-term) between the goals of safety and production. The cultural accommodation between the pursuit of these goals must achieve a delicate balance. On the one hand, we have to face the fact that no organization is just in the business of being safe. Every company must obey both the ‘ALARP’ principle (keep the risks as low as reasonably practicable) and the ‘ASSIB’ principle (and still stay in business).”• • • •

OPERATIONAL READINESS 377 verbal or written hand over. The investigatio n found that vital communications systems on Piper Alpha had become too relaxed, with the result that the Work Permit was left on the manager’s desk instead of it being personally gi ven to him to enable proper communication at the subsequent shift change. If the system ha d been implemented properly, the initial gas release would not have occurred. However, on ce this had occurred, many other factors conspired together to cause the fatalit ies and loss of platform (CCPS 2008). Lessons Safe Work Practices. Good safe work practices are needed to control hazards due to maintenance work and make sure equipment is re ady before starting up. These work practices need to include communication between th e people doing the work and production personnel. In Piper Alpha, the night shift crew was not informed that the relief valve had been removed and the pump was not ready to be returned to operation. Additionally, the blind flange put in the line was not properly inst alled, so it could not hold the pressure. Emergency Management. The Offshore Installation Manager (OIM) did not order an evacuation immediately resulting in his fatality shortly after. Fire boats responding to the event waited for orders from the OIM, which delayed response. Many of the evacuation routes were blocked. Other platforms in the area were feedin g material to Piper Alpha and did not turn off their feeds, providing a continuing source of fu el to the fire. Even though they could see the fire on the horizon, they believed they needed permission from onshore management to turn off their feeds. The workers on the platform were not adequately trained in emergency procedures, and management was not trained to provide good leadership during a crisis situation. Evacuation drills were performed, but not every week as required by regulations. A full drill had not taken place in over three year s. The place where the crew gathered was not safe. Smoke could enter and this caused the fata lities. After Piper Alpha, the U.K. Government required that there be a Temporary Safe Refuge (TSR) protecting staff sheltering there from explosions, fire, and toxic smoke until safe evacuation can be organized . The Piper Alpha fire and explosion led to development of stronger offshore safety requirements in the U.K. Offshore Installation s (Safety Case) Regulations. The Safety Case regulations are goal-based and replaced the prev ious prescriptive regulations. A Safety Case is the documentation that a production organiza tion must submit in the U.K. to demonstrate that their operation is safe. Another change made was having responsibility for enforcing safety case moved from the U.K.’s Department of Energy to the Health and Safety Executive (HSE) to avoid potential conflicts between production and safety. Introduction to Operational Readiness Chapters 10 through 15 addressed project design , methods to identify hazards and analyze risk, and risk prevention and mitigation meas ures. Whether these concepts are applied to a new project, following a change, or during mainte nance, the facility will be started. This chapter addresses the topic of verifying the facility is ready for a safe start up and safe operation. Process safety incidents occur five times more often during startup than during normal operations. (CCPS 1995)

220 Cyber Security (Protection of critical information systems including hardware, software, infr astructure, and data from loss, corruption, theft, or damage). Crisis Management and Emergenc y Response Plans (process by which an organization deals with a major event that threatens to harm the organization, its stakeh olders, or the general public). Policies and Procedures (policies are the rules that govern how a company conducts business; whereas procedures are a set of steps for administering a process). Information Security (the practi ce of preventing unauthorized access, use, disclosure, disrup tion, modification, inspection, recording or destruction of information). Intelligence (Information to characterize specific or general threats when considering a threat ’s motivation, capabilities, and activities). Inherent Safety (a concept and approach to safety that focuses on eliminating or reducing the ha zards associated with materials and operations used in the pr ocess where this reduction or elimination is permanent and inseparable). 9.5 ASSESSING SECURITY VULNERABILITIES Leaders in the chemical industry have recognized the potential for chemical facilities, or chemicals themselves, to be used as weapons by terrorists or other criminals. The in dustry also understood the need to expand existing security programs to address these new and serious threats. U.S. processing industries built on existing process safety management systems to develop security management systems that included requirements to assess and prioritize potential security risks posed by chemical facilities, and to implement measures to address those risks. Examples include the Responsible Care® Security Code, adopted by the Chemical Manufacture rs Association in June 2003 (Ref 9.1 ACC) and API’s Security Guidelines for the Petroleum Industry (Ref 9.3 API) The American Institute of Chem ical Engineers supported the development of such security management systems by publishing Guidelines for Analyzing and Managing the Security Vulnerabilities of Fixed

Piping and Instrumentation Diagram Development 64 For example, if an item was added only for the ease of maintenance, another duty may be placed on its shoul-ders during normal operation. Later in this book, each piece of equipment and oppor - tunities for merging them are discussed. 5.6 Dealing with Common Challenges in P&ID De velopment During the development of a P&ID there are occasion-ally some challenges to find a better option among the available options. Sometimes these challenges are in the designer’s mind and are resolved easily, but sometimes a challenge can be the subject of heated debates between stakeholders. Following are a few listed and discussed. ●“Should I add this item or not?”The components and items should be added to give the operator enough flexibility. A plant with not enough resources is difficult to operate, and it is also the case for a plant with more than enough pipe cir - cuits, control valves, alarms, and SIS actions. For example, a plant with too many alarms will overload the operator, which results in operators losing a sense of urgency in the case of an alarm (Figure 5.27). However, the designer should be careful of not falling in the trap of “adding does not hurt!” This is a popular statement when P&ID developers try to bypass the complete evaluation of the need for an item in the sys - tem and placing it in the system regardless. However, although adding an item might not increase the capital cost of the project (if it is small and inexpensive), it will increase the operating cost because of the required inspection, maintenance, probable utility or chemical usage, and so on. In addition to that, any new item in the system is an opportunity for mistakes, cross con-tamination, and leaks. ●“Based on my past experience… ”The inherent creativity required in developing P&IDs may become to hinder, if for every single case one refers to past experience. Every past experience should reevaluated and tailored before being applied to new situations. Unlikely as it may seem, the “this is what has been done before” mentality is not the most effi-cient way to develop P&ID. On the other hand, tech-nological innovations, availability of materials, quality of raw material, and the required quality of products, Train 1 Train 2Unit A Unit A' Unit B' Unit C' Unit D'Unit B Unit C Unit D Crosso verC rosso verC rosso verC rosso ver Crosso verProduct ProductFeed Feed Figure 5.26 A “ train” in a plant. Few tools Bad operable plantConfusion Flexibility in operationToo many toolsFigure 5.27 The “sweet spot” for providing items for a plant.

194 | 5 Aligning Culture with PSMS Elements From the perspective of process safety culture, training provides the skills that give leaders the ability to empower individuals to successfully fulfill their process safety responsibilities , and the confidence that they may defer to expertise . Training also provides opportunities for leaders to reinforce the imperative for process safety and maintain the sense of vulnerability . Every employee and contractor at a facility requires some form of process safety training. The need to train operators, m echanics, supervisors, and other production personnel should be clear. B ut even if the individual never works outside of the administrative office, they still need to be skilled in the necessary emergency procedures and understand the hazards m anaged at the location. Some office workers m ay require m ore training. For exam ple, procurement professionals need to understand the process safety implications of changes in the sources of spare parts, replacement equipm ent, and raw materials. Com panies manage process safety training initiatives differently. It is not unusual, for a common training group to organize and m anage training in all skill areas. Other approaches include m anaging all skills by department, and managing process skills through the process safety function. Each has benefits and potential drawbacks, but with strong culture, leadership , and open and frank communication , all approaches can work. If training is m anaged outside of the sphere of process safety culture, it may be necessary to harmonize the training culture with the process safety culture. For example, if the overall training culture focuses on checking the boxes of required training – ethics, non-harassment, and so on – it will take some effort to establish a process safety training effort that focuses on com petency and culture. That effort is necessary to prevent the overall training culture from undermining the process safety culture.

114 INVESTIGATING PROCESS SAFETY INCIDENTS • Subtle changes in process variables • Unexpected relationships be tween certain parameters • Reliability of specific instrumentation • Unexpected problems and associat ed changes in the process made during the initial startup of the system • History of previous problems and actions taken to avoid/rectify problems If a similar incident occurred in the past, it might be appropriate to re- interview those witnesses involved to gain insights into this investigation. A list of potential witnesses is provided in Figure 7.2. Employees Contractors and Third Parties On-shift operators Statutory compliance officer/ Safety, Health and Environment officer/ Fire engineer/officer Off-shift operators First responders/emergency response personnel Maintenance personnel Contract maintenance Process engineers Manufacturer’s representatives Operations management Personnel previously involved in operation/ maintenance of the system, including former employees and personnel involved in the initial start-up of the system Maintenance management Personnel involved in previous incidents associated with the process Chemistry and other laboratory personnel Janitorial, delivery, and other service personnel Warehouse personnel Original design/installation contractors or engineering group Procurement personnel Security force (roaming guards or sentries) Quality control personnel Off-site personnel and visitors Research scientists Members of the community Figure 7.2 List of Potential W itnesses

420 PROCESS SAFETY FOR ENGINEERS: AN INTRODUCTION Figure 20.6. Coffeyville Refinery 2007 flood (KDA) Emergency Response Planning Emergency response plans should be develo ped collaboratively with experts aware of potential hazards, operations personnel that co uld be involved in an emergency response, and emergency responders (internal and external ). The Local Emergency Planning Committees (LEPC), more than 3000 across the US, develop em ergency response plans and interact with stakeholders. The following are the steps in developing an emergency response plan. Identify accident scenarios based on hazards. Emergency response plans can address a few scenarios involving each type of hazard to cover the range of potential scenarios. Process safety emergency scenarios can be selected fr om hazard identification studies (refer to Chapter 12) and from industry incident history. Other emergencies, such as those noted in section 20.3, may be identified through focusing on the specific hazard. The CCPS Monographs on Assessment of and Planni ng for Natural Hazards and Risk Based Process Safety During Disruptive Times , and CCPS Guidelines for Analyzing and Managing the Security Vulnerabilities of Fixed Chemical Sites provide helpful guidance. (CCPS 2020 and CCPS 2003) Plan response actions. Response actions should be identified, reviewed, and optimized in advance of a potential incident as opposed to tr ying to decide what to do in the heat of the moment. Response actions include, but are not limited to, the following. Emergency recognition and reporting – Identi fying what is considered an emergency and when and to whom it is to be reported.

242 BP has published Inherently Safer Design Guidelines for New Projects and Developments . DuPont (Ref 10.6 Clark) describes how ISP (Inherently Safer Processing) is integrated in to the overall corporate PSM program, based on a checklist an d a corporate training program. A semi-quantitative ISP scoring system is used by corporate R&D to ensure that it is appropriatel y considered at the earliest stage of a process life cycle. In addition, the Contra Costa County, CA, Health Services Department has issued a guidance docum ent for the IS review of existing facilities, and new facilities at the chemistry-forming, facilities design scoping and development, and basic project design stages. IS analyses must be performed for all situations where a “major chemical accident or release”—as defined in the stan dard—could reasonably occur. This document also includes guidance for evaluating the feasibility of recommendations, and for IS review documentation. IS reviews for existing processes can be conducted as part of an initial/five-year PHA or as a separate study, using a ch ecklist or guideword analysis that incorporates IS. See Appendix A for a detailed example of an IST checklist. Table 10.3 that appears la ter in this document offers sample Guidewords, while Table 10.5 sh ows a sample Guideword Matrix. 10.5.1 Inherent Safety Review Objectives The objectives for an inherent safety review are to employ a synergistic team to: Understand the hazards. Find ways to eliminate or reduce those hazards. The first major objective for the inherent safety review is the development of a good understanding of the hazards involved in the process. Early understanding of th ese hazards provides time for the development team to implement re commendations from the inherent safety effort. Hazards associated with flamma bility, pressure, and temperature are relatively easy to identify. Reactive chemistry hazards are not. They are frequently difficult to identify and understand in the lab and pilot plant. Special calorimetry equipment and expertise are often needed to fully characterize the hazards of runaway reactions and

Piping and Instrumentation Diagram Development 226 the functionality of the spring. The other feature of a bel- lows‐type pressure relief valve is that the pressure down-stream of the relief valve (or backpressure) doesn’t impact the set pressure of the relief valve. Pilot‐type relief valves also have this feature. Schematics of the different types of PRDs are depicted in Figure 12.12. 12.12.2 Rupture D isks There are two types of non‐reclosable PRDs. In “rupture disks, ” a “hole” is covered by a disk that will rupture at a specific pre‐set pressure, and release the pressure. The sec - ond type of non‐reclosable PRDs is very similar to a spring‐loaded PRV, but the spring is replaced by a “buckling/breaking” pin. Between these two non‐reclosable PRDs, rupture disks are more common. Rupture disks are manufactured in main three forms: flat, forward dome, and backward dome. These three types of rupture disks could be in the form of a solid sheet, a hinged or scored sheet, with a cutting edge, and in composite. The available types of rupture disks are shown in Table 12.10.12.12.3 Decision Gener al Rules Deciding on PSD types are based on quantitative and qualitative parameters. As many criteria go back to the design stage of project, they are not discussed here. 12.13 PRD Identifiers As it was stated in Chapter 4, the identifiers of PRDs – as an item of instrumentation – on P&IDs are PRD sym-bols, PRD tags, and PRD technical information. 12.13.1 PRD Symbols and  Tags Table 12.11 shows P/V RD symbols and tags. As can be seen in the table, when there is a need to have both a pressure relief valve and a vacuum relief valve, these two devices can be merged together to save money on nozzles and other operating costs. The PVSV was invented for this purpose. PVSVs (pressure/vacuum safety valves, or as some companies call them, PVRVs [pressure/vacuum relief valves]) are devices that protect Inlet connection Conventional typeSpring-type SchematicDead weight-type Pilot-type Outlet connection Inlet connection Balance typeOutlet connection Inlet connectionInlet connection Outlet connectionOutlet connection Figure 12.12 Differ ent types of relief valves.

3. Options for supporting human performance 27 Table 3-3: Example of a rule-based mistake Event Formosa Plastics Vinyl Ch loride Monomer Explosion What happened? The operator was going to drain a flush out of the reactor to prepare for cleaning it. The reactors spanned two levels of the process area, with the bottom valve controls on the lower level. An operator walked downstairs and mistakenly went to the wrong reactor. Because the reactor still contained highly hazardous material (Vinyl Chloride Monomer), the bottom outlet valve was interlocked closed. The operator used an emergency air supply to force open the outlet valve on the active reactor. This allowed the Vinyl Chloride Monomer to escape from the vessel. A cloud of Vinyl Chloride Monomer spread across the floor. The supervisor ran downstairs to investigate, then returned upstairs to try to reduce the speed of the released chemical. The monomer was ignited, causing an explosion. A mistake The bottom outlet from the vinyl chloride reactor vessel was mistakenly opened by an operator. The supervisor tried to lower the pressure in the vessel to slow down the release of the chemical, by ordering operators to open valves. He did not command an immediate evacuation of the unit. The flammable gas exploded, killing five people, seriously injuring two and destroying the unit. Causes The investigation found that staff had not been trained properly to order an immediate evacuation, which could have saved lives. Further reading and video U.S. Chemical Safety and Hazard Investigation Board. Vinyl Chloride monomer explosion. [21] 3.3.5 Common causes of rule-based mistakes Rule-based mistakes often involve misunderstanding of what is happening and/or making the wrong decision about what to do next. Common causes of rule-based mistakes are: 3.3.5.1 Missing, confusing or incomplete procedures If the task is infrequent or must be perf ormed quickly, a person may rely too much on their knowledge to make decisions or judgments or they may quickly improvise plans of action, instead of recalling inst ructions and procedures. A person may incorrectly assume that the task can be done in the same way as a similar task, or they may not know what the correct task steps are.

Chapter No.: 1 Title Name: <TITLENAME> c18.indd Comp. by: <USER> Date: 25 Feb 2019 Time: 07:41:29 AM Stage: <STAGE> WorkFlow:<WORKFLOW> Page Number: 381 381 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 18.1 Introduction In this chapter we cover some systems that couldn’t be categorized in the previous chapters. 18.2 Safety Issues The main purpose of taking care of “safety” in process plants is preventing injury, and in its worst case, preventing death. Safety issues should be addressed in all aspects of process design and also P&ID development. The first step in upholding safety is understanding hazards and their relation to injuries. Safety should be considered in all P&ID development activities. No effort is made to make this section a fully exhaustive section on safety of process plants. 18.2.1 Diff erent Types of Hazards Hazards can be arbitrarily classified into three groups based on the initiators: mechanical hazards, chemical hazards, and energy hazards. Mechanical hazards are the hazards causes by mechan- ical systems and devices. The examples are: impacts, penetrations, compressions, rolling‐overs, falling (including slipping and tripping). Chemical hazards caused by chemicals in different forms include: liquid, gas, vapor, fume, and dust. Energy hazards are caused by light, optical radiation, contact with hot or cold surfaces, or noise. Hazards caused by biological matter are not generally categorized as safety hazard but are known as health hazards. 18.2.2 Hazards and I njuries Before the injury all effort should put into preventing an injury by reducing the risk of injury. After the injury, all effort should be aimed at mitigating and limiting the consequences of an injury.Table 18.1 shows these two concepts.In the left column  –  or before the accident  –  all the efforts are to minimize the hazard to prevent the injury. In the right column – or after the accident – the injury has happened and efforts should focus on minimizing the extent and breadth of the injury. Let’s start with the left column.In the scope of preventing injury what can be imple- mented during the design of a plant is firstly reducing or eliminating the hazard. Eliminating the hazard can be done by passively eliminating hazardous matter. The passive prevention of injury generally goes into the deep concepts of process and generally cannot be implemented in the plant during P&ID development. Such strategies can be implemented in the BFD (block flow diagram) or PFD (process flow diagram) develop-ment stages. During the P&ID development stage of projects, active methods – or placing barriers – is the main strategy to prevent injury and reduce hazards. The third strategy to reduce hazards is implementing rules and standard operation procedures and forcing operators to follow them. This the weakest way of dealing with hazards and also doesn’t have any impact on P&IDs. Therefore we focus only on the first item of preventing injuries actively by “masking” hazardous matters. In the right column we only put few of the actions. Out of these actions, providing safety showers and eye washers have P&ID footprint and will be discussed here. 18.2.3 Mechanical Hazar ds There are different “guards” available to protect personnel against mechanical hazards. However they are generally not shown on P&IDs. The examples are different types of “machine guards” including shaft guards, belt guards, coupling guards, etc. The majority of mechanical hazard barriers are offered by equipment vendors.18 Ancillary Systems and Additional Considerations

94 Guidelines for Revalidating a Process Hazard Analysis Experience shows that a well-organized and conducted Update can effectively address the changes that o ccur over a revalidation cycle. However, some companies require a Redo of the PHA after a specific number of cycles (e.g., every second or third revalidation), as discus sed in Section 3.3.6. This is done for several reasons: • Sometimes overlooked errors in a PHA will perpetuate from one Update to the next. • Changes may have been missed in a previous Update and a new team, unbiased by previous reviews, is more likely to capture these changes for review. • Nuances of risk understanding, tolerance, and general risk practices can change within a company over time, and a Redo allows the PHA to be recalibrated for cons istency with those evolutionary changes. • Team members should carefully think through the entire range of potential process upsets and judge the current risk of loss scenarios as an educational benefit to the entire team. 5.1.3 Combining Update and Re do in a Revalidation In most cases, the evaluation of the prior PHA and events since it was conducted identifies a range of issues that should be remedied during the revalidation. At one end of the spectrum, as illustrated in Figure 5-2, the deficiencies may be beyond repair (e.g., an inappropriat e core methodology was used) or an Update would take more time and effort than starting over, and a Redo should be performed. At the other end of the spectrum, the prior PHA was excellent and the few changes that have occurred can be easily incorporated with an Update . Many revalidations fall somewhere betw een those two extremes, with some deficiencies that can be most efficiently repaired by using the Redo approach while the balance of the PHA is Updated . Thus, in practice, two approaches are discretely applied in varying proportions; they are combined, not blended. Thus, Update is applied to those portions of the PHA that are being preserved or edited

494 PROCESS SAFETY FOR ENGINEERS: AN INTRODUCTION Table E.3 continued Threshold Release Category Material Hazard Classification Option 1 Material Hazard Classification Option 2 Threshold Quantity (outdoor) Threshold Quantity (indoorb) TRC-7 Liquids with Flash Point ≥ 23 °C (73 °F) and < 60 °C (140 °F) H226 Flammable liquid and vapor, Flammable liquids (cat 3) Tier 1: ≥ 2000 kg (4400 lb) or ≥ 14 oil bbl Tier 2: ≥ 200 kg (440 lb) or ≥ 1.4 oil bbl Tier 1: ≥ 200 kg (440 lb) or ≥ 1.4 oil bbl Tier 2: ≥ 100 kg (220 lb) or ≥ 0.7 oil bbl Liquids with Flash Point > 60 °C (140 °F) released at a temperature at or above Flash Point H227 Combustible liquid, Flammable liquids (cat 4) [**R eleased at or above flashpoint**] Liquids with Flash Point > 93 °C (200 °F) released at a temperature at or above Flash Point Crude Oil <15 API Gravity (unless actual flashpoint available) Crude Oil <15 API Gravity (unless actual flashpoint available) UNDG Class 2, Division 2.2 (non- flammable, non- toxic gases) excluding air H270 May cause or intensify fire; oxidizer Oxidizing gases (cat1) UNDG Class 2, Division 2.2 (non- flammable, non-toxic gases) excluding air Other Packing Group III Materials (excluding acids/bases) H272 May intensify fire; oxidizer, Oxidizing liquids and Oxidizing solids (cat 2,3) H312 Harmful in contact with skin, Acute toxicity, dermal (cat 4) TRC-8 Liquids with Flash Point > 60 °C (140 °F) and < 93 °C (200 °F) released at a temperature below Flash Point H227 Combustible liquid, Flammable liquids (cat 4) [**Released below flashpoint**] Tier 1: N/A Tier 2: ≥ 1000 kg (2200 lb) or ≥ 7 oil bbl Tier 1: N/A Tier 2: ≥ 500 kg (1100 lb) or ≥ 3.5 oil bbl Strong acids/bases (see definition 3.1) H314 Causes severe skin burns, Skin corrosion/irritation (cat 1A) H370 Causes damage to organs, Specific target orga n toxicity, single exposure (cat 1)

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

5IF
"NFSJDBO
*OTUJUVUF
PG
$IFNJDBM
&OHJOFFST 261 11 THE IM PACT OF HUM AN FACTORS “For a long time, people were sayi ng that most accidents were due to human error and this is true in a sense but it’s not very helpful. It’s a bit like saying that falls are due to gravity.” —Trevor Kletz Humans are involved in all aspects of the workplace. Humans manage facilities, design equipment, operate equipment, and maintain equipment. Yet historically, incident investigators have overlooked or provided cursory treatment of human factor contributions to incident causation. Contributions made by mechanical issues related to pressure vessel failures, pipe leaks, process upsets, mitigation system ma lfunctions, etc. are often readily identified. However, the real difficulty is to answer why these deficiencies occurred, and the answer is often rela ted to human behavior. For instance, a broken shaft may be obvious but to identify why the shaft broke may involve more rigorous examination. Were comp any inspection, material selection, operational controls, production proc edures, standards, priorities, etc. contributing factors? The shaf t may have broken due to poor supervision of operations or maintenance procedures, an engineering design that made it all but impossible to inspect the shaft, material selection that is no longer compatible with current production rate s, etc. All underlying factors should be probed for why it happened. Meaningful solu tions can be developed only after the investigator understands the true under lying causes. In many investigations, however, the why as it relates to huma n factors is sometimes underdeveloped. Incident investigation teams should attempt to determine what management system improvements coul d be made to remedy the particular human performance problem asso ciated with the incident under investigation. Oversimplifying a human performance to “human error” is an easy mistake to make but can be avoi ded if proper technique is used. In almost every case, there are underlying reasons for the human performance beyond the simple assumption that the worker failed to follow procedure. A system failure, design flaw, incorrect procedure, workload imbalance, or training deficiency may be the foundation of the performance problem. A good root cause identification proc ess should identi fy the underlying reasons. A good investigation recommendation seeks to set up the human for future success.

72 PROCESS SAFETY IN UPSTREAM OIL & GAS ●Kolskaya jackup rig, Russia 2011, 53 fatalities Offshore, harsh weather events might re sult in rupture of the drilling riser due to drill vessel movement or collision with a service vessel. Rupture of seabed pipework can be caused by subsea or loop currents, seabed movement, or collisions with dragged objects from facilities that have lost station. Production riser failures and subsea infrastructure vulnerabilities are more critical during the production phase rather than during well construction. This discussion is provided in Chapter 6. Onshore, harsh weather can also cause hazards to onshore facilities, such as toppling of poorly anchored land rigs. Key Process Safety Measure(s) Safe Work Practices: Many companies establish a weather window both onshore and offshore to limit operations such as use of cranes, people transfer by boat, and helicopter operations that might be adversel y affected by strong winds, high seas, etc. Emergency Management : It is common in areas subject to hurricanes or typhoons for MODUs to be moved before a storm stri kes, and this reduces personnel risks due to structural failure or sinking. The North Sea, Canada, and Alaska all have long periods of harsh weather, but not with as strong winds as during a hurricane, and evacuation or rig relocation is not generally required. 4.2.12 SIMOPS Risks Ineffective management of Simultaneous Operations (SIMOPS) can introduce new and significant risks versus each operation on its own. SIMOPS occurs when two or more separate operations are occurring that can interact in potentially unexpected ways and create a hazard. Examples are drilling a well while producing from other wells in close proximity at the same time and conducting hot work on a production facility while drilling into a hydrocarbon b earing zone. Refer to Section 5.2.5 for more details on the topic of SIMOPS. Key Process Safety Measure(s) See Section 5.2.5 for a listing of process safety measures. 4.3 APPLYING PROCESS SAF ETY METHODS IN WELL CONSTRUCTION Various process safety methods and tools are applicable to upstream operations. This section outlines a few of the more important tools relevant to managing the risks of well construction. To avoid repetition, the methods for well construction, onshore production, offshore production, and the design activity (Chapters 4, 5, 6, and 7, respectively)

INVESTIGATION M ETHODOLOGIES 31 of a n a l y s i s i s u p t o t h e group and does not always ensure reaching root causes. 3.1.5 Process of Elimination Process of elimination is another tool that can be used after brainstorming, as well as in structured approaches, to arrive at causal factors. Process of elimination is an integral part of scientific methodologies. It is valid to eliminate (disprove) hypotheses based on information obtained during an investigation. However, it is not sufficient to conclude that the one remaining hypothesis, for which there is no support, is the cause just because all other hypotheses have been elimin ated (NFPA 921, 2017). Any hypothesis must have a factual basis including evidence, observations, analysis and testing. Readers are ca utioned that process of elimination alone is not sufficient to reach a cause determination. 3.1.6 Timelines Most methodologies make use of a chronological list of events and conditions leading up to the incident. While a variety of formats have been used by investigation teams, the basic concept of a timeline remains unchanged (see Se ction 6.2.1). 3.1.7 Sequence Diagrams Several investigative tools employing graphic displays of incidents have been developed, but only a few are us ed in the chemic al industry. Although diagrams and charts had been in us e before 1970 to depict a sequence of events, the National Transportation Safety Board (NTSB) introduced Multilinear Event Sequencing (MES) concepts in the early 1970s to analyze and describe incidents. Another method is the Sequentially Timed Events Plot (STEP) (Benner, 2000; Hendrick, 1987). MES and STEP were originally developed for incidents othe r than process incidents and are discussed in more detail below. Multilinear Events Sequencing (MES) When applying the MES tool, investigators convert observed data into events and arrange the events on a matrix with time and actor coordinates. An event is defined as one actor plus one action. Actors can be people or things, and actions are what the actors did. As da ta defining an actor and what the actor did are acquired, each new event is posi tioned on its actor row on the matrix and positioned horizontally under the ti me it started. This displays what

Piping and Instrumentation Diagram Development 10 Were they not design work?” However, the word design in IFD has a specific meaning. It means “design by groups other than Process discipline. ” After issuing IFD P&IDs, the Process group lets other groups know that “my design is almost done and is firm, so all other groups can start their designs based on these (fairly) firm P&IDs. ” This is an important step because groups other than Process, including Instrumentation and Control, Piping, Mechanical, Electrical, and Civil can only start their (main) design based on a firm process design. If others start their design before a firmed‐up process design, it may end up being costly because every change in the process design will impact other groups’ designs. However, it should be noted that after the issue of IFD P&IDs, it is not the case that process design is finished because process still continues its work but at a different and slower pace. After IFD P&IDs, other groups do not expect the Process group to make big changes to the P&IDs. All the steps up to the IFD version of P&IDs fall under basic engineering or front‐end engineering and design (FEED), and all activities after IFD fall under detailed engineering. One important activity that should usually be done before the IFD version of P&IDs is the hazard and oper - ability study, or HAZOP . The HAZOP is an activity that seeks to identify flaws in design. It is a structured and systematic investigation technique to discover flaws in a specific process design. Generally, a HAZOP study is conducted in the form of a multiple‐day meeting with people from different groups present. The HAZOP study does not necessarily propose solu- tions to mitigate a process flaw; rather, it identifies the flaws and lists them in a HAZOP recommendation list. It is then the responsibility of the designer to address these flaws after the HAZOP meetings and close out the HAZOP issues. In an ideal world, the HAZOP would be done before the IFD version of P&IDs because the HAZOP meeting may impact the process design heavily, and it is a good idea to keep all the big process changes handled before the IFD version of P&IDs. However, some companies decide to have HAZOP meetings after IFD P&IDs for different reasons, including a tight schedule or a lack of detailed P&IDs from vendors. When a company wants to start a HAZOP study on a P&ID set, they may decide to do it on the latest and greatest version of the P&IDs, either officially issued or not. If the decision is to do the HAZOP on officially issued P&IDs, the revision of the P&IDs is Issued for HAZOP , or IFH. Not all companies issue an IFH version of P&IDs for the purpose of the HAZOP study, and instead they do the HAZOP on the latest available P&IDs. As was mentioned, all the activities after the IFD ver - sion of P&ID are part of detailed engineering. The client decides which activities should be done during the FEED stage of the project and which activities can be left for the detailed‐engineering stage. A client can decide how complete a P&ID should be at each milestone. However, there is one thing that is almost universally accepted: There should be no contact with vendors during the FEED stage of a project, and all vendor contacts can start during the detailed‐engineering stage to eliminate ven-dors’ involvement in process selection and design. This also means that all the information on the P&IDs up to the IFD version comes from the engineering com-pany’s experience and knowledge, and if there is a need for vendor information, the engineering company uses general vendor information or catalog information. Later, during detailed engineering, all the assumed vendor‐related information will be evaluated against the actual information provided by the selected vendor, and the information will be fine‐tuned. This concept shows the importance of previous expe- rience for P&ID development. There is one big exception to this rule and that is items with a long lead time. Long‐lead items are the equipment whose delivery to site is long (maybe 2 years or mor e). For long‐lead items, contact with the vendor can be started even during the early stages of the project or the IFR version of the P&IDs, which minimizes the impact of long‐lead items on the project schedule. Long‐lead items are generally the main equipment of a plant and are large or expensive ones. These may be different from plant to plant, but in general, equipment such as boilers, distillation towers, and furnaces can be considered long‐lead items. The next, and possibly last, P&ID milestone is IFC. Basically, from a P&ID point of view, the detailed‐engi-neering activity consists of improving the P&ID from the quality of IFD to the quality of IFC. As it was mentioned previously, during P&ID develop- ment, there could be several economic go or no‐go gates put in place by the client. At each of these “gates, ” the client needs a cost estimation report for the project to check if they want to continue the project, cancel it, or put it on hold. Therefore, there are usually three cost estimates during P&ID development. Each cost estimate can be done based on a copy of the P&ID set, or the cli-ent may ask for an official issue of the P&IDs for the pur - poses of cost estimation. For cost estimation purposes, an engineering company may issue P&IDs as Issued for Estimate, or IFE (Figure 2.2). IFR IFA IFD IFC Figure 2.1 The P&ID milest ones.

26. Learning from error and human performance 349 Figure 26-3 continued (adapted from [114] )1••1 2• •• 3••3 •• •• • •5 •• •••• •• • • • 7 •• ••2 4 6 Understand what motivated the action 7Understand how priorities set by supervision and management could have contributedWork withj individuals involved to reinforce the appropriate behaviors Encourage use of formal Continuous Improvement processConsult Human Resources for advice on whether disciplinary measures appropriate6Review and address what made it difficult to meet expectations in this caseWork with those involved to agree how this situation could be managed to meet expectations in the future Investigate factors which made the situation more likely (e.g. equipment, procedures, design, distractions, fatigue, etc.) •Where the individuals have a history of errors in different circumstances, consult Human Resoures for advice on appropriate performance improvement measuresEncourage a "stop and consult" attitudeWork with those involved to understand why this became the preferred approach 5Investigate why the practice became routing and how widespread it isCoach appropriate behavior with those involved Encourage use of formal Continuous Improvement processEncourage individuals involved to act as role-models for appropriate behavior Consult HSE team for advice on tackling group non-conformanceConsullt Human Resources for advice on whether disciplinary measures are appropriate4Investigate factors which triggered error or made it more likely (e.g. equipment, procedures, design, distractions, fatigue, etc.)Work with those involved to understand where other errors and problems could occur Identify tasks which would have a serious outcome in case of errorWhere the individuals have a history of errors in different circumstances, consult Human Resoures for advice on appropriate performance improvement measures Redesign tasks to eliminate and detect errors and recover without harmEncourage people to "stop and consult" when something is new Address selection, training, assessment and quantity of people required to fulfill the expectationsProvide appropriate traning assessment and resources for individuals involvedAssess and coach supervision and managers on leadershipDefine and test figure of authority's action with this process Clarify and verify expectations are met •Work with thouse involved to understand where there are misunderstandings or conflict in expectationsImprove management of procedures or consider alternate means of controlAddress conditions people work underWork with people involved 4 3 Now test supervisor / line manager / others contribution5

1 • Introduction 7 compiled from successful experien ce in many industrial organizations across a broad range of industries in different jurisdictions around the world. The approach continues to evolve today. For readers unfamiliar with these pillars and elements, re fer to Chapter 10 for an overview, as the discussions and lessons le arned from start-up and shut-down incidents will be based on the CCPS RBPS foundation. In addition, note that there is no hyphen— by design —when the “CCPS Risk Based Process Safety (RBPS)” approach is being discussed. However, a Risk- Based Inspection (RBI) program, for example, used in maintenance- related efforts applies a hyphen between risk and based. 1.6 Incident discussions and guidance This guideline uses in cidents, from both pu blished investigation reports and internal company inci dent information, that provide details on what went well and what went wrong during the start-up or shut-down. The anonymous company incidents submitted to this book or located in generic incident databases are presented for sharing. Everyone learns from experience. The goal of sharing incidents is to prevent others from learning from the bad experience the hard way. The collective global goal is to reduce the process safety risks an d prevent incidents that cause harm to people, the environment, and the business. As the cases presented are reviewed, it should be noted that: 1. The guidance—these learnings—are framed within the CCPS Risk Based Process Safety (RBPS) approach described in Chapter 10, And most importantly: 2. The year of these cases is noted since those that occurred before the publication of the initial CCPS RBPS guidance in 2007

108 Guidelines for Revalidating a Process Hazard Analysis * Note: In the example in Table 6-1, the revalidation team has decided to Update the Facility Siting Checklist but Redo the Human Fa ctors Checklist. It is not typical for the same revalidation to Update one checklist and Redo another, but it is certainly possible. For example, in this case, the fa cility may have made si gnificant progress in its human factors programs and therefore want ed to perform this checklist from the beginning (Redo) to gauge progress without influence from the previous checklist responses. 6.1.2 Selecting Team Members The same issues and considerations di scussed in Section 3.1.2 regarding the qualifications of the prior PHA team ar e equally relevant to the revalidation team. At a minimum, the revalidation team must have the same set of skills and qualifications required by local regu lations for any PHA team. Globally, the minimum required skills usually include: • Engineering expertise • Operations expertise • Expertise in the analysis method being used (e.g., HAZOP) Beyond that, depending upon the specif ic scope of the PHA, there may be requirements (See Section 2.2.) for additional team skills, such as: • Maintenance expertise • Instrumentation and controls expertise • Process chemistry expertise • Human factors expertise • Expertise in risk analysis (e.g., LOPA) The team composition will likely have to be modified slightly when the complementary analyses are performed, part icularly the facility siting checklists. Team members with knowledge of th e emergency response plan and its execution, communication systems, an d other non-process knowledge will be Multiple PHAs for a Single Process If multiple PHAs exist for a single process that was started up in stages, the PHA revalidation offers the opportunity to combine these PHAs and ensure no scenarios were missed.

48 | 2 Core Principles of Process Safety ignoring of any outside advice and a form of self-isolation with respect to new or different process safety ideas. Maintaining a sense of vulnerability also requires that organizations be vigilant for new or previously unrecognized causes of process safety incidents. When new issues are discovered organizations should then extend the scope and application of their PSM S to cover these new issues. Examples of such extensions are: Many organizations have extended the use of their Management of Change (M OC) program to include certain types of organization and personnel changes. Organizational Management of Change (OMOC) was not originally part of the intent of MOC, but many facilities and com panies have recognized how turnover in certain jobs, overall staffing, and other sim ilar changes can affect the quality of the PSMS.Many organizations voluntarily perform Layer of Protection Analysis (LOPA) as part of their Hazard Identification and Risk Analysis (HIRA)/Process Hazard Analysis (PHA) to provide additional study of the number and quality of their safeguards for possible hazard scenarios that m eet certain risk criteria as measured in their HIRAs/PHAs. Com placency and an uncontrolled can-do attitude are part of hum an nature. They can be reinforced by the social conditions within an organization, but mostly they represent human traits that are common to all people to some degree. Combatting these characteristics can be difficult, even when the risks are high. When com peting pressures, such as production are also present, a can- do attitude can become a com plicating negative trait.• •

Fundamentals of Instrumentation and Control 259 non‐sealed flows are the flows where liquid is in a por - tion of the liquid conductor. For example the flow in open channels is a always non‐sealed type conductor. Table  13.19 gives a non‐exhaustive list of common flow meters for sealed conductors. Liquid flow in non‐sealed conductors also needs to be measured. The fundamental principle of measuring liq-uid flow in open channels is to measure the liquid level, and then convert the level to the corresponding flow rate. Therefore the flow sensors in open channels are nothing other than level sensors.Table  13.20 shows two types of open channel flow sensor. Flow meter arrangements: there are some cases that a flow meter needs a fluid velocity higher than the pipe flow velocity to be able to sense the flow. In such cases the flow meter cannot be placed directly on the pipe. The pipe size should be shrunk to a smaller size (possibly one or two sizes smaller than the pipe size) and then the flow meter can be installed correctly. To do that a combination of a reducer–enlarger can be used (Figure 13.30).Table 13.18 Lev el sensors. Type P&ID schematic Unique advantage Unique disadvantage Application Contact typeStatic pressure type LI 15Simple system Relays on density of the liquid that could be changingBy default choice Bubbler type LT 15Good choice for slurry and precipitating liquids ●Needs utility air connection ●Relies on density of the liquidSlurry liquids, water tanks Float type LISimple to operate Limited range Small tanks Non‐Contact typeUltrasonic type LT 15 ●No contact with the process materialThe atmosphere should be transparent free of dust and liquid drops ●The atmosphere should be with fairly constant composition ●Liquid surface should be free of ripples and foams ●Relatively inexpensive Radar type(microwave) LT 15 ●No contact with the process material ●Elevation of interface in multi‐layered fluids (oily water, water and sludge) can be measured. ●The measurement is not affected by the atmosphere condition ●More expensive Last resort

EDUCATION FOR MANAGING ABNORMAL SITUATIONS 99 Example Incident 4.4 – Air Fran ce AF 447 Crash, June 2009 (cont.) The report goes further to identify several contributing factors related to crew recognition and management of the situation, including: Incorrect actions taken by the crew upon auto pilot disconnection destabilized the flight path. Failure of the crew to initiate procedures upon losing flight speed. Failure of the crew to recognize the stall position in a timely manner. This incident prompted increased re porting from airline operators of similar problems with pitot tubes in heavily icing conditions and led to a prohibition of certain models of probes as a precautionary measure. In addition, the maintenance interval for pitot cleaning was reduced. Lessons learned in relation to abnormal situation management: Among the 25 safety recommendations issued by BEA, the following were made with regards to crew instruction & training: Knowledge: Improve crew knowledge of aircraft systems and changes in their characteristics in degraded or unusual situations Skill Development and Training: Improve flight simulators for a realistic simulation of abnormal situations.

Ancillary Systems and Additional Considerations 397 The off‐line corrosion monitoring program is a type of off‐line monitoring program, which was discussed in the previous section. An off‐line corrosion monitoring program is a set of hardware (system) and procedures to measure and report the “corrosion rate” in a specific location of a plant. A corrosion monitoring program specifies the required corrosion coupon system, procedures to trans - fer the coupon to the lab, applying a test procedure to measure the corrosion rate, and sending the results to the appropriate parties. The only footprint of a corrosion monitoring program on the P&ID is the corrosion coupon. Therefore our discussion is limited to the corrosion coupon. A “corrosion coupon” is a piece of material with the spe- cific shape and specific weight. The material of the corro-sion coupon is generally selected the same material of pipe or equipment. Corrosion coupons are located in locations to provide meaningful information. They are generally located at locations that are suspected as having a high corrosion rate, upstream of some critical equipment and mainly on pipes. The locations where the fluid is stagnant are not good candidates for a corrosion coupon. They should also be located in accessible, safe, and comfortable locations for operators.Figure 18.23 shows two schematics of a corrosion cou- pon on P&IDs. 18.7 Impact of the Plant Model on the P&ID Generally speaking the plant model is developed based on the P&ID and other documents like the plot plan. However, there are cases that the route is reversed, which means the P&ID needs to be changed because of the plant model (Figure 18.24). Space constraints can dictate changes on the P&ID but not all of them are acceptable from a process viewpoint. Figure 18.24 Direc t route and reverse route.CC SP. 316SP 102SP 87 CC 1/uni2033 FP Figure 18.23 Cor rosion coupon.

KEY RELEVANCE TO PROCESS PLANT OPERATIONS 63 Example Incident 3.10 – Tower Flooding Poor separation in a distillation column can be the result of insufficient flows of vapor/liquid due to low reflux and/or low boil-up from the reboiler. In one typical case, the tower was providing poor separation of the individual components. The operator responded by adding more heat to the reboiler to provide more energy for the separation. This increased the tower overhead temp erature/flow and a corresponding high level was reached in the overhead receiver, which was then addressed by adding more reflux. However, the separation of components in the tower worsened, and so the control room operator repeated the more-reboiler/more-ref lux cycle over a period of an hour, with progressively worse results. This was a case of tower ‘flooding’ in which liquid and/or vapor rates are too high. This leads to excessive liquid on individual trays, resulting in poor liquid-vapor disengagemen t, high pressure drop, and poor overall separation of components. The solution was to remove the heat source to the tower, let everything fall to the bottom, an d start it back up again. Lessons learned in relation to abnormal situation management: Understanding Abnormal Situations: This is less about having a procedure to address the issue, and more about abnormal situation identification and training in the principles of distillation column operation. Knowledge and Skill Development: Ad ditional knowledge and training may have prevented the lead-up to the flooding. Learning from others—experience is a valuab le knowledge-sharing tool. Cultural influences may play a factor in choosing the interface between the control panel operator an d the control panel, in order to allow more rapid detection of an abnormal situation. The following discussion between a licensor’s re presentative and the control panel operators on a new unit illustrates this concept in Example Incident 3.11.

B.2 Advancing Safety in the Oil and Gas Industry – Statement on Safety Culture |271 Maintenance activities not prioritized and executed as planned. Processes and procedures not routinely assessed for accuracy, completeness, or effectiveness. B.3 References B .1 Canadian National Energy Board (CNEB ), Advancing Safety in The Oil and Gas Industry - Statement on Safety Culture, 2012.• •

EVIDEN CE IDEN TIFICATION , COLLECTION & M ANAGEM ENT 143 in the work area is also critical. Fina lly, it is common for the team to work extended hours in a variety of weather conditions. The team leader should watch for signs of fatigue, as this can affect the safety of the team members and the quality of the investigation. The team leader should also set a ri gorous standard for consistent and proper use of personal protective equipment and team members should approach each task with awareness and a high degree of caution to help prevent injuries and minimize unnecessary hazard exposure. If the incident has led to an interruption of production, the investigation team may have to deal with pres sure from management to resume operation. For smaller incidents, prod uction may have resumed before the start of the investigation, or it could have continued throughout the occurrence if process integrity was not compromised. In these cases, the investigation team may have to rely on the support of operations and maintenance personnel to help with initial acquisition and preservation of some of the data from the operating plant. The investigation team should provide guidance to these personnel r egarding the key issu es of evidence preservation. This may in clude an explanation on the protocols that have to be used, as discussed in 8.3.2. For major investigations, production may be interrupted for some period of time following the incident. Pressures to resume production may be apparent from the start of the investigat ion and may increase as time passes. For example, once one or two causal factors are identified, facility staff may pressure the team to release the system for production. They perceive that “the cause” of the occurrence has been iden tified, and therefore the investigation must be nearly complete. However, the team usually has a great deal of work to perform to iden tify the remaining causal factors and the root causes of the occurrence. The team leader may need to oppose requests to conduct repairs or resume operations until th e required data is collected and compiled. In some cases, the process, or portions of the process, may be released back to the manufacturing management for repair and resumption of operations before th e collection of data is complete. The decision to release these portions, begin cleanup, and start rebuilding should be based on a number of factors including: Is it safe to reenter the area? Have sufficient data been collected? Has sufficient knowledge been ga ined about the causes of the incident to ensure the safety of the operation?

CONSEQUENCE ANALYSIS 289 do not have sufficient stored energy to repres ent a threat from shock wave beyond the plant boundaries. However, these types of incidents ca n result in domino effects particularly from the effects of the projectiles produced. Several different methods can be used to estima te projectile size and trajectory, but these have a high uncertainty as the specific way in which a vessel will fail is not known. These methods are more suited for accident investigat ions, where the number, size and location of the fragments is known. Very few Chemical Pr ocess Quantitative Risk Assessment (CPQRA) studies have incorporated projectile effects on a quantitative basis. BLEVE and Fireball A BLEVE is a sudden release of a large mass of pressurized superheated liquid to the atmosphere and was discussed in Chapter 4. A BLEVE occurs when an external fire, either through thermal radiation or direct flame impingement, weakens the vessel above the liquid level as the vapor space provides less internal cooling and the vessel wall fails, typically when it reaches 550°C (1022°F). As hydr ocarbon fires burn at 1150°C (2102°F), there is only a short period, often only 15 minutes, before a BLEVE ma y occur. Note at 550°C (1022°F) the ultimate tensile strength of steel is reduced by half and this fully exhausts the design safety factor in shell thickness. A special type of BLEVE involv es flammable materials, such as LPG. At the beginning of a BLEVE, a fireball is formed qu ickly due to the rapid ejection of flammable material as it flashes due to depressurization of the vessel. Ignition occurs as the cause of the failure is an external fire. This is followed by a much slower rise in the fireball due to buoyancy of the heated gases. Methods to determine consequences from a BLEVE are discussed in CCPS Guidelines for Chemical Processe s Quantitative Risk Assessment and CCPS Vapor Cloud Explosion, Pressure Vessel Burst, BLEVE and Flash Fire Hazards . (CCPS 1999 and CCPS 2010) BLEVE models are a blend of empirical correlatio ns (for size, duration, and radiant fraction) and more fundamental relationships (for view fa ctor and transmissivity). BLEVE models require the material properties (heat of combustion an d vapor pressure), the mass of material, and atmospheric humidity. Fragment models are fairly simplistic and require vessel volume and vapor pressure. The output of a BLEVE model is usually the radiant flux level and duration. BLEVE models require some care in application, as errors in surface flux, view factor, or transmissivity can lead to significant error. A BLEVE and fireball are significant threats to firefighters as they approach an emergency scene. Understanding when potential for a BLEVE exists and planning an appropriate response are important to the safety of the firefighters. Water spray can be used to cool the area of flame impingement – if the water can be applied wi thout putting firefighters at risk. Protecting vessels that could be exposed to external flam e impingement with fireproofing is also a means to reduce the vessel wall heating and delay or prevent a BLEVE. Vapor Cloud Explosions (VCE) Dispersion analysis can be used to define the ex tent of the flammable portion of a vapor cloud. If the vapor cloud is ignited before it is dilute d below its lower flammability limit, a VCE or flash fire will occur. Vapor clouds are normally ignited at the edge as they drift to an ignition source such as a fired heater or a vehicle. The effect of ignition is to terminate further spread of the cloud in

360 Human Factors Handbook Individuals should engage in self-reflection following involvement in an incident to assess what went wrong, and what they did or could have done differently. The self-reflection can take the form of a group discussion, as points discussed with others can offer additional insight into learning from incidents. Learning culture and psychological safety are required for individuals to engage in an open and honest discussion. Lessons learned should be applied in pr actice. Individuals should be open to change – that is, that they are willing and interested in changing their thinking and behavior. There should be a sense in an organization of “chronic unease” and readiness to change. It is important to ma intain ‘chronic unease’ at a certain level, to keep people thinking about potential situations and be alert to danger (what could go wrong). 26.7.2 Tools for learning Learning from incidents is a crucial elem ent of process safety. The learning does not stop once the incident investigation is completed (i.e., root causes were identified and improvements proposed). Th e lessons learned from error should be shared and applied to ensure employees’ full understanding of the issues and in order that change may begin. Lesson sharing includes: • Immediate incident notification and interim updates. • Lessons learned from an incident investigation. • Lessons learned from a review of incident trends. Information and updates shared should be written in a simple comprehensible format, and should contain incident descriptions and actions taken. This information should also offer feedback on the effectiveness of the undertaken actions. This is shown in more depth in Table 26-4. Chronic unease is the experience of unease and discomfort regarding the management of risks. It is defined as a healthy scepticism about the true standard of safety performance. It is about probing deeper and understanding the risks, not just assuming that just because systems are in place everything will be “ok”. See Chapter 18 for more information on psychological safety.

6 • Recovery 103 controls. It is the combination of the effectively designed, implemented, and sustained controls that helps reduce the process safety risks. Many facilities have started to anticipate security threats and prevent or mitigate the consequences once the facility perimeter has been breached. In particular, geopolit ical unrest and wars in countries around the world have made terrorist targets out of refineries, natural gas facilities, and other manufactu rers using toxic, flammable, or explosive materials. Cyberattacks have also been successful in reducing productivity as well as jeopardizing the process safety of processes handling hazardous materials and energies. For this reason, there are many defensive tactics that can be implemented to help reduce the likelihood of an attack , and in the event the facility is targeted, designing the emergency response capability to reduce the consequences of the release is key. Additional details are provided in other publications [55] [56, pp. Chapter 35, pp. 2-8] [57] [58]. 6.4 Managing abnormal operations Abnormal operations are defined as “the operating mode that occurs during normal operations when there is a process upset and the process conditions deviate from the normal operating conditions” (Table 2.2). These upsets to the normal processing conditions may result in deviations that can range from relatively small deviations to large deviations that may become to o difficult for the recovery efforts to manage. Thus, a “normal” operatio ns ventures into the “abnormal” operations territory, and if the process upset begins to or exceeds the safe operating limits, “emergency” operations are implemented (Figure 6.2). This section describe s how abnormal operations can be effectively managed for successful recovery efforts, beginning with a description of the abnormal si tuation (Section 6.4.1).

330 PROCESS SAFETY FOR ENGINEERS: AN INTRODUCTION 4. The HAZOP identified safeguards. However, none of the safeguards intended to prevent the tank overflow met the criteria of an IPL. A dike is provided that would contain the overflow and prevent a fire over a large area. This dike meets the IPL criteria and has a probability of failure on demand of 1 x 10-2. 5. Combining the initiating event frequency of 1 x 10-1 per year and the probability of failure of the dike of 1 x 10-2, the frequency is 1 x 10-3 per year. 6. Using the company’s criteria, this would require risk reduction typically by the implementation of additional IPLs. The anal yst could then consider the addition of a safety instrumented system to prevent the overflow. (Safety instrumented systems are discussed in Chapter 15.) With the addition of such a system with a probability of failure on demand of 1 x 10-2, then the total probability of failure on demand would then be 1 x 10-4 and the frequency of the mitigated scenario, 1 x 10-5 per year. Per the company’s risk criteria, this is a tolerable risk. This is a very simple example. In reality, many factors can make a LOPA more complex. These include consideration of the following. The number of potential initiating causes in various modes of operation such that they are all accurately included and not double counted. Inclusion of IPLs that truly meet the requirements for an IPL. The use of enabling conditions or condition modifiers that are required to realize the consequence of concern, e.g. a process in recycle mode or the probability of maintenance personnel being present, resp ectively. A full description of these are given in Guidelines for Enabling Conditions and Conditional Modifiers in Layers of Protection Analysis. (CCPS 2013) What a New Engineer Might Do New engineers frequently participate in small uni t projects or major capi tal projects which can include the use of hazard identification studies and risk assessments. This work can include plotting risks on a risk matrix in support of pr ioritization of resources all the way to gathering data for use in a QRA. As with consequence an alysis, using the best data possible supports a quality QRA. Researching data sources to find cu rrent, relevant data is an important activity that is frequently supported by new engineers. One thing that new engineers do not typically do is conduct detailed QRAs or analyze their results. The QRA results can be significantly influenced by the data, assumptions, and parameters used in the modeling. It can be easy to generate results, and it sometimes takes an experienced analyst to recognize that something is amiss in those results. Seeking the advice and review of an experienced analyst is a good approach in building risk analysis skills.

204 8.3 American National Standard s Institute/Instrument Society of America (ANSI/ISA), Application of Safety Instrumented Systems for the Process Industries, ANSI/ISA-84. 00.01-2004, Instrument Society of America, 2004. 8.4 American Petroleum Institute, Management of Hazards Associated with Location of Process Plant Portable Buildings, API-753, 2012. 8.5 American Petroleum Institute, Management of Hazards Associated with Location of Process Plant Tents, API-756, 2014. 8.6 American Society for Testing and Materials (ASTM International). CHETAH: Chemical Thermodynamic & Energy Release Evaluation, Ver 10.0, 2016. 8.7 American Society for Testing and Materials (ASTM International), E2012-06 Standard Guide for the Preparation of a Binary Chemical Compatibility Chart, ASTM International, 2006. 8.8 Bodor, N., Design of biologic ally safer chemicals, Chemtech, 25 (10), 22-32, 1995. 8.9 Bretherick, L., Handbook of Reactive Chemical Hazards, 5th Edition, London, UK: Butterworths, 1995. 8.10 Burch, W., Process modi fications and new chemicals, Chemical Engineering Progress, 82 (4) 5-8, 1986. 8.11 Center for Chemical Process Safety (CCPS 1999), Avoiding Static Ignition Hazards in Chemical Operations. American Institute of Chemical Engineers, 1999. 8.12 Center for Chemical Process Safety (CCPS 2000). Guidelines for Chemical Proce ss Quantitative Risk Analysis 2 nd Ed., American Institute of Chemical Engineers, 2000. 8.13 Center for Chemical Process Safety (CCPS 1995), Guidelines for Chemical Reactivity Evaluation and Application to Process Design. American Institute of Chemical Engineers, 1995.

Table C-1 continued HF Competency Performance/ Knowledge Criteria Level 1 - Operator Level 2 - Supervisor* Level 3 - Manager** Operational competency Supporting operational competency Understand the process of determining competency Is involved in the process of determining competency for safety critical tasks Can determine competency requirements by conducting task analysis, perform learning needs analysis, and select assessment learning methods Able to review the effectiveness of competency process Identify training requirements Is able to identify training needs requirements Understands the importance of training Recognizes personal need for training Can recognize personal and team need for training, by conducting training needs analysis Can suggest/ recommend forms of training Able to assess effectiveness of training Develop Competency Understands the need and process of developing competency Understands the importance and process of developing and maintaining competency Monitors and advises on the importance of developing competency Able to assess effectiveness of strategies to maintain and develop competency

1 • Introduction 9 Figure 1.1 Three types of facility operations and their corresponding transient operating modes. (Adapted from [15, p. 22] ) The ten transient operating modes discussed in this guideline are introduced in Table 1.1. The Appe ndix provides the summary of a detailed incident review focusing on published incidents that occurred during the transient operating modes listed in Table 1.1, and includes additional guidance on how to mo re effectively manage unexpected situations, especially those than may cause or may happen during transient operating modes.

80 | 3 Leadership for Process Safety Culture Within the Organizational Structure Trait theories, which tend to reinforce the idea that leaders are born not m ade, m ight help in the selection of leaders, but they are less useful for developing leaders. One ideal leadership style would not suit all circumstances. Many theories assert that leaders can change behavior to fit circumstances at will. However, many find it hard to do in practice, due to unconscious beliefs, fears or ingrained habits. Thus, he argued, leaders need to work on their inner psychology. None of the older theories successfully address the challenge of developing “leadership presence,” that “certain something” in leaders that comm ands attention, inspires people, wins their trust, and makes followers want to work with them. Leadership of Process Safety As noted above, process safety leadership differs from general leadership only in focus. But leaders have struggled to include process safety in their focus. Stricoff (Ref 3.14) stated: “The connection between leadership and process safety has not always been clear. Leaders often struggle to identify how or whether they affect process safety outcom es. The head of Transocean, for exam ple, recently testified that while he wished his crew had done m ore to prevent the 2010 Deepwater Horizon disaster, his organization had found no failure of management. To m any leaders, the idea that some events will ‘just happen’ despite leadership efforts is (and should be) deeply troubling. “New research is showing that leaders play a critical and very specific role in catastrophic event prevention through their effect on culture. Of the 10 most recent events investigated by the U.S. Chem ical Safety Board, each had • • • •

Piping and Instrumentation Diagram Development 42 is followed by the B/L P&IDs. The third and the main part of a set of P&IDs is the sheets that relate to the main process of the plant, or system P&IDs. The process P&IDs show the route that the raw materials follow to be converted into product(s). Utility and auxiliary P&IDs are the last groups of P&IDs.In a design project, with this sequence, the network P&IDs are mainly dependent on the plot plan and the location of equipment should be finalized to be able to develop utility and interconnecting P&IDs. All or majority of auxiliary P&IDs are created during the detailed engineering stage of projects, when the development of other P&IDs are near the end. When designing a process plant, all the P&ID sheets of the plant should ideally be issued at once (simultaneously) as Issued for Construction (IFC). In the real world, how - ever, such a thing may not be possible and figuring which P&IDs to issue first depends on the critical nature of the construction for the items on a specific P&ID sheet. Generally speaking, pipe rack P&IDs should be issued first. P&IDs of large items are issued early, too, if they are not the vendor responsibility. The other high priority P&IDs are the ones for utility generation systems. Because the utility systems are usually the first systems that come into operation for commissioning, they should be constructed first. However if the utility generation systems are generic with low complexity, then they can be issued with lesser critically important P&IDs. Such a priority in issuing P&IDs is for the IFC version only. A P&ID set can be named based on not only its con- tent but also its purpose. Each of the P&IDs mentioned thus far can be for a greenfield project or brownfield project. Brownfield projects can be upgrading or opti-mizing projects. In brownfield projects each of the dis - cussed P&IDs can be converted to demolition P&ID and tie‐in P&ID. In demolition P&IDs, the part of equipment that needs to be removed from the plant is specified somehow (e.g. hatched lines). In tie‐in P&IDs, different tie‐ins are added to show the pipes that need to be con-nected to a new item in the plant.4.6 P&IDs Pr epared in Engineering Companies Compared to Manufacturing or Fabricating Companies The P&IDs prepared by manufacturing or fabricating companies can be different than the P&IDs prepared by engineering companies. P&IDs created by engineering companies are prepared for the purpose of erection, installation, and start‐up and should be kept in the plant for the life of the plant, whereas a P&ID made by manufacturing companies are prepared solely for construction. The differences can be summarized as follows: 1) The P&I D set by manufacturing companies generally do not have auxiliary P&IDs. All the required details are shown on the main P&ID set. In many cases, ven-dors are not responsible for auxiliary systems. 2) The P&I D set by manufacturing companies tend to have more technical information. It is not strange to see the pressure range of a pressure gauge on a P&ID prepared by a manufacturing company. This is because manufacturing companies try to put as much as information on their P&IDs for other disciplines. Auxiliary Utility Interconnecting, B/L Process Legend, list Figure 4.30 A giv en set of P&IDs. Product Raw material Process5 1 Figure 4.31 Pr ocess P&IDs within a complete set. P–2600–1/2 Grundfos vertical inline Centrifugal charge pump CRN 64–2–2, 4/uni2033–150#RF, 339 USGPM 125 ft head c/w baldor electric motor, 15HP, 3450RPM, 254TC frame, 3PH/208–230/460V/60Hz, Class I div II (TEFC) Maximum discharge pressure: 190 ft headO Figure 4.32 Sample pump callout in a manufac turer P&ID.

6.2 Assess the Organization’s Pr ocess Safety Culture |211 conference rooms usually used by managem ent. The interview setting should be private, avoiding areas where others may be present or where passers-by may look in. When interviewing managers, obtain a brief understanding of titles, responsibilities, and reporting relationships. This will help the interviewer understand how the process safety culture flows through the organization and where roadblocks may exist. Individual interviews should generally be conducted by a single interviewer. This helps create a more trusting environment and avoids the potential for interviewees to feel ganged-up on by m ultiple interviewers. More than one interviewer could be used when interviewing executives, as they are less likely to be intimidated and it should be more time-efficient. Where hourly employees are accom panied by a union representative, talk to the representative in advance to request they provide support only and do not seek to influence the interviewee. Design the interview protocol to provide prom pts for the interviewer rather than detailed questions, and make it easy to record responses. This will allow the interviewer to focus on the interviewee rather than on the notes. Using a paper notebook or electronic tablet is generally the least intim idating to the interviewee. Since notes taken by this method will by nature be m inim al, a few undisturbed minutes following each interview should be planned to record additional notes and observations. The use of clipboards, though convenient, can convey the sense that the interviewee who is being evaluated, not the culture. Using a laptop com puter as the source of the notes should also be avoided, as the screen acts as a barrier between interviewer and interviewee. Audio or video recording of the interviews should be strictly avoided. Group interviews should include participants from the sam e level of the organization. This helps avoid potential reluctance to offer input in front of a supervisor. Interviewers should also be

188 PROCESS SAFETY FOR ENGINEERS: AN INTRODUCTION Figure 11.7. Schematic of centrifugal pump (Kelley) Figure 11.8. Single and double mechanical seals (Berg)

106 | 8 Landmark Incidents that Everyone Should Learn From Table 8.1 (Continued) Incidents with Key Findings that Everyone Should Know Incident Prominent Findings and Causal Factors Piper Alpha, North Sea off Aberdeen, Scotland, 1987 • culture • management of change • safe work practices • conduct of operations • Shut-down authority Texas City, TX, USA, 2005 • facility siting (e.g., of trailers) • culture • conduct of Operations • operating Procedures • asset integrity • operational readiness • safe design Buncefield, Hertfordshire, UK, 2005 • HIRA (insufficient layers of protection) • asset integrity • conduct of operations • vapor cloud explosions West, TX, USA, 2013 • stakeholder outreach • facility siting • emergency management • HIRA • chemical reactivity hazards NASA Space Shuttles, USA; Challenger 1986 and Columbia 2003 • culture • conduct of operations • HIRA Fukushima Daiichi, Japan, 2011 • culture • preparation for natural disasters • emergency management • stakeholder outreach 8.1 Flixborough, North Lincolnshire, UK, 1974 When temporary bypass piping failed, a vapor cloud explosion resulted in the deaths of 28 workers (UKDOE 1975). Many other workers suffered injuries, and significant onsite and offsite property damage occurred. The temporary piping had been installed to bypass the fifth oxidation reactor in a chain of six, which had been removed for repair. See Appendix index entry J119

154 PROCESS SAFETY FOR ENGINEERS: AN INTRODUCTION Table 9.2 continued Tier 4 Process Safety Culture Process safety culture survey scores Process hazard analysis Percentage of total PHAs documenting use of complete Process Safety Information (PSI) during the PHA Number of PHA Recommendations Facility Siting Risk Assessments Percentage of total PHAs documenting Facility Siting risk assessments Operating Procedures and Maintenance Procedures Percentage of total number of operating or maintenance procedures reviewed/updated Asset Integrity Percentage of total inspections of safety critical equipment completed on time Percentage of time plant is in production with items of safety critical equipment in a failed state Process Safety Training and Competency Assurance Percentage of individuals who completed required process safety competency training on time Management of Change Percentage of MOCs that satisfie d all aspects of the site’s MOC procedure. Percentage of identified changes that used the site’s MOC procedure prior to making the change. Action Item Follow-up Percentage of process safety action items that are past due Fatigue Risk Management Amount of overtime Number of extended shifts What a New Engineer Might Do New engineers are frequently involved in the collation of performance metrics. The calculations should be accurate and use th e precise definitions provided to support comparison of performance as opposed to comparison of data anomalies. A new engineer should be very familiar with the relevant documents described in this chapter. A common responsibility of early career engineer s is to develop source models to calculate release amounts from various aperture releases, vessel overflows, and other loss of primary containment events. Once release amounts are known, the API RP 754 criterion for PSE is used to classify the incident as Tier 1, 2, 3, or near miss. Often, engineers face a short time frame to return classification due to company or regulatory requirements for reporting. Leading and lagging indicator data are tracked and the data analyzed to identify trends and make suggestions for improvement.

462 PROCESS SAFETY FOR ENGINEERS: AN INTRODUCTION Figure 23.1. CCPS Vision 20/20 What a New Engineer Might Do A new engineer can benefit from reviewing the CS B investigations and videos relevant to this chapter as listed in Appendix G. Tools Resources to support process safety culture include the following. CCPS Vision 20/20 Assessment Tool. This tool is intended to help a company assess its process safety implementation as compared to the Vision 20/20 elements. It can be used in various operating locations or parts of a business to compare implementation across the company. The tool is available at https://www.aiche.org/ccps/vision-2020 .

132 PROCESS SAFETY FOR ENGINEERS: AN INTRODUCTION power plants were located on the northeaste rn coast of Japan. Fukushima Daiichi was operated by Tokyo Electric Power Company (TEPCO). Refer to Figure 8.1. The Fukushima Daiichi design used boiling water reactors. The reactors were a closed loop system. Water boiled in the reactor producing stea m that drove turbines to generate electric power. The steam was then condensed using cold water from the ocean, and then fed back to the reactor again. Figure 8.1. Fukushima Daiichi nuclear reactor design (IAEA 2015) The Great East Japan Earthquake occurred at 4:46 PM. It was a magnitude 9.0 and lasted more than 2 minutes causing damage to struct ures and power infrastructure. Units 1, 2, and 3 were running at the time and shutdown automa tically due to the earthquake seismic motion. A tsunami was created by the earthquake with the waves arriving 40 minutes after the initial shock. A wave of 14 to 15 m (46 to 49 ft) o verwhelmed the Daiichi seawalls and flooded the site. This caused significant dama ge, loss of power, loss of cont rol, and eventual loss of reactor containment. Following the earthquake, TEPCO set up an emergency response center in Tokyo to manage the response and an on-site emergency re sponse center at the Daiichi site. Evacuation and shelter-in-place orders were issued over the next three days. After inserting the control rods (rods composed of chemical elements used to control the nuclear fission) to stop the reaction, heat co ntinued to be generated. Cooling systems were powered or controlled by electrical power. The earthquake damaged off-site power supply resulting in a total loss of power supply to the pl ant. This loss of power isolated the units from their turbines resulting in increased temperature and pressure in the reactors. The operators followed appropriate procedures for the earthq uake and loss of power in shutting down, isolating, and activating cooling systems. The incident progression is shown in Figure 8.2.

351 the attention devoted to it in the po litical arena, IS remains more of a philosophy than a codified proc ess with a well-established and understood framework for evaluati on and implementation. Both industry and regulators lack tools and measures to compare the inherent safety of different options or to determine what is “feasible.” Therefore, policy debates over how best to encourage IS continue to be frustrating for all concerned. 14.2 EXPERIENCE WITH INHERENT SAFETY PROVISIONS IN UNITED STATES REGULATIONS Unlike other process safety issues, IS is not easily regulated. For example, when the United States EPA promulgated its Risk Management Program (RMP) rule in 1996, so me commenters recommended the Agency require facilities to conduct “technology options analyses” to identify inherently safer approaches . The US EPA declined to do so, stating that: “PHA teams regularly suggest viable, effective (and inherently safer) alternatives for risk reduction, which may include features such as inventory reduction, material substitution, and process control changes. These changes are made as opportuniti es arise, without regulation or adoption of completely new and unproven process technologies. EPA does not believe that a requirement that sources conduct searches or analyses of alternative processing technologies for new or existing processes will produce additional benefits beyond those accruing to the rule already.” In 2017, a final revised RMP Rule was published in the Federal Register (Ref 14.8 Revised RMP Rule). This represented the review ordered by then-President Obama’s 2013 Executive Order (Ref 14.17 Executive Order) to the federal agen cies responsible for regulating the safety and security of the chemic a l i n d u s t r y i n t h e w a k e o f t h e ammonium nitrate fire and explosion in West, TX in 2013. One of the revisions in the final revised RMP Rule was a provision to perform a Safer Technology & Alternatives Analysis (STAA) as part of the PHAs of RMP-covered processes. The revised final RMP Rule was delayed several times following its publication, but in 201 8 the U.S. Court of Appeals for the District of Columbia Circuit vacated the delay and ordered the final rule be implemented.

125 6.3 Center for Chemical Process Safety (CCPS), Guidelines for Technical Planning for On-Site Emergencies . New York: American Institute of Chemical Engineers, 1995 6.4 Center for Chemical Process Safety (CCPS), Guidelines for Chemical Reactivity Evaluation and Application to Process Design . New York: American Institute of Chemical Engineers, 1995. 6.5 Center for Chemical Process Safety (CCPS), Essential Practices for Managing Chemical Reactivity Hazards. New York: American Institute of American Institute of Chemical Engineers, 2003. 6.6 Forsberg, C.W., Moses, D.L., Le wis, E.B., Gibson, R., Pearson, R., Reich, W.J., et al., Proposed and Existing Passive and Inherent Safety- Related Structures, Systems, and Components (Building Blocks) for Advanced Light Water Reactors. Oak Ridge, TN: Oak Ridge National Laboratory, 1989. 6.7 Hendershot, D.C., Safety cons iderations in the design of batch processing plants. In J. L. Woodward (Ed.). Proceedings of the International Symposium on Preventing Major Chemical Accidents , February 3-5, 1987, Washington, D.C . (pp. 3.2-3.16). New York: American Institute of Chemical Engineers, 1987. 6.8 Kletz, T.A., Plant Design for Safety . Rugby, Warwickshire, England: The Institution of Chemical Engineers, 1991. 6.9 Kletz, T.A., Process Plants: A Handbook for Inherently Safer Design. Philadelphia, PA: Taylor & Francis, 1998. 6.10 Kletz, T.A. and Amyotte, P., Process Plants: A Handbook for Inherently Safer Design, Se cond Edition. CRC Press, 2010. 6.11 Luyben, W.L. an d Hendershot, D.C., Dynamic disadvantages of intensification in inherently safer process design. Ind. Eng. Chem. Res., 43 (2), 2004. 6.12 Norman, D.A., The Psychology of Everyday Things . New York: Basic Books, 1988. 6.13 Raghaven, K.V., Temperature runaway in fixed bed reactors: Online and offline checks for intrinsic safety. Journal of Loss Prevention in the Process Industries, 5 (3), 153-159, 1992. 6.14 Siirola, J.J., An industrial perspective on process synthesis. In AIChE Symposium Series, 91, 222-233, 1995.

11 1.20 National Academy of Sciences (NAS), The Use and Storage of Methyl Isocyanate (MIC) at Ba yer CropScience, National Academies Press, 2012. 1.21 Rogers, R.L., Mansfield, D.P., Malmen, Y., Turney, R.D., and Verwoerd, M. (1995). The INSIDE Projec t: Integrating inherent safety in chemical process development and pl ant design. In G.A. Melhem and H.G. Fisher (Eds.). International Symposium on Runaway Reactions and Pressure Relief Design, August 2-4, 1995, Boston, MA (pp. 668-689). American Institute of Chemical Engineers, 1995. 1.22 Rolt, L.T.C, The Railway Revolution: George and Robert Stevenson (pg.147). New York : St. Martin’s Press, 1960. 1.23 Tickner, J. The case for inherent safety. Chemistry and Industry, 796, 1994. 1.24 U.S. Chemical and Hazard Investigation Board (CSB), West Fertilizer Company Fire and Explosion, Final Report, 2013. 1.25 Vaughen, B. K., and Klein, J. A., What you don’t manage will leak: A tribute to Trevor Kletz. Process Safety and Environmental Protection, 90, 411-418, 2012a. 1.26 Vaughen, B. K., and Kletz, T. A., Continuing our process safety management journey. Process Safety Progress, 31(4), 337-342, 2012b.

Chapter No.: 1 Title Name: <TITLENAME> p05.indd Comp. by: <USER> Date: 25 Feb 2019 Time: 12:33:14 PM Stage: <STAGE> WorkFlow:<WORKFLOW> Page Number: 379 379 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 Part V Additional Information and General Procedure Part 5 has two chapters, Chapters 19 and 20. In Chapter 19 several general procedures are provided. First of all a general methodology is provided for P&ID development of a new item (not familiar for the designer). Then a general procedure for P&ID checking and reviewing is provided.At the end, the quality required for each stage of P&ID is provided. Chapter 20 is devoted to several P&ID examples.

6 | 1 Introduction The shared beliefs and values may create a culture that is either positive or negative, either strong or weak. A strong positive process safety culture would generally exhibit norms such as: Always doing the right thing even when nobody is watching or listening, Not tolerating deviance from approved policies, procedures, or practices, Maintaining a healthy respect for the risks inherent to the processes, even when the likelihood of serious consequences is very low; and Perform ing actions safely, or not performing them at all. Conversely, a negative or weak culture would generally exhibit norm s such as: Tolerating deviance from approved policies, procedures, or practices, Allowing such deviance to become regular occurrences, Exhibiting com placency regarding the operation’s process risks; or Allowing short-cuts to occur to get something done more quickly or m ore cheaply. The CCPS Culture Subcommittee distilled the published definitions listed above, along with their personal ongoing experience in building and strengthening process safety culture. For purposes of this book, a sound or strong positive process safety culture is: From this starting point, Chapter 2 will describe core principles of process safety culture. Chapter 3 will discuss the leadership The pattern of shared written and unwritten attitudes and behavioral norms that positively influence how a facility or company collectively supports the successful execution and improvement of its Process Safety Management System (PSMS), resulting in preventing process safety incidents. • • • • • • • •

Utilities 363 17.2.5 Connection Details of Utility to Process U tilities are used for the purpose of process. Utility streams could be connected and hard‐piped to the process or could be separated from the process. A utility stream that is not hard‐piped to the process ends in the “utility station” (US). Utility stations will be discussed in Chapter 18. If a process needs to be supplied continuously with a utility stream, it should be hard‐piped. However, for a non‐continuous requirement of a utility stream the utility pipe could be hard‐piped to the process or only ending in the US, depending on the frequency of usage. When a utility pipe is connected to a process pipe or equipment, adequate provisions should be considered to make sure no backflow of the process stream occurs and no utility contamination is probable. When the utility stream comes from the US, there is already a check valve installed on the stream and no additional check valve is needed near the process item (Figure 17.6). Connecting the distribution network to the coolecion network will be discussed in section 17.15. 17.3 Different Utilities in Plants There is no standard list of utilities for all plants; how - ever, we can make a list of common utilities in plants. They include: 1) Instrumen t air (IA) 2) Utility air (U A) 3) Utility w ater (UW) 4) Pot able water 5) He at transfer media 6) Condensat e collection network 7) Fue ls 8) Inert g as 9) Va por collection network 10) Emergenc y vapor/gas release collection network 11) Fir e water 12) Sur face drainage collection network 13) Elec tricity Electricity is not a process utility. The generation and distribution of electricity is not shown on P&IDs, there-fore it is not discussed here.In the next sections, we will explain each of these utili- ties briefly. 17.4 Air as a Utility in Process Plants There are at least two types of air used in process plants; they are instrument air and plant air. In some companies these two air systems are completely separate systems. This means there is one instrument air generation system and another one as a plant air generation system, and each of them has their associated distribution system. However, in some other plants they are both integrated into one sys - tem and one system provides both instrument air and plant air. In such cases, however, it should be made sure that preference is given to instrument air rather than plant air. This means if overuse of plant air starts to cause decreased pressure in the instrument air header there should be a control system to cut off the plant air branch and prevent plant air users from using plant air to make sure that instrument air is always available. This is because instrument air is more important than plant air. Instrument air is a motive gas for control valves, switching valves, and some flow meters. While plant air is the air that is used for purposes used in the plant other than those for instru-ment air. Plant air can be used as a motive gas for opera-tion of an air operated pump or it could be used for operation of air cushions in silos. The main purpose of plant air and instrument air is providing a flow of air that is dust free and water droplet free, and within good temperature. 17.4.1 Instrumen t Air (IA) Instrument air is almost always necessary in process plants. IA is used to actuate control valves and switching valves remotely. Therefore IA works as the “nerving sys - tem” of a plant. Basically, instrument air is needed wherever we have a controlled system in a plant. This is because the majority of control valves and switching valves in current industry are pneumatic. There are some non‐pneumatic control valves and switching valves available, but they are cur - rently not popular.Permanent connection Process UtilityTemporary connection Process UtilityFrom utility stationFigure 17.6 Per manent versus temporary utility users.

E.29Disempowered to Per form Safety Responsibilities by Omniscient Software |315 E.29 Disem powered to Perform Safety Responsibilities by “Om niscient1“ Software A plant sustained a small leak on the process side of a heat exchanger. Action was quickly taken to repair it, but during the shutdown, the coolant dropped the exchanger temperature dangerously low, embrittling the m etal. As the process restarted, the heat exchanger ruptured, releasing a flammable vapor cloud. The vapor cloud traveled 170 m eters before finding an ignition source. The m assive gas cloud exploded and then caught fire, killing 2 workers and injuring 8. B ecause the plant was the sole supplier of natural gas to the region, the entire region had no gas for cooking, and factories em ploying 250,000 workers were left idle. A corporate audit of the plant conducted just 6 m onths before the incident declared that the plant’s process safety m anagement system was in order. However, the incident investigation team found (Ref E.3) significant deficiencies in process hazard analyses, training, documentation, workforce involvement and com munication, and management oversight. The Royal Investigation Comm ission noted that the com pany had a world class computer-based system to manage its process safety programs, but concluded that the com pany’s use of it was flawed in that personnel over-relied on checking the boxes specified by the system rather than assuring actual safety, effectively failing to empower individuals to successfully fulfill their safety responsibilities. What other culture gaps m ight have contributed to this incident? What culture factors led the PHA team to fail to understand the hazards and risks they were evaluating and develop insufficient actions? Was failure to ensure open and frank communications and 1 The word om niscient is used here in its literal sense, and does not in any way refer to the software com pany Omniscient Software Pvt. Ltd. Actual Case History

240 PROCESS SAFETY FOR ENGINEERS: AN INTRODUCTION Key Points: Hazard Identification and Risk Analysis. If you don’t identify a hazard, you won’t manage it. The gas plant #1 hazard identification study had been planned, but never carried out. Operators were not aware of the potential hazard of the heat exchanger failing due to brittle fracture and did not know how to respond appropriately. Management of Change. MOC is not just about equipment. Managing the changes in process safety tasks in job descriptions is key. The plant’s process safety engineering staff was relo cated and the role that they filled in Management of Change review was not managed and not replaced. The Supervisors and operators were not prepared for the increased troubleshooting responsibilities. Process Safety Competency. Plant personnel were unaware of the issue of brittle fracture potential when normal steel is reduced to -40°C temperature. Esso argued this persuasively in their evidence at the subsequent enquiry. This process sa fety information should have been understood by plant personnel. Detailed Description The plant involved, Plant No. 1, was a lean oil absorption plant, which separated methane from LPG by stripping the incoming gas with a hydr ocarbon stream called “lean oil”. Methane rises to the top of the towers, with heavier hydroc arbons dissolving in the liquid hydrocarbon condensate, see Figure 12.2. Plant No. 1 had a pair of absorbers operating in parallel. Each absorber had a gas/liquid disengaging region at the base where a mixtur e of gas and liquid hydrocarbons entered the absorbers. During the previous night shift, th e hydrocarbon condensate level had started to increase in the base of Absorber B. As the norm al disposal of condensate to Gas Plant No. 2 was not available, the alternative condensate disposal route was to a Condensate Flash Tank, see Figure 12.3. Under this set of circumstances, it was normal to increase the temperature at the base of the absorber, but this was not done . The inlet to the Condensate Flash Tank was protected against excessively low temperatures by an override on the absorber level controllers. The consequence; therefore, was that the disposal rate of condensate from the absorber became less than the inlet flow, resulting in a buildup of liquid condensate in the absorber base.

192 PROCESS SAFETY FOR ENGINEERS: AN INTRODUCTION the fluid. The next block, Materials of Constructi on, is important to safe processing. Use of the incorrect material of construction can lead to loss of containment. Engineering standards that include design considerations for pumps are: API STD 610 “Centrifugal Pumps for Petr oleum, Petrochemical and Natural Gas Industries, Eleventh Edition” (ISO 13709:2009 Identical Adoption) API STD 617 “Axial and Centrifugal Co mpressors and Expander-compressors” API STD 674 “Positive Displacement Pumps-Controlled Volume for Petroleum, Chemical, and Gas Industry Services” API STD 685 “Sealless Centrifugal Pumps fo r Petroleum, Petrochemical, and Gas Industry Process Service” Figure 11.11. Example application data sheet (OEC Fluid Handling)

333 needed for the defrost cycle of a fr ost-free refrigerator (Ref 13.24 Chemistry and Industry). People must recognize that they are making a tradeoff when they replace CFCs with other materials. While the alternative materials are safer with respect to long-term environmental damage, they are often more hazard ous with respect to flammability and acute toxicity (Ref 13.16 Hendershot 1995). 13.4 INHERENT SAFETY AND HEALTH CONFLICTS 13.4.1 Water Disinfection Substituting bleach for chlorine in drinking water and wastewater treatment facilities can reduce risk at the water treatment plant but may increase the amount of chlorine requ ired at the bleach manufacturing site. The amount of chlorine need ed at the treatment plant—whether from chlorine gas or bleach—will depend on the amount of water to be treated, so the total amount of elemental chlorine required will remain the same. The difference is the way in which the fac ility receives the chlorine and whether a change from elemental chlorine to bleach will reduce the overall risk, or just shift the risk from one place to another. This well-publicized inherently sa fer modification can be used to highlight the site-specifi c challenges to identifying IS opportunities. Converting from elemental chlorine to bleach will reduce the hazard associated with a release of the mate rial to the population around the water treatment plant, whether from the chlorine/bleach storage vessel or the process of connecting and disc onnecting the transported chlorine or bleach to the water treatment process. Due to the economics of sewer and water distribution, such plants are generally located in close proximity to the populations they se rve. Because the same amount of chlorine will be needed to treat a given quantity of water, and there is less chlorine in a container of bleach than in the same size container of chlorine gas, more containers of bl each will be required. The increased probability of release due to the increased number of connections/disconnections necessita ted by an increased number of shipments must be balanced against the reduced potential consequence from the release of a less hazardou s material. The reduction in hazard (in this case, 2nd order IS) could requ ire an increase in the layers of protection surrounding the hazard to reduce the increased probability of release. The goal is for the facility to make sure that the overall risk is

15 2.2 INHERENT SAFETY DEFINED in·her·ent : Adjective. Existing as an essential constituent or characteristic; intrinsic. From the Latin inharens , inhaerent -, present participle of inhaerere, to inhere. W h a t d o w e m e a n w h e n w e s p e a k o f “ i n h e r e n t s a f e t y ” o r “ i n h e r e n t l y safer?” “Inherent” has been defined as “existing in something as a permanent and inseparable element, quality, or attribute” (Ref 2.1 American Heritage). Inherent safety is a concept, an approach to safety that focuses on eliminatin g or reducing the hazards associated with a set of conditions. A chemical manufacturing process is inherently safer if it reduces or eliminates the hazard s associated with materials and operations used in the process and this reduction or elimination is permanent and inseparable. Th e process of identifying and implementing inherent safety in a sp ecific context is called inherently safer design (ISD). A process with reduced hazards is described as inherently safer compared to a proc ess with only passive, active, and procedural controls. Since the 2nd Edition of this book was published in 2009, additional definitions of inherent safety, in herently safer, inherently safer technologies (IST), and inherently safe r design (ISD) have appeared in the technical literature. However, for the purposes of this book, a general definition of inherent safety is adapted from the 2010 CCPS project for the U.S. Department of Homeland Security to define “inherently safer technologies.” (Ref 2.13 CCPS DHS) This definition is as follows: “The application of inherent safety concepts permanently eliminates or reduces hazards to avoid or reduce the consequences of incidents. Inherent safety is a philosophy, applied to the entire life cycle of chemical processes, including design, construction, operation, maintenance, and decommissioning, as well as all mode s of operation of these processes including manufacture, transport, stor age, use, and disposal. It is an iterative process that considers options, including eliminating a hazard, reducing a hazard by having less of the hazardous materials,

34 | 3 Obstacles to Learning litigation may be worse if the company knows of hazards or improvement opportunities but fails to address them. Some attorneys and managers discourage continuous learning for this reason. This approach virtually guarantees eventual litigation, however. Ultimately, the unknown, unresolved problems become incidents that draw the regulators’ attention. It is much better to seek knowledge and address gaps so that there are fewer incidents requiring legal defense. Many regulations and standards will consider processes and equipment that were designed to an earlier version of a standard current at the time to be in compliance even if the standard is later changed. This acceptance of legacied designs can allow a plant to be in full compliance, but not meet the company’s risk criteria. As CCPS discusses in Vision 20/20 (the organization’s guiding vision for process safety by the year 2020), it is important to monitor standards for changes and determine if process or equipment improvements are needed for legacy-compliant designs (CCPS 2014). 3.3 Obstacles Common to Individuals and Companies “It Can’t Happen Here” Attitude—Loss of the Sense of Vulnerability A sense of vulnerability is an essential characteristic of a good process safety culture. We all know the importance of maintaining a healthy respect for process hazards and using that as motivation to faithfully execute our roles with professionalism. Catastrophic incidents are infrequent, however, and that can drive us to relax our sense of vulnerability, leading to complacency and a false sense of security—which in turn can compromise performance and demotivate efforts to improve. Ivory Tower Syndrome Many companies have teams of highly competent process safety professionals focused on advanced learning. Often, however, these individuals are effectively walled off from both operations and corporate oversight roles. This can create a significant gap between what the corporate experts have learned and what the company practices. In some cases, the walls are real organizational obstacles, while in other cases they are built by the personalities of the individuals involved. In either case, the company cannot benefit from what its experts have learned. To obtain the maximum learning benefit from these experts, companies should

APPLICATION OF PROCESS SAFETY TO WELLS 65 Casing: API RP 100-1 (API, 2015) for onshore hydraulic fracturing notes that casing design and selection is critical to well integrity including well control. It must be designed to withstand all anticipated loads while running into the hole, as well as loads during drilling, completions, workovers, interventions and production. The prime design factors on casing are ratings for tension, burst and collapse pressure. The selection of casing material is important to avoid corrosion and loss of containment events. IADC (2015a) and API 5CT (2019b) provide guidance on material selection to deal with sour gas, CO 2, chlorides, temperature, carbonate concentration, and produced water contaminants. Cement: Cement is a critical barrier element in achieving isolation and multiple local factors can affect cement integrity. The fa ilure of the cement job to achieve the required isolation in temporary abandonment of the Deepwater Horizon rig was a significant contributing factor to the blowout. Dusseault et al (2000) discuss mechanis ms causing onshore oil wells to leak, especially those related to cement failure s. They identify cement shrinkage as an important factor, and this leads to channeling and high cement permeability. Inadequate design or installation and contamination are all important factors in cement failures. The BOP: While multiple responses are possible to a kick event, one common response is to circulate mud using the Driller’s Method prior to use of the BOP. This requires two complete separate circulatio ns of drilling fluid in the well. The first circulation removes influx with original mud weight, while the second uses kill weight mud. If this is not successful, the well is sealed using the BOP. A simpler method, the Engineer’s Method (aka ‘Wait and Weight’), requires only one circulation of a heavier mud weight material. As previously noted, a BOP normally requires manual actuation, except in the case of a drive off or drift off event offshore or a loss of control signal. To be effective in stopping a blowout, the BOP must be actuated in a timely manner. A blowout preventer will not stop the blowout if it is not actuated in a timely manner. This was apparent in the Deepwater Horizon event, where the drill string was pushed off-center inside the BOP during the event and could only be squeezed but not cut by the blind shear ram (DNV, 2011). If the BOP fails to seal and well fluids ri se to the surface, then a diverter valve can be actuated directing well fluid flow overboard (offshore) or to a flare or burn pit (onshore) away from the locations wh ere crews are working. This reduces the risk of harm to people but does not eliminate the hazardous situation entirely. Key Process Safety Measure(s) Process Safety Competency : Well construction is dynamic and manually controlled. The competence of those involved to be ab le to conduct operations and detect and respond promptly when an unplanned influx occurs is key to safe well construction operations.

376 Net operating costs Change in the cost of material s including transportation and handling related costs Change in energy consumption Change in human costs such as number of operators, training Any other direct manufacturing costs Net regulatory compliance cost, change in fees Demolition and future cleanup and disposal cost All of the above criteria requir e a quantitative justification (cost/benefit analysis) Generally, an IS measure is feasible if it has been successfully applied to similar processes or similar situations unless there are unique circumstances at the facility. The ju stifications should highlight those unique circumstances and how they relate to the feasibility factors. (Ref 14.13 NJ IST) It is important to point out that, while the Prescriptive Order and its IST review requirement were driv en by security concerns and the potential for intentional releases, efforts to comply with the Prescriptive Order also served to address the po tential for accidental releases. In other words, the required review would address the full range of IST strategies, and not only substitution, minimization, and moderation—the most effective strategies to reduce security-related risk. Results of Implementing IST Under the Prescriptive Order . According to NJDEP, of the 157 facilities subject to the Prescriptive Order, more than 98% complied within the 120-day deadlin e established in the Order. Of those, 32% provided a schedule to implement additional IST or other risk reduction measures, and 19% identified additional IST or risk reduction measures. The remaining 49% of the facilities had no additional recommendations and 80% of the facilit ies concluded that at least some of the IST or risk reduction measures identified during the IST evaluation were infeasible for their operat ions. (Ref 14.12 Sondermeyer ) Based on the results of the IST review program required under the Prescriptive Order, New Jersey believe s that evaluating inherently safer technology is not overly burdensome on industry and is an effective tool for critically evaluating the risk reduction opportunities available at a

102 PROCESS SAFETY IN UPSTREAM OIL & GAS with personnel to identify these locations. The company must implement a system to ensure that, over time, unoccupied sp aces do not become occupied as that invalidates the safe spacing decisions. Vapor Cloud Explosion – Short Primer Leak of hydrocarbon vapors or mists of flammable liquids disperse downwind. If this vapor cloud is ignited in unconges ted space, then a flash fire event results. This is primarily a hazard to anyone trapped within the cloud. People close by wearing normal PPE, but not in the cloud, may not be seriously impacted. A much worse outcome occurs if the clou d disperses into congested space and a vapor cloud explosion results. Process Safety Issues : The Flixborough chemical pl ant incident in 1976 was the first well documented VCE event and it, plus some serious offshore explosions such as Piper Alpha, led to much experimental work to understand the mechanism involved. Hundreds of larg e-scale experiments were carried out, mainly at Spadeadam in the UK and in Texas in the US. These showed that congestion from process equipment cause s the flame front to accelerate from low speeds (under 10 m/s [33 ft/s] as in flash fires) to higher speeds (300 m/s [984 ft/s]) that result in damaging ove rpressures. This event is termed a deflagration. Deflagrations only generate overpressures for that part of a flammable cloud within the congestion, so the whole mass released may not contribute to the blast propagation. It had been argued that some major explosion events were detonations. This was not initially accepted until detailed investiga tion of the Buncefield storage tank event in 2005 showed that it was a DDT event – deflagration to detonation transition. That event was caused by a tank overflow, spilling gasoline over a wing girder of the tank, and forming a mist of gasoline vapor and droplets at ground level. This was ignited and the flame front was accelerated, not by equipment congestion, but by dense foliage in a hedge row. A detonation event is more serious than a deflagration as the flame speeds are much higher (1000+ m/s [3281 ft/s]) and that all the material in the flammable range contributes to the explosion, not just the portion in congested space (Hansen and Johnson, 2015). RBPS Application Hazard Identification and Risk Analysis : Any team member in a HAZOP or Facility Siting Study can identify the potential for a vapor cloud explosion. Analyzing the severity of the explosion and estimating the risk of the explosion should be conducted by a tech nical expert in this field.

KEY RELEVANCE TO PROCESS PLANT OPERATIONS 31 for a limited time, it does not always work as intended. Guidelines and/or training for continued operation and/or safe shutdown in this circumstance should be developed. Further aspects are discussed under 3.4.2.3. The example described in Example Incident 3.1 involves a power failure just to the process cont rol system and the unforeseen consequences of restoring the supply. Example Incident 3.1 – Control System Power Failure A batch process using toxic and fla mmable chlorocarbons suffered a failure of power supply to the DCS. The process continued to operate safely without the DCS for a short time, since the system was set up so the controls would fail to a safe position. The reactor agitator control systems went to a stay-put mode, so the reactants continued to be mixed and the exothermic re actions were in control. When power to the DCS was restored, the op erators then turned it back on, which forced all control parame ters and variables to their initialization positions. This cau sed a number of problems, including a zero-speed for the agitators. As a result, the plant experienced a near-miss, as a chemical stop agent had to be used to kill the reaction in order to prevent a runaway reaction. Lessons learned in relation to abnormal situation management: Abnormal Situation Recognition: Operating teams must be aware of the failure modes of equipment due to loss of services (e.g., power, steam, instrument air). Failure modes due to loss of services (power, steam, instru ment air, or other utilities) should be identified during the risk assessment and be understood by designers and operating/maintenance teams. Procedures: Should be available, and training should be provided to handle such failures. Training should include failure of one service that can lead to ca scading failure of other services (e.g., power, stea m, air, water).

236 beyond doing the minimum required to give the appearance of compliance. Figure 10.1 illustrates the required foundation of Figure 10.1: Management leadership is the foundation for process safety management management leadership (Ref 10.3 Au ger). Managing process safety is based on a corporate mission which management implements. By their words and actions, managers at all levels of the corporation must understand the benefits of IS an d show their commitment to such programs in tangible ways. The la tter should include providing the necessary resources, such as personnel, time, and funding, for IS-related activities like reviews and evaluati on/ implementation of follow-up recommendations that arise. 10.4.2 Incorporating IS into Normal Design Process As stated previously, IS should be made a part of the normal design process, by incorporating IS th inking into design processes and standards. Only by implementing IS consistently throughout the

9 • Other Transition Time Considerations 163 Table 9.1 Definitions of the Proce ss or Equipment Life Cycle Stages. Definitions of the Process or Equipment Life Cycle Stages 1When the engineering design concepts, th e process design parameters, and the equipment design specifications are established and the process knowledge and the process design basis are documented. 2When the equipment is fabricated per the fabrication design specifications (Note: fabrication may occur at the equipment's location in the process unit) 3When the assembled equipment is insta lled at its designated location in the process unit per the insta llation specifications. When the installed equipment is approved for safe operations. Commissioning steps include: 1) verifying that the equipment and process unit meet their performance specifications: 1.1) testing the equipment, the contro l systems, the protecti on layers, and the utilities, 1.2) training all operations and mai ntenance personnel on their tasks and procedures, and 2) safely introducing the chemical s to the equipment and process units. The transient operating time when the pr ocess chemicals are introduced to the equipment and process units for the first time after the new, unused or modified equipment has been fabricated and installed. 5 When the equipment and the pr ocess units are safely operated 6When the equipment is safely maintained using the established, scheduled Inspection, Testing, and Preventi ve Maintenance (ITPM) program. 7When proposed changes to equipment design, process design, engineering controls, or administrative (procedural) controls ar e reviewed, approved, and prepared for commissioning. When the equipment's or pr ocess unit's useful life is over and the decision has been made to remove the equipment or process unit from normal operations (its end-of-life stage). The transient operating time when the process chemicals are removed from the equipment and process units and the equi pment are prepared for their end-of- life project stages. A decommissioning stage: When the equipment or process unit may be potentially re-commissioned at a later date. A decommissioning stage: When the equipment or process unit is dismantled and individual components of the equipment or individual equipment from the process unit may be re-used. A decommissioning stage: When the equipment or pr ocess unit is essentially dismantled and for scrap and or material recycling. Fabricate Initial start-upCommissionLife Cycle Stage 4 8Design Operate Maintain Change Decommission Shut-down for decommissioning Mothball Dismantle Demolish Construction Install

KEY RELEVANCE TO PROCESS PLANT OPERATIONS 37 Example Incident 3.3 – BP Texas City 2005 The 2005 BP Texas City refinery ex plosion on the Isomerization Unit occurred during startup, when th e Raffinate Splitter fractionation column overfilled (illustrated in Figure 3.1). The control board operator had failed to open the co lumn bottoms line to tankage as required by the operating procedur e. The bottoms temperature of the Raffinate Splitter exceeded the maximum temperature specified in the operating procedure. Although the column top pressure was normal, the hydrostatic head of liquid in the overhead line was sufficient to open the three relie f valves at a lower level on the overhead line. Consequently, the liquid feed flowed directly to an atmospheric Blowdown Drum, allowing the liquid feed and vapors to release from the elevated vent stac k on the Blowdown Drum and pool around the drum before finding an ignition source (CSB 2007). Figure 3.1 BP Texas City Raffinate Splitter

APPLICATION OF PROCESS SAFETY TO WELLS 59 4.1.4 Drilling the Well: Casing Casing is normally composed of sections of steel pipe screwed together. Usually multiple diameters of casing are employed in decreasing size with well depth to allow new casing sections to pass through already set sections. Typical casing naming conventions in order of depth are: 1) conductor, 2) surf ace, 3) intermediate, 4) production, and 5) reservoir liner. Not all wells have all these types of casing; some wells have drilling liners at intermediate depths in addition to the reservoir liner and some wells have the production casing thru the reservoir and no liner. The conductor casing protects the well fr om collapse from loose near-surface aggregates and serves as the foundation fo r the well. For onshore wells the conductor is often 15-30 m (50-100 ft) deep, for offshore wells it may be 300 m (1000 ft) deep. Surface casing comes next, and it protect s local groundwater resources from potential contamination from well fluids and typically extends at least 15 m (50 ft) below any potable groundwater unless lo cal regulations require more. Surface casing is usually cemented all the way back to the surface or seabed (i.e., a layer of cement outside the casing separating it from the formation), completely isolating any groundwater resource. Testing of cemen t integrity is required once completed. In some cases, the total well depth is safely drilled from the surface casing alone, but usually another deeper intermed iate casing is required. Intermediate casing protects the wellbore from multiple problems and ensures that the pore pressure fracture gradient limitations are not violated. The production casing is usually run from the wellhead to the full design depth of the well. An additional process safety hazard is du e to the mechanical handling of casing or drill pipe segments near to producing facilities. Dropped segments can rupture pipework or vessels and create a LOPC event. This is a SIMOPS issue and special controls are needed, which are discussed further in Section 5.2.5. 4.1.5 Drilling the Well: Cement Cementing and cement compositions are discussed in Chapter 9 of the SPE Petroleum Engineering Handbook (2007). Cement is used to permanently seal annular spaces between the casing and the borehole walls. Cement is also used to seal formations to prevent loss of drilling fluid and for operations ranging from setting kick-off plugs to plug and abandonment. Various additives are used to control density, setting time, strength, and flow properties. The cement slurry, commonly formed by mixing cement, water and assorted dry and liquid additives, is pumped into place and allowed to solidify (typically for 12 to 24 hours) before additional drilling activity resumes. 4.1.6 Drilling the Well: The BOP The blowout preventer, BOP, is a safety device that forms part of the well barrier system (see Figure 4-2). The terms BOP, blowout preventer, blowout preventer stack and blowout preventer system ar e used interchangeably. Note the BOP normally requires manual actuation, except in the case of a drif t off event offshore,

10. Implementing Inherently Safer Design 10.1 INTRODUCTION To be most effective, implementation of inherently safer designs/technologies requires a systematic management approach, sound technical basis, and manage ment and cultural emphasis on inherent safety as an important organizational value and tool for reducing the risk of process-related inci dents. At a corporate level, it first requires management commitment and leadership to provide resources and establish policies and procedures to integrate the use of inherent safety into the framework of the company’s overall process safety management program. This may be also driven by regulatory requirements to implement ISD in ad dition to a company’s commitment to reducing risk. As a general philoso phy, its concepts should be woven into the way that fac ilities are designed, constructed, operated and maintained, so that company practiti oners can continuously look for and identify way s to maximize the inherent safety of an operation . This is very similar in concept to the cu rrent industry emphasis on “Lean Manufacturing,” which involves an overall management approach to eliminate waste in over-productio n, waiting time, transportation, processing, inventory, motion and sc rap, and on “Kaizen,” a philosophy of continual improvement in busi ness processes to accomplish this objective. It is now more common for compa nies to build inherently safer design principles into their proces s safety management systems. For example, this can be accomplished by incorporating inherently safer design concepts into existing safety and process hazards reviews. Companies may wish to enhance their existing review systems with inherent safety reviews at key poin ts in the process life cycle. This chapter discusses methodologies to co nduct inherent safety reviews at three key stages of the life cycle: 1.During product and process development; 230 (VJEFMJOFT
GPS
*OIFSFOUMZ
4BGFS
$IFNJDBM
1SPDFTTFT
"
-JGF
$ZDMF
"QQSPBDI #Z
$$14 ¥

5IF
"NFSJDBO
*OTUJUVUF
PG
$IFNJDBM
&OHJOFFST

CASE STUDIES/LESSONS LEARNED 181 7.1.7.4 Procedures: The report refers to flight procedur es that should be followed in the event of unreliable airspeed indication . This provides a table of flight settings to ensure that the aircraft operates within its safe flight envelope. This procedure does not appear to have been followed by the pilots, although following analysis with reports and statements from other crews, it states: ... although technically adequate, details of the procedure continue to be understood to differing degrees by crews, who do not always consider their application necessary, and even sometimes consider them to be inappropriate at high altitude… Some crews mentioned the difficulty of choosing a procedure bearing in mind the situation (numerous warnings) . The development of procedures shou ld always involve the people who are using them, to make them ac curate, meaningful, and usable. 7.1.7.5 Communications: During the handover before the Capt ain left the cockpit, he did not specify which of the two co-pilots woul d be his designated relief, nor did he provide any instructions for crossing the ITCZ. In particular, he did not comment on the meteorological si tuation which was about to be encountered during the ITCZ cros sing. He also did not provide instructions concerning the tactics for crossing the ITCZ, or on the PF’s decision to try to climb above the cl oud mass. This may have given rise to possible issues regarding hierar chy in the cockpit between the two pilots and perhaps among the three of them after the Captain returned. The report referred to a deteri oration in the quality of the communications between the PF and th e PNF, associated with the stress of the situation. Training in simulators using suitab le, realistic scenarios can provide useful experience for pilots in stressful situations. 7.1.7.6 Training/Knowledge & Skills: When the autopilot disconnected and the control laws were reduced, the aircraft was stable, but the sudden intr oduction of control inputs rapidly brought it outside the flight envelope. Pilots are not used to flying under

232 IS reviews of new and existing facilities, including recommendation follow-up tracking (may be incorporated into project design reviews and facility PHAs) Implementation of IS concepts in on-going aspects of an operation (i.e., management of change, maintenance, SOPs) A systematic management review process, including performance metrics, performanc e assessment, and review and implementation of improvements Amyotte, et al. (Ref 10.1 Amyotte) describe how IS concepts can be integrated into each element of a process safety management program, including maintenance and operating procedures, training, management of change, and incident investigation. Dowell (Ref 10.11 Dowell) emphasizes the need to inte grate the many environment, safety and health elements, including those involving process safety management, into a comprehensive ma nagement system so that they become part of a company’s way of doing business, and not separate management programs. IS is unlikely to succeed as a separate program; it needs to be integrated into an overall Process Safety Management (PSM) program. The process hazard analyses (PHAs) will drive understanding of the process hazard s/risks, after which IS techniques could be used to control or eliminate the identified hazards/risks. In this way, inherent safety becomes an inherent p a r t o f t h e company’s system for improving pr ocess safety and reducing risk. Application of IS principles to Ris k Based Process Safety elements is addressed more comprehensively in Chapter 11. Like other management systems, an IS management system moves ISD from the conceptual stage to im plementation in relevant process safety-related activities. Without a management systems approach, IS will remain as a concept only, and not a functional element of the organization’s safety efforts. 10.3 EDUCATION AND AWARENESS 10.3.1 Making IS a Corporate Philosophy Once management has established it s commitment to the principles of inherent safety, including the deve lopment of a management system

134 INVESTIGATING PROCESS SAFETY INCIDENTS The interviewer should express appr eciation for the witness’s time, information, and cooperat ion and gain consent to contact the witness later if necessary for a follow-up interview, even if this is co nsidered unnecessary. If the interviewer asks permission for fo llow-up interviews with only some of the witnesses, those witnesses may f eel they are being singled out. Finally, the investigator should review the notes with the witness. During this review, numerous clarifications and additional details are usually provided. I t i s c o m m o n f o r a w i t n e s s t o r e c a ll additional information after the interview is over. Astute investigators anticipate this human trait and provide a clearly understood and easily accomp lished mechanism for the witness to contact the interviewer later. Always close an interview by inviting the witness to return or contact the investigator if he remembers something else, or would like to otherwise modify or add to the interview results. Provide the investigator’s contact information to the witness. 7.4 CONDUCTING FOLLOW -UP ACTIVITIES Once the interview is complete, the investigator should perform a few additional tasks immediately after the witness leaves the room: • Review the interview process against the plan • Organize the information received • Identify any key points that confirm or conflict with previous information • Record the findings. Findings would include such items as observations, specific insights, and a list of items to be followed-up on in later interviews or investigation activity. Where relevant, the investigator should add content to a timeline, based on the witness statement (See Chapter 9. 2.1 for more details on timeline development.) Finally, the information from the interview should be communicated to the remainder of the investigation team.

154 INVESTIGATING PROCESS SAFETY INCIDENTS Table 8.4 Examples of Position Data • As found position of every valve related to the occurrence • As found position of controls and switches • Condition of relief devi ces (e.g., open/ closed) • Tank levels • Pointer needle positions from locally mounted temperature, pressure, and flow devices. • Location of flame and scorch marks • Position and sequence of layers of materials and debris • Direction of glass pieces • Missile mapping • Locations of parts removed from the process as part of maintenance • Locations of personnel involv ed in the maintenance and operation of the process • Locations of witnesses/ witness views • Location of equipment that should be present but is missing • Smoke traces • Location or position of chemicals in the process • Melting patterns • Impact marks • Assembly of equipment • Locations of training aids and procedures/checklists Position data is one of the most fragile types of data. It can be lost through many activities including: • Emergency response activities • Fire extinguishment • Removal of the injured • Stabilization of the system, including repositioning of valves/switches/controls, draining of tanks • Witness movement • Restoration/stabilization/demolition work • Degradation from weather • Investigator actions Typically, position data are recorded by documenting visual observations via photography/ video, drawings, ma ps, and measurements. An example photo that documents an as-f ound valve position is provided in Figure 8.3.

Pipes 81 But is this spec break only a nonphysical border or does it have some representation on a pipe like a flange? In a majority of border cases, only a flange is enough on a border, but in a spec break, the border could be more complicated. There are at least three different types of borders for spec break: flange (Figure  6.29a), blocking valve (Figure  6.29b), and blocking valve‐check valve (Figure 6.29c). A process engineer decides about the type of spec bor - der based on judgment or consultation with the project documents. As a rule of thumb, a process engineer decides by default to use a flange for spec break. But if the spec is changed to a robust spec, the designer may choose to use a blocking valve or even blocking valve‐check valve combination for spec break border. For Figures  6.29b and c, the question that arises is which pipe spec should cover the border system (valves)? Or where should the pipe spec border be, on the right side or left side of the valves (Figure 6.30)?The common practice is to always cover the valves with the more robust pipe spec. For example, in Figure 6.30, the designer should investigate if spec A or spec B is more robust. The more robust one should be included in the P&ID. But the question is how we can recognize which is more robust or less robust. This is not always easy, so it is recommended to consult with a piping material engi-neer on the project. Figure 6.31 shows a pipe spec border at the middle of a pipe, which is wrong because it should be at least on a 2/uni2033-CH-5327-BBABBA CBA2/uni2033-CH-5007-CBAFigure 6.28 Pipe spec bor der and its effect on the pipe tag. AB(a) (b) (c)AB AB Figure 6.29 (a–c) P ipe spec break border systems.Spec ASpec B Spec ASpec B(b)(a) Or Figure 6.30 (a, b) Options f or placing a pipe spec border on the border system. Spec ASpec B Spec ASpec B(b)(a) Figure 6.31 (a, b) M istakes in placing a pipe spec border.

152 American Society for Testing and Materials (ASTM) International, CHETAH: Chemical Thermodynamic & Energy Release Evaluation (Ref 8.6 ASTM CHETAH). US Seal (www.usseal.com/jmchem.html). iProcessamart.com (www.iprocessmart.com/techsmart/compatibility.htm) IDEX Health and Science (www .idex-hs.com/education-and- tools/educational-materials/chemical-compatibility) Reactivity Testing. There are a number of testing methods available to determine the thermal stability and the onset temperature of exothermic reactions, as well as the rate of reaction and heat generated per unit mass of the material(s) in volved. These are summarized below, and described in full by (Ref 8.20 CCPS 2003a), Englund (Ref 8.35 Englund 1990), and Fauske (Ref 8.39 Fauske): Differential scanning calorimetry Differential thermal analysis Insulated exotherm test Decomposition pressure test Carius sealed tube test Mixing cell calorimetry Vent sizing package Accelerating Rate Calorimeter® Reactive System Screening Tool/Advanced Reactive System Screening Tool Table 8.4: Reactive Combinations of Chemicals Substances Type of Hazard A + B = H a z a r d o u s E v e n t Acids Chlorates Spontaneous ignition Chlorites and Hypochlorite Spontaneous ignition

4.7 References |155 4.14 Chemical Safety and Hazard Investigation Board, Investigation Report – West Fertilizer Company Fire and Explosion , 2013 4.15 Chemical Safety and Hazard Investigation B oard, Safety Bulletin – Dangers of Propylene Cylinders in High Temperatures 4.16 International Atomic Energy Agency, NS Tutorial, Section 6., Developing Safety (6.2.1 How to Measure Safety Cultur e) , 2001. 4.17 Center for Chemical Process Safety (CCPS), Guidelines for Risk B ased Process Safety, American Institute of Chemical Engineers , 2007.

122 Guidelines for Revalidating a Process Hazard Analysis Table 6-4 continued Process/System Characteristic Examples of Change Siting Addition, relocation, or remova l of temporary structures or trailers Addition or demolition of permanent structures Changes in occupancy Changes to emergency vehicle access routes Increased vehicle traffic Equipment relocation Electrical classification changes New adjacent units or structures Increases in, or addition of new, vulnerable exposures surrounding the process unit (either on-site or off-site) Process changes that introduced new scenarios with potential siting impacts Identifying Documented and Controlled Changes. Appendix C contains an Example Change Summary Worksheet that could be used as an aid for this task. This worksheet is organized by P&ID number (first column) so that the revalidation team can evaluate each chan ge in the context of the P&ID-based review during the revalidation work sessions. The worksheet provides for documentation of the nature of the chan ge, the source of information regarding the change (i.e., in what documentation was the change discovered), and explanatory comments. The last column of the worksheet in Appendix C allows documentation of the proposed action for addressing the change during the revalidation. Table 6-5 is another example of how to document changes reviewed in a revalidation. Note that in this table, the MOC Reference and Change Description columns are completed during preparation, and the PHA Revalidation Team Comment should be filled out during the PHA sessions with the team. The MOC Detail column provides the minimum detail required, and the addition of the PHA Revalidation Team Comment column makes the documentation a stronger record of how the Update approach was applied. This information would be useful in the PHA revalidation report document.

Figure 6-1: Selecting a type of job aid for operational use Key: CK = Checklist. GC = Grab card. DFC = Diagnostic flow chart DT = Decision tree Info = Information (e.g., chemical safety datasheet) Log = Operational log M = Manual PTW = Permit to work SH = Shift Handover SOP = Standard Operating Procedure WI = Work Instruction

128 PROCESS SAFETY FOR ENGINEERS: AN INTRODUCTION Material Handling Guides. Organizations such as the American Chemistry Council and chemical manufactures publish guides on safe handling of chemicals. These are typically short documents that address the handling, storage, transportation, and compatibility with other chemicals and materials of construction. These are available through an internet search on the chemical name or manufacturer. What a New Engineer Might Do A new engineer may work closely with a chemis t when dealing with chemicals to understand the hazards associated with the pure substances , and mixtures. They may also be involved in handling and processing chemicals or designing systems and procedures for others who handle and process chemicals. In either case, an engineer has a responsibility to understand and manage the hazards associated with chemicals. This includes researching chemical data and communication of the chemical hazards usin g the sources and systems identified in this chapter. Through this, the engineer will protect th emselves, as well as others working with the chemicals. Tools The chemical hazards data sources and communication systems discussed in this chapter are themselves the tools that support the identification and understanding of chemical hazards. Summary It is imperative that engineers understand the hazards associated with the chemicals they are including in process designs. It is also impe rative that they communicate these hazards to those who are handling these chemicals in the wo rkplace. Many data sources are available to support the identification and communication of chemical hazards. Many of these are now aligning their categorizations and communications with the UN Globally Harmonized System of Classification and Labelling of Chemicals (GHS). (UN) Exercises List 3 RBPS elements evident in the Conc ept Sciences explosion summarized at the beginning of this chapter. Describe their shortcomings as related to this accident. Considering the Concept Sciences explosion, what actions could have been taken to reduce the risk of this incident? What pictogram is used in the Globally Harmonized System (GHS) for potassium permanganate? For anhydrous zinc chloride ? What pictograms are used for acute toxicity? Is anhydrous ammonia a fire hazard or a toxicity hazard? Draw the NFPA 704 diamond for it. For a small spill of boron trifluoride at ni ght, how far downwind should people be protected? The MIC release in Bhopal, India was summarized in Chapter 6. For a large release of MIC at night, to what distance downwind should people be protected?

112 PROCESS SAFETY IN UPSTREAM OIL & GAS ●Helicopter landing and take-off operations and offshore vessel operations can potentially impact process equipment or risers through collisions. ●Process facilities are often contained within enclosures to protect equipment and workers from the weathe r and this increases the risk of small leaks accumulating to flammable c oncentrations that might otherwise disperse safely in an open design. It is worth noting that some onshore facilities, especially in harsher weathe r locations, are also contained within enclosures and share this risk. ●SIMOPS is a key challenge for offshore. Refer to Chapter 5 where this topic was discussed. The following sections highlight some of the more common hazards associated with areas of an offshore facility. A Hazard Identification and Risk Analysis (HIRA) is recommended to determine the possible hazards and incident scenarios applicable to a specific offshore facility. Industry codes such as API RP 14C, 14J and ISO 17776 are helpful for hazard identification. The UK HSE provides a wide range of useful notes and regulatory guidance on offshore hazard management on its webpage (www.hse.gov.uk/). 6.2.1 The Well Risks Loss of well control can occur during the production phase in addition to during well construction. This may be due to interventio ns or workovers as an extension of the well operations or due to production probl ems or collisions. These were discussed in Section 4.2.2 summarizing the SINTEF blowout database. Key Process Safety Measure(s) Asset Integrity and Reliability : There are several asset integrity activities undertaken to ensure that the wellbore maintains its integrity during production operations. Emergency Management : The BOP is replaced during production by a Christmas tree and this is closed to prevent a potential loss of co ntainment. Subsurface safety valves (SSSV) are often installed into th e well as an additional barrier. A loss of well control does not automatically mean a loss of containment to the environment. If one barrier is lost, the well should be shut in until actions are implemented to stabilize the well and restore the lost barrier. 6.2.2 The Production and Export Risers Risks The production riser takes production from wells on the seabed or from nearby facilities up to the topside production facility. It also takes cabling and other services down to the wellhead or other facilities. Export risers send oil or gas down to export

9 • Other Transition Time Considerations 174 commissioning and initial start-up stages. Although there may be fewer members and groups associat ed with the commissioning team, robust and clear handovers should be established for safe start-ups. At this point, it is worth recognizing that no start-up-related pre-plans can anticipate all situations that actually occur during the start-up. Some general guidance for effectively managing these unexpected situations is provided in the Appendix. 9.3.3 When managing new equipment or process units Larger capital projects will range from new, major processing equipment, such as a distillation tower , to a process unit or facility. As was noted in Chapter 5, these projec ts need to have the rigorous and disciplined project management approach for their commissioning and initial start-up stages. Due to the larger number of team members and groups associated with the co mmissioning and initial start-up efforts, robust and clear handovers should be established for safe start-ups. In such cases, a sp ecial commissioning team with a commissioning project manager should be used to help manage the risks associated with the many insp ections, tests, and equipment– and process unit–related verification steps (refer to Section 9.3.1). 9.4 Incidents and lessons learned, commissioning and initial start-ups Details of some commissioning-relate d incidents are included in this section. The incident summary is provided in the Appendix.

Provisions for Ease of Maintenance 135 The other parameter is the pressure of the fluid. If the pressure is higher we may need a more positive isolation system. For example isolation of high pressure steam should be stronger than low pressure steam. The other location that we generally put an isolation system in is when a utility stream goes from one area to another. The type of isolation system depends on the fluid type and pressure, and on the border of the areas, unless two areas are so interrelated that one cannot be run when other one is shut down. This concept is shown in Figure 8.6 for a high pressure steam stream. Isolation may also be needed as part of the safe shut - down of a unit. In such cases the isolation is done fully automatically and probably without usage of blinds (Figure 8.7). 8.7.3 Plac ement of an Isolation System The answer to the third question is that isolation system should be added to all downstream and upstream con-necting pipes, and as close as possible to the equipment. Some companies, however, challenge this, and ques - tion whether it is a really necessary to put an isolation valve on a pipe that goes to atmosphere (Figure 8.8). 8.7.4 Inbound Versus Outbound Blind Location The question arises when talking about the blind is its location with respect to the to‐be‐isolated system. The blind (either of spectacle type or spade type) should be placed on the isolating valve but the question is whether it should be placed closer to the to‐be‐isolated equip-ment, inbound, or away from the to‐be‐insulated system, outbound (Figure 8.9). The difference is that in outbound blinding the to‐be‐ isolated system doesn’t need to be emptied before blind-ing (Table 8.4). Some companied prefer outbound blinding because there is more flexibility in operation, but some other companies prefer inbound blinding because it allows “the only correct isolating sequence. ” 8.7.5 Mer ging Isolation Valves There is one opportunity for saving isolation valves. This is a good cost saving strategy if pipes are of large size, e.g. larger than 12″. This concept was covered in Chapter 7. Y 218I/PY 219 To-be-isolated syste mI/P YBlock valves Bleed valve220I/PFigure 8.7 Aut omatic double block and bleed system. To Atm. Figure 8.8 Loca tion of isolation systems. To-be-isolated syste m Inbound bl ind Outbound blindTo-be-isolated syste m Figure 8.9 Inbound v ersus outbound blind location.

120 •Using immersion heaters that cannot add enough energy to cause a fire in the material being heated or damage the container. •Using a heating medium for distilla tion reboiler at a temperature such that it cannot overpressure the tower in case of loss of cooling flow to the condensers. •Limiting process heating using steam, when possible, to the saturation temperature, which adds the needed amount of heat and no more. If the heating me dium maximum heat flux cannot be reduced, the heat transfer ar ea should be adjusted to limit the energy transfer. •Limiting pump or compressor disc harge pressures to less than the downstream relief valve setpoints or the maximum allowable working pressure of any downstream components. •Ensuring that residual heat ca nnot be transferred inadvertently to a material via conduction or radiation, such as a hot vessel wall that transfers heat to a materi al that is sufficient to cause a runaway reaction (Ref 6.9 Kletz 1998). Some of the examples above also relate to designing equipment that is robust enough to withstand the maximum achievable temperature or pressure. 6.11 SIMPLIFICATION OF THE HUMAN-MACHINE INTERFACE 6.11.1 Overview In the previous sections, the focus has been on simplifying designs to eliminate or reduce the chemical/stor ed energy hazard. This section will address the simplification of the hum an-machine interface; i.e., how humans interact with the process, in order to reduce the likelihood of errors. The human-machine interface in cludes all aspects of the process (equipment layout, accessibility, operability, maintainability, functionality of controls, etc.), not simply the computer screen or control panel from which the process is op erated. This can be considered a subset of Human Factors. For a more comprehensive discussion of Human Factors, see the CCPS book, Human Factors Methods for Improving Performance In the Process Industries , 2006.

D.1 The HRO Concept |281 regulations shrink operating margins, the conflicts could be intensified. Additionally, the technical regulations governing nuclear power go much further in establishing design, construction, operations and maintenance standards. This is possible due to the relatively limited scope of technologies that are practiced. Again, this can be helpful in that lessons learned can be systematically incorporated into standards and com municated, but also provides less flexibility than chem ical, oil, and gas facilities need. Finally, with the unforgiving social and political environment of nuclear power, regulatory agencies place resident inspectors onsite at every nuclear power plant. This is an additional source of ever-present independent oversight with direct authority to order immediate shutdown if deemed necessary. In chem icals, oil, and gas, regulators can usually order shutdown in cases of situations deemed imm ediately dangerous. However, inspectors are in facilities only occasionally, and in such cases a court order m ay be needed. HROs perform m uch more intense indoctrination of personnel than chem icals, oil, and gas. Indoctrination begins on the first day of em ployment, where new hires are constantly and forcefully reminded that the stakes are higher than other work places. Training and qualification program s are m uch more structured in HROs than in other industrial sectors. For exam ple, control room operators in nuclear power facilities must be granted form al reactor operator licenses based on a training and qualification process specified by regulation. In the chem ical, oil, and gas sector, com panies m ay have an internal qualification program for operators, but there are very few exam ples of form al training required of operators. An exception to this is a certification required by the State of California, USA, for wastewater treatment plant operators.

276 | Appendix C As Low as Reasonably Practicable C.4 International Electrotechnical Commission, Functional Safety: Safety Instrumented Systems for the Process Industry Sector , IEC 61511, 2003. C.5 American National Standards Institute, Functional Safety: Safety Instrumented Systems for the Process Industry Sector , ANSI/ISA 84.01- 2004, 2004. C.6 American Petroleum Institute, Risk Based Inspection , API RP-580, 1st Ed, 2002.

5.1 Senior Leader Element Grouping |165 Senior leaders need to conduct the corporate risk review process regularly and follow up to close gaps. Management needs to oversee all functions and hold them accountable for perform ing their specific roles in the PSMS. The HR departm ent needs to consider process safety com petencies when screening and hiring new em ployees. Engineering must follow the applicable standards and RAGAGEPS when designing and installing equipment. Management and workers both have responsibilities for conduct of operations. Management defines procedures and standards, and controls for their consistent use. Workers commit to following the procedures and standards without variation by shift or unit. B oth comm it to performing their duties alertly, with due thought, full knowledge, sound judgment, and a strong sense of pride and accountability. Signs of effective conduct of operations include: Consistent practice of established work processes and procedures, which are followed, Effective shift turnover practices, Consistent and proper use of safe work permits to control work, Effective and consistent use of interlocks, bypassed only with proper evaluation, Consistent use of bonding and grounding, Excellent general housekeeping, Few overdue action items; and No ad hoc trials or modifications. Conduct of operations is clearly linked to combatting the norm alization of deviance. It starts with an insistence that procedures should be followed. This must be supported with procedures that can easily be followed. Com mon problem s with procedures include confusing form at, language that is not easily • • • • • • • • • • • •

PROCESS SAFETY AND MANAGEMENT OF ABNORMAL SITUATIONS 13 operations personnel, including l essons learned and actions taken to improve managing of similar situations in the future. 2.3 ADVERSE OUTCOMES OF ABNORMAL SITUATIONS The frequent occurrence of abnormal situations increases the likelihood of process safety incidents at a facility. An abnormal situation often occurs as an early step in a series of events that lead to serious incidents. Industry and insurance company su rveys have indicated that the cost of the consequences of proce ss upsets and other unplanned events can range from $100,000 to many millio ns of dollars. While the cost of equipment damage may be claimable, depending on insurance coverage, the actual cost to compani es is likely significantly higher due to policy deductibles, business interru ption that may not be claimable, and possible reputational damage. A 2020 insurance study analyzed 137 incidents between 1996 and 2019 that resulted in major losses (> $50 million) in the onshore oil, gas, and petrochemical industries (Jarvi s & Goddard 2020). Figure 2.3 from the study shows the breakdown of cause of loss between mechanical integrity failure, unsafe main tenance, and operations. Figure 2.3 Breakdown by Loss Type

194 hazardous materials from a decommissio ned process that is left in place is the elimination of a hazard and a fi rst order IS concep t as described in Chapter 2. Even if some small am ount of hazardous materials cannot physically be removed because they have solidified and adhered to inside surfaces of the equipment or cannot be reached without dismantling the equipment, the removal step is a strong application of Minimization during decommissioning. Documentation of Status . The exact state of the decommissioned equipment must be clearly docume nted, using the management of change process or an equivalent, so that at some point in the future, possibly years downstream, any actions taken to recommission, modify, or dismantle the equipment can be done safely. This is a form of the inherently safer strategy of Simplification . Example 8.1 A 50-gallon stirred pot reactor was used for the production of sodium aluminum hydride. In the presence of water, sodium aluminum hydride reacts exothermally enough heat to cause the hydrogen that is released to explode. The reactor was emptied, cleaned thoroughly (by report), and then placed in an outdoor surplus equipment yard with the nozzles open to “weath er” the equipment. About one year later, a maintenance man was orde red to clean up the reactor in preparation for reuse. He was told to put on full protective fire gear before opening the vessel. He did not don this PPE and proceeded to open the vessel and wash it out with a fire hose. An explosion resulted when water dislodged crusted-ov er sodium aluminum hydride trapped in a nozzle. The worker was burned, requiring a two-week hospital stay and several months of recuperation. Attention must be given to the lo ng-term protection of people or the environment from the hazards of abandoned equipment. Equipment that meets the criteria for disposal in a landfill, i.e., it has been properly cleaned, may not be suitable for other uses. Problems such as the one related in the fo llowing example can be avoided by

69 6 IMPLEMENTING THE REAL MODEL “I read, I study, I examine, I listen, I reflect, and out of all this I try to form an idea into which I put as much common sense as I can.” —Marquis de Lafayette, French Nobleman and Military Officer Successful execution of the REAL Model requires both individual evaluation and corporate change. The most basic requirements for implementation are: • leadership support and involvement at all levels. • enough people with the proper knowledge and experience evaluating external and internal incidents • a workforce interested and motivated to improve process safety performance. If you are reading this book, you are probably a member of the second or third group in the list. For you to have success in achieving lasting improvement for the company or plant, it will be important for you to obtain not only leadership support and involvement in setting objectives and driving the changes that the model identifies but also the needed financial and human resources. If the company leadership team members have not yet bought into their roles in driving the PSMS, getting this support is an important first step. CCPS provides three helpful resources: • Process Safety Leadership from the Boardroom to the Frontline (CCPS 2019a). This book lays out the business case for process safety, describes what leaders at each level must do to fulfill their roles, and helps dispel many misconceptions leaders may have. Driving Continuous Process Safety Improvement From Investigated Incidents By CCPS and EI © 2021 the American Institute of Chemical Engineers

APPENDIX D – EXAM PLE CASE STUDY 377 The No. 2 diesel fire water pump was down because its batteries were dead. The dead batteries were detected and recharged during a monthly check two months prior to the incident, but they were not replaced or rechecked after that. Interviews suggest that the fire water pumps had not been repaired due to a mechanical department perception that, because of budgetary pressures the expensive repairs requ ired delaying until the first of the year. It is interesting to note that although several people knew that one fire water pump was impaired, no one person in the department knew that both pumps were impaired. In interviews, several upper management representatives stated that fire water pump repairs would be critical and would be completed immediately, so there is a mismatch between the employee and management perspectives on the severity of the budget constraints. (Asset Integrity & Reliability; Process Safety Culture) iv Catalyst Storag e Tank Failure The catalyst storage tank failed earlie r than would have been expected had the fireproofing insulation been in good condition and the relief valve been adequate for the fire case. Witnesses indicate that several sections of the insulation had either fallen off or had been removed from the tank 2–3 months prior to the incident. The in sulation had not been repaired. (Asset Integrity & Reliability) v Relief Valve Sizing A check of the catalyst storage tank relief valve sizing calculations indicates the valve was large enough for the fire case assuming the tank had fireproofing insulation, but it was undersized for an un-insulated vessel. The original relief valve design calculations could not be found. The relief valve may also have been compromised by improper maintenance or pluggage. The last relief valve preventative maintenance and pop test occurred five years prior to the inciden t. No records were found for years prior to this pop test. (Mechanical integrity) Although the system failed below its design pressure, the overfilling of Kettle No. 3 caused a higher than normal pressure in the system. There were several causal factors for the Kettle No. 3 system being filled completely: vi Operator Error The control room operator did not stop filling Kettle No. 3 at the normal level of 85%. (Human Factors: An operator error, but one that would be expected to occur over the normal life of a process) vii Safety Critical Equipment Inhibited The Kettle No. 3 high-level alar m was bypassed, so it did not annunciate or log to the DCS alarm log. The operators bypassed the alarm because it

238 Human Factors Handbook Common tactics are: • To limit the requirement for remote or verbal communication for safety critical tasks, such as by use of logs and shift handover forms. • To “chunk” information. An everyday example is the “chunking” of telephone numbers into three or more strings of three or four numbers each. • Speak at a moderate pace and with moderate volume. • Repeat aloud what has been said, to help reinforce the memory of the communication. This also allows the receiver to control the pace or the speed at which the speaker says each chunk of information. • Use familiar words, abbreviations, and codes that require less mental effort to memorize. • Ensure the time allowed for communication is enough for the recipient to make a record of each chunk of information, before communicating th e next chunk of information. Some examples of communication protocols include: • Having a word and/or time limit for each safety critical communication, such as 15 words or 30 seconds. • Requiring long messages to be chunked, with each chunk recorded or logged before saying the next chunk. 19.5.1 Repeat-back procedures Repeat-back may be implemented as a formal procedure for safety critical communications. Repeat-back helps store information in memory and creates an opportunity for the sender of information to spot that they have not been heard correctly, and to correct the communication error. Asking someone to confirm they ha ve heard and understood a message by saying “yes” is not reliable. The receiver may not realize that they have misheard the message. When using repeat-back: • The sender starts by saying the receiver’s name and then states their message. • The receiver repeats the message back. • The sender confirms the accuracy of the repeat-back or repeats the messa ge if it is not accurate. Use common plant lexicon.

4 • Process Shutdowns 61 involve contractors and other personnel responsible for construction, installation, and commissioning. For this reason, an effective “Pre- Operations Plan” will have been developed before the construction stage begins (i.e., during stages 1-3, Figure 4.2). These key activities for these earlier stages are described in more detail elsewhere [31, p. 174]. 4.4.6 Resuming operations after the handover When restarting the process equipment after a project-related or maintenance-related shutdown, the operations group depends on an effective handover from the grou p or groups working on the equipment. The special administrativ e procedures used to prepare the equipment for the duration of th e shutdown, if any, should be addressed so that the equipment is returned to its normal, safe, at- rest and idle state before start-up. This includes, for example, removing all blinds that had been added to isolate the equipment from other parts of the process equi pment during the project or maintenance work. When all the special equipment preparations have been reversed, the operations group then can use their normal start-up procedures to restart the pr ocess safely (see Chapter 3). Often the final step used to close-out a larger project is to review the project's successes and challenge s, including a lessons learned review to improve the facility’s project management system performance for future projects . Any issues which had to be addressed and solved during the projec t, including those during shut- down and start-up, are lessons th at capture the knowledge gained from the project’s experiences. At th is point, the project team officially phases out, the facility is hand ed over completely to the operations group, and the project is closed. This concludes the overview section of the project life cycle. In summary, the basic stages for every pr oject are as follows: planning

280 Human Factors Handbook Figure 22-2: Human Errors – categories In the Milford Haven Refinery explosio n, three human errors were evident: • Diagnostic error – misinterpretation of an abnormal event – operators were not able to recognize the severit y of the situation or to correctly interpret the alarms. • Decision error – incorrect decisions were made by individual(s), and not evaluated by a team. • Action error – individuals continued operating in a highly hazardous scenario for several hours.

2 • Defining the Transition Times 21 Table 2.2 Definitions for the transi ent operating modes (Continued). A transient operating mode: The time that may require procedures in addition to the normal start-up procedures before restarting the equipment after a planned project or maintenance shutdown. Note: If other groups were involved in the planned shutdown, such as engineering, maintenance, or contractors, the special permits and handover procedures implemented for the shutdown-related activities must be reversed, reviewed, and authorized before resuming operations. This includes performing, as needed, equipment Pre-start- up Safety Reviews (PSSR) and Operational Readiness Reviews (ORR). A transient operating mode: The time that requires procedures in addition to the normal shut-down procedures for stopping the equipment in preparation for a major project, a major process unit shutdown, or a major facility turnaround or outage. Note: If other groups are involved in the extended shutdown, such as engineering, maintenance, or contractors, special permits and handover procedures must be in place beforehand, as needed, before performing the shutdown-related activities. A transient operating mode: The time that requires procedures in addition to the normal start-up procedures before restarting the equipment after a major project, a major process unit shutdown, or a major facility turnaround or outage. Note: If other groups were involved in the extended shutdown, such as engineering, maintenance, or contractors, the special permits and handover procedures implemented for the shutdown-related activities must be reversed, reviewed ,and authorized before resuming operations. This includes performing, as needed, equipment Pre-start- up Safety Reviews (PSSR) and Operational Readiness Reviews (ORR). A transient operating mode: The time when the operations team can use normal or specially-designed shut-down controls and procedures. These shut-downs can occur when: 1) the process cannot be successfully recovered from an abnormal situation and the normal shut-down procedures can be used 2) there is time to prepare the facility for a pending natural hazard (e.g., a hurricane or cyclone) Note: If other groups are involved in the unscheduled shutdown, such as maintenance, special permits and handover procedures must be in place beforehand, as needed, before performing the shutdown-related activities7Shut-down activated for an unscheduled shutdown4Start-up after a planned shutdown 5Shut-down designed for an extended shutdown 6Start-up after an extended shutdown

364

6.5 References |237 6.5 REFEREN CES 6.1 Center for Chemical Process Safety (CCPS), The Business Case for Process Safety, 3rd ed. , American Institute of Chemical Engineers, 2007 6.2 Baker, J .A. et al., The Report of BP U.S. Refiner ies Independent Safety Review Panel , J anuary 2007 (B aker Panel Report). 6.3 Contra Costa County (CCC) Industrial Safety Ordinance , County Ordinance Chapter 450-8 (as amended). 6.4 Center for Chemical Process Safety (CCPS), Guidelines for Auditing Process Safety Management Systems, American Institute of Chemical Engineers , 2010. 6.5 UK HSE, A review of safety culture and safety climate literature for the development of the safety HSE Health & Safety Executive culture inspection toolkit, Research Report 367, 2005. 6.6 UK HSE, Development and validation of the HMRI safety culture inspection toolkit, Research Report 365, 2005. 6.7 UK HSE, High Reliability Organisations – A Review of the Literature , Research Report HR899, 2011. 6.8 Canadian National Energy B oard (CNEB ), Advancing Safety in the Oil and Gas Industry - Statement on Safety Culture, 2012. 6.9 Center for Chemical Process Safety (CCPS), Vision 20/20 Self- assessment Tool , American Institute of Chemical Engineers, 2015 6.10 Mathis, T., Galloway, S., STEPS to Safety Culture ExcellenceSM, Wiley, 2013. 6.11 Hopkins, A., Disastrous Decisions – The Human and Organisational Causes of the Gulf of Mexico Blowout , CCH Australia Limited, 2012. 6.12 Blair, E., American Society of Safety Engineers, Building Safety Culture – Three Practical Strategies , Professional Safety, November 2013.