Datasets:

hndc

/

AI4Hazard-small

like 0

TOOLS AND METHODS FOR MANAGING ABNORMAL SITUATIONS 141 Some general aspects of the Human Machine Interface (HMI) to consider: Are indicators for operation critical and safety critical process control parameters and process cond itions included, and if so, are they visually correct and readily observable? Are the graphics overwhelming due to an excessive number of low priority parameters and alarms? Is an alarm summary list always di splayed to the operator for quick review? Is there consistency between the control screen graphics with respect to layout, symbols, and colors? Is there good resolution and minimal glare? Are the procedure steps for manually initiated control panel actions proven and straightforward to bring the process back to a safe state? Is there an overview screen (lev el 1 graphic) available for rapid access from all other screens that provides key overall operating parameters in the event of an abnormal situation? Is it simple for the operator to drill down from the level 1 graphic into specific areas of interest? Can trends on critical process parameters be created and observed? Can historical operation repo rts be generated for review? The considerable variety of HMI designs can lead to human factors issues. For example, in some designs, red indicates “off” or “closed” and green means “on” or “open.” Ot her designs may adopt exactly the opposite convention, where red is “d anger”, i.e., “on” or “open”. Other systems use grey and black (similar to a P&ID). Control stations are continuing to advance with improvem ents in resolution, video display, and human interface such as a touc h screen. Some standardization has taken place more recently , as detailed in 5.3.2.

219 Figure 9.1: Security Layers of Protection or Defense in Depth 9.4 COUNTERMEASURES Countermeasures are actions taken to reduce or eliminate one or more vulnerabilities. Countermeasures in clude security officers, security barriers, technical security systems, cyber security, response, security procedures, and administrative controls to address the following strategies: Physical Security (Security systems, protective personnel, and architectural features that are intended to improve protection). Technical Security (Electronic systems for increased protection or for other security purposes including access control systems, card readers, keypads, electric locks, remote control openers, alarm systems, intrusion detection equipment, annunciating and reporting systems, central stations monitoring, video surveillance equipment, voic e communications systems, listening devices, computer secu rity, encryption, data auditing, and scanners).

PROCESS SAFETY REGULATIONS, CODES, AND STANDARDS 43 Table 3.2. Sources of process safety related codes and standards and selected examples Source Examples American Chemistry Council “Responsible Care® Management System” and RC 12001 (ACC) American National Standards Institute (ANSI) ANSI/ISA-84.00.01-2004 Parts 1-3 (IEC 61511 Mod) “Functional Safety: Safety Instrumented Systems for the Process Industry Sector”. ISA-84.91.xx - a series of normative standards and guidelines regarding safety controls, alarms, and interlocks. American Petroleum Institute Recommended Practices (API) API 520, “Sizing, Selection, and Installation of Pressure Relief Devices” API 521, “Pressure Relieving and Depressuring Systems” API RP 752, “Management of Hazards Associated with Location of Process Plant Buildings” API RP 941, “Steels for Hydrogen Service at Elevated Temperatures and Pressures in Petroleum Refineries and Petrochemical Plants” American Society of Mechanical Engineers (ASME) “Boiler and Pressure Vessel Code” “Process Piping” B31.3 ASTM International Test methods for chemical and combustible dust properties Canadian Standards Group “Process Safety Management Standard” CSZ Z767 (CSA) Compressed Gas Association safety standards (CGA) Factory Mutual Data Sheets (FMG) FM Data Sheet 7-82N, “Storage of Liquid and Solid Oxidizing Materials” International Electrotechnical Commission (IEC) IEC 61508, “Functional Safety of Electrical/Electronic/Programmab le Electronic Safety-related Systems” (E/E/PE, or E/E/PES), 2010. IEC 61511, “Functional safety - Safety instrumented systems for the process industry sector, 2003”. International Organization for Standardization (ISO) ISO 9001 – “Quality Management Systems” ISO 14001 – “Environmental Management Systems” ISO 31000 – Risk Management ISO 45001 – “Occupational Health and Safety Management Systems”

17. Error management in task pla nning, preparation and control 201 greater sleepiness during the daytime. Task s requiring high levels of attention may best be scheduled for periods of alertness, such as 08:00 to noon, and not scheduled for periods of sleepiness, such as during night shifts. Table 17-3: Example tactics for enabling attention Task breaks Enable people to take a break before losing attention Noise Low levels of ambient noise Task sharing Switch tasks between people Task enrichment Redesign the task to increase the level of stimulation Ambient environment Maintain temperature (e.g., around 65 oF to 72oF/18 oC to 22oC) and humidity Lighting Higher levels of ambient lighting Task scheduling Schedule tasks requiring attention to higher energy times of the circadian rhythms Shift design Adopt good shift design to minimize fatigue See Chapter 15 for more information on shift design Automate high attention tasks Via control systems Alert or alarms Use to reduce demands on monitoring

Piping and Instrumentation Diagram Development 280 14.8.2 Selective Control When we talk about selective control, we mean that we want to select one signal from multiple signals to be pro- cessed by the controller. We can do this by using either a high selecting (> or HS) or low selecting (< or LS) opera-tor, which is situated on a primary signal (sensor signal). Selective control is used where there are several sen- sors installed to measure one unique parameter of a sin-gle stream, or “comparable streams, ” and one single sensor signal needs to be “selected” to be sent to the con-troller for control purposes. The selective control could select the highest value, lowest value, or a mid‐range value as the “selected sensor signal” for the controller. “Comparable streams” are streams that are completely identical. Generally speaking, there are no “comparable FY FCf(x) FY RSP“Non-prepared” SP FFCRSPFigure 14.16 Ra tio control symbology. Scheme 2 Less Po pular ÷ FY FCFFC Works as “ ” SP (Desired Ratio)Figure 14.17 Ra tio control: alternative scheme. FCFCFYf(x) RSP AC Figure 14.18 Ra tio control: FF + FB .

54 | 4 Examples of Failure to Learn 4.34 Laris, M., Duncan, I. and Aratani, L. (2019). FAA Administrator Dickson pressed on agency’s prediction of ‘as many as 15’ Max crashes possible. Washington Post (11 December 2019). 4.35 Maritime Executive (2018). BP's Deepwater Horizon costs reach $65 billion. www.maritime-executive.com/article/bp-s-deepwater-horizon- costs-reach-65-billion. (Accessed April 2020). 4.36 Noack, R. and Mellen, R (2020). What We Know About the Beirut Explosions. www.washingtonpost.com/world/2020/08/05/faq-what-we- know-beirut-explosions. (Accessed August 2020). 4.37 Quintão, A. (2019). CPI da Barragem de Brumadinho. www2.camara.leg.br/atividade-legislativa/comissoes/comissoes- temporarias/parlamentar-de-inquerito/56a-legislatura/cpi- rompimento-da-barragem-de-brumadinho/documentos/outros- documentos/relatorio-final-cpi-assembleia-legislativa-mg. (Accessed August 2020). 4. 38 Rddad, Y. and Snyder, J. (2019). Driver killed after fertilizer truck explodes in South Arkansas; Area evacuated after blast that was heard miles away. Arkansas Democrat Gazette (27 March 2019). 4. 39 Robertson, P.K. de Melo, L., Williams, D. et al. (2019a). Report of the Expert Panel on the Technical Causes of the Failure of Feijão Dam I. Agência Nacional de Mineração. 4. 40 Robertson, P.K. (2019b). Transcript of Video presentation by Dr. Peter K. Robertson, Ph. D., Chairperson of the Expert Panel on the Technical Causes of the Failure of Feijão Dam I. Agência Nacional de Mineração. 4.41 United Kingdom Department of the Environment (1975). The Flixborough Disaster. London, UK: Her Majesty’s Stationary Office. 4.42 Venkataraman, K. (2020). Gas leak at Haldia Plant, 3 workers die. Indianexpress.com/article/cities/kolkata/gas-leak-at-haldia-plant-3- workers-die/ (Accessed June 2020).

OPERATING PROCEDURES, SAFE WORK PRACTICES, 401 CONDUCT OF OPERATIONS, AND OPERATIONAL DISCIPLINE Figure 19.3. Conduct of operations model (Forest 2018) Conduct of Operations - The embodiment of an organization's values and principles in management systems that are developed, implemented, and maintained to (1) structure oper ational tasks in a manner consistent with the organization's risk tolera nce, (2) ensure that every task is performed deliberately and correctly, and (3) minimize variations in performance. (CCPS Glossary) Operational Discipline - The performance of all tasks correctly every time. (CCPS Glossary) Conduct of operations institutionalizes the pursuit of excellence in the performance of every task and minimizes variations in performa nce. Personnel at every level are expected to perform their duties with alertness, due thought, full knowledge, sound judgment, and a sense of pride and accountability. Conduct of operat ions is the execution of operational and management tasks in a deliberate and structured manner. The conduct of operations includes operating procedures and safe work practices, and much more. Operational discipline is closely tied to an organi zation’s culture. It refers to the operations being conducted correctly, every time, by everyone in the organization. Operational discipline should not be confused with blindly following a procedure. Operators should think about what they are doing. Some procedures, such as for critical operations, are intended to be used in-hand (o ften in a checklist form at) and followed exactly as written. Other procedures provide general guidance and should be thoughtfully followed, but not literally step by step. For procedures th at are expected to be followed step by step, operators should be trained on how to obtain approval for a deviation from the procedure.

209 8.56 Lewis, R., Sax’s Dangerou s Properties of Industrial Materials, Wiley, 2005. 8.57 Lewis, B., and von Elbe, G., Combustion Flames and Explosions of Gases, 3rd Edition, Academic Press, 1987. 8.58 Medard, L., Accidental Explosions, Volume 2: Types of Explosive Substances. Ellis Horwood Limited, 1989. 8.59 Mizerek, P., Disinfecti on techniques for water and wastewater. The National Environmental Journal, 22-28, 1996. 8.60 National Fire Protection Association (NFPA 2018). Guide for Venting of Deflagrations. NFPA 68, 2018. 8.61 National Fire Protection Association (NFPA 2014), Explosion Prevention Systems, NFPA 69, 2014. 8.62 National Fire Protection Association (NFPA 2018a). Flammable and Combustible Liquids Code . NFPA 30, 2018. 8.63 National Fire Protection Association (NFPA 2004). Recommended Practice for the Classification of Combustible Dusts and of Hazardous (Classified) Locations for Electrical Installations in Chemical Process Areas, NFPA 499, 2004. 8.64 National Fire Protection Association (NFPA 2006). Standard for the Prevention of Fire and Dust Explosions from the Manufacturing, Processing, and Handling of Combust ible Particulate Solids NFPA 654, 2006. 8.65 National Institute for O ccupational Safety and Health (NIOSH), DHHS (NIOSH) Publicatio n No. 2005-149, Pocket Guide to Chemical Hazards, 2005. 8.66 National Institute for O ccupational Safety and Health (NIOSH), Registry of Toxic Ef fects of Substances (RTECS); www.cdc.gov/niosh/rtecs/default.html 8.67 American Institute of Chem ical Engineers (AIChE), Center Chemical for Process Safety (CCPS), Chemical Reactivity Worksheet 4.0.3 (CRW 4.0.3) (2019)

30 | 2 Core Principles of Process Safety The Imperative for Process Safety did not measure up to the imperative for occupational safety, a theme that will be often repeated in this book. However, all the other cultural gaps trace back to a failure to Provide Strong Leadership for process safety. Workers attempting to control the well did not have the physical capability to perform the tasks required. This speaks to a weak Sense of Vulnerability as well as insufficient Understanding and Action on Hazards and Risks . These two gaps also influenced the over-reliance on the un-tested B lowout Preventer as a single safeguard. Investigators also noted numerous short-cuts from intended procedures that had become routine (Normalization of Deviance ). This incident also highlighted issues in Open and Frank Communication between the owner-operator and the contractor, leading to poorly defined assignment of process safety roles. In this book, the terms process safety leaders and leadership apply to the senior executives of the com pany and the line organization. Clearly, process safety professionals are also expected to provide leadership as well. However, time and again, experience shows how driving process safety excellence through the line organization achieves the best results. The Organization of Economic Cooperation and Development (Ref 2.8) highlighted the vital role senior leaders play in a process safety program : “Strong leadership is vital, because it is central to the culture of an organisation, and it is the culture which influences employee behaviour and safety. Process safety tasks may be delegated, but responsibility and accountability will always remain with the senior leaders, so it is essential that they promote an environment which encourages safe behaviour.

Part I Normal Operations Guidelines for Process Safety During the Transient Operating Mode: Managing Risks during Process Start-ups and Shut-downs . By CCPS. © 2021 the American Institute of Chemical Engineers

Appendix B - Major accident case studies 387 B.3 Longford gas plant explosion, Australia, 1998 As shown in Figure B-3, the Esso Longford gas explosion in Australia was an industrial accident with major consequences [20]. The incident resulted in two fatalities and eight injuries. It also cut th e gas supply for the state of Victoria for two weeks. A failure of a warm liquid (lean oil) syst em caused the temperatures of a heat exchanger to drop and become intensely cold and therefore brittle. When operators tried to reintroduce the warm oil, the brittle vessel fractured and released large quantities of hydrocarbon vapor, which found an ignition source, and exploded. Some key events and failures are noted after Figure B-3. Figure B-3 Longford Esso Gas Plant explosion (reproduced from www.icheme.org) The steps leading to this accident are as follows: • The accident was preceded by unsu ccessful attempts to repair leaky vessels. A heat exchanger, GP922, had developed flange leaks. Attempts were made to repair it. While the repairs were underway, another heat exchanger, GP905, lost its supply of lean oil. The temperature of GP905 had dropped to -54 °F (-48 °C), which is below its normal operating temperature of approximately 212 °F (100 °C). • The low temperature was caused by loss of lean oil flow in pump GP1. Hot lean oil flowing via GP905 would normally keep the temperature at optimal level.

9 • Other Transition Time Considerations 160 provided in Section 9.9. This chapter concludes with a discussion on how the RBPS elements apply to the start-ups and shut-downs associated with these specific transi tion times in life cycle-related projects. 9.2 A Life Cycle overview As was noted earlier, there are transi tion times in a project’s life cycle that have had significant inciden ts due to lack of proper project management during the handovers (C hapter 4). The two transition times discussed in this chapter are the initial start-up time, when the equipment or processes are being sta rted up for first time, and when the shut-down is being performed for the last time, when the equipment or process is to be mothballed or permanently removed. As was noted earlier, a well-conducted hazards review, such as a HAZOP, helped reduce the percentage of incidents involving incorrect design and straight mechanical failur es [2, p. 4]. The project-related transition times, the initial start-ups and final shut-downs, are depicted in Figure 9.1 on the illustration of the project life cycle (Figure 4.2). Thus, the initial start-up stage, depict ed as “Startup” project stage #6 in Figure 9.1, is when the chemical s and energies are being introduced for the first time. The shut-down activities associated with the “End of Life” projects are depicted in project stages #10 and #11. These project-related transition times du ring the project’s life cycle—the initial start-up and final shut-down—will be discussed in context of the equipment and process life cycle next. An equipment and process life cycl e can be represented with the eight distinct stages as illustrated in Figure 9.2: 1) design; 2) fabricate; 3) install; 4) commission; 5) opera te; 6) maintain; 7) change; and 8) decommission. (Sometimes the fabricate and install stages are combined and referred to as the construct stage. See definitions summarized in Table 9.1.).

Evaluating the Prior PHA 53 Note: The suggestions for evaluating HAZOP deviations discussed in the next topic are also germane. If docu mentation issues are identified, and particularly if they are common, consideration should be given to Redoing the analysis. In fact, it is common for teams to Redo checklists even when using the Update approach elsewhere. For example, if an Update primarily focuses on incorporating MOCs into the PHA, the fact that field labelling has degraded due to weathering may be overlooked. One way to ensure the revalidation team considers the current condition of labelling is to Redo the human factors checklist. HAZOP. The HAZOP method uses deviations created by combining guidewords (high, low, no, other than, etc.) with process parameters (flow, level, temperature, pressure, composition, etc.). Questions that should be asked when reviewing the quality of HAZOP studies include: • Are the nodes or sections define d in a manner that deviations have clear meaning? If the HAZOP worksheet is quite complex, it may be that the prior PHA team used process sections that were too large. Therefore, each deviation had to be repeated multiple times to ensure all relevant haza rds were addressed (e.g., High pressure in Vessel A, High pressure in Vessel B, High pressure in Vessel C). If the reviewer sees such repetition in a very large node, they should carefully look for instances that were missed. Conversely, deviations may have been misunderstood because the process sections were too small. In such cases, the causes, consequences, and safeguards ma y be divided among so many small nodes that the important eq uipment and system interactions were overlooked. • Have valid deviations that do not result in a consequence of interest been deleted, or were those rows left blank? Documenting which deviations do not cause issu es of concern, and why they are not of concern, is also valuable information for subsequent PHA revalidation teams to consider. There should be an entry in each row. If not, the PHA revalidation te am will need to address this issue by Updating those deviations. The deviations within the HAZOP worksheet can be sampled to evaluate the level of detail and logic of the analysis. Note: The questions listed apply to the What-If/Checklist and FMEA methods as well, with minor differences in column titles for each method. Relevant questions include:

262 INVESTIGATING PROCESS SAFETY INCIDENTS Historically, investigations have attempted to identify causal factors. This has helped ensure that specific cause is not repeated, preventing accidents. However, if the investigation root causes include human factors, then the identified issues that promp ted the human performance for that incident when addressed will apply to th ose for other potential incidents with similar performance requirements. Th is potentially prevents many more incidents. It can also improve employ ee morale, increase productivity, and complement positive cultural change. This chapter addresses the following human factors topics: • Human factors concepts • Incorporating human factors into the incident investigation process 11.1 HUM AN FACTORS CONCEPTS The term Human Factors is defined differently by va rious organizations. The CCPS glossary defines human factors as: “a discipline concerned with designi ng machines, operations, and work environments so that they match human capabilities, limitations, and needs. Includes any tech nical work (engineering, procedure writing, worker training, worker selection, et c.) related to the human factor in operator-machine systems” (CCPS, 2018). The UK Health and Safety Execut ive (HSE, 1999) defines it as: “environmental, organizational and job factors, and human and individual characteristics which infl uence behaviour at work in a way which can affect health and safety.” A common model for human factors in the process industries is shown in Figure 11.1. This model is included in the CCPS book Human Factors Methods for Improving Performance in the Process Industries (CCPS, 2007) and is based on the IOGP model (IOGP, 2005).

Appendix B - Major accident case studies 389 B.4 Milford Haven refinery explosion, Wales, 1994 On July 24th, 1994, a large explosion occurred at the plant of Texaco Refinery, Milford Haven in Wales, which caused inju ry to 26 people [87]. The blast from the explosion damaged properties in a 10 mile (16 kilometer) radius and was heard 40 miles (64 kilometer). The si te suffered severe damage to the process plant, the building, and storage tanks. The event was preceded by a severe electr ical storm that caused disturbance to the plant, affecting the vacuum distillati on, alkylation, and Butamer units, as well as the fluidized catalytic cracking unit. The explosion occurred some five ho urs later. The direct cause was a combination of failures in management, equipment, and control systems during the plant upset. These failures led to th e release of approximately 22 tons (20 tonnes) of flammable hydrocarbon from the outlet pipe of the flare knockout drum. The released hydrocarbon formed a clou d of vapor and droplets that found a source of ignition and consequently exploded. It took two days before the fires were finally extinguished. Figure B-4 The explosion and fires at Milford Haven (reproduced from UK HSE, [87]). Some key events and failures leading to the explosion are noted next: • On Sunday July 24th, 1994, at about 07:20, an electrical storm approached the Milford Haven area. This caused a series of interruptions to the power

38 runaway reaction. The processes described may involve trade-offs with other risks arising from other hazards. For example, the non-volatile solvent in the first example may be extremely toxic, and the solvent in the remaining examples may be water. Decisions on process design must be based on a thorough evaluation of all hazards involved. Hendershot (Ref 2.15 Hendershot 1995) compares the inherent safety characteristics of air an d automobile transportation, and concludes that automobile transportati on is inherently safer for reasons such as: The automobile, on the ground, will coast to a stop in case of engine failure, while the airpla ne will rapidly descend and may not be able to land safely. The automobile travels at a lower speed. The automobile contains a sma ller inventory of passengers. The control of an automobile is simpler (in two dimensions) compared to the airplane, which must be controlled in three dimensions. However, the benefits of air transp ortation, primarily speed, make it an attractive alternative for longer trips. These benefits have justified the expenditure of large amounts of money to provide extensive layers of protection to overcome the inherent hazards of air travel, including extensive redundancy in aircraft design and construction, the management of airspace by the air traffic control system, rigorous maintenance of aircraft and equipme nt, and many other systems. The result is that air travel, while inherently more hazardous, is, in fact, safer than automobile travel for long trips, even though the rate of fatal auto accidents dropped from1982-2005. Simila r situations can be expected to occur in the chemical/process indu stry (Ref 2.16 Hendershot 2006). In developing the definition of IST for the U.S. Department of Homeland Security that has been adap ted as the definition of inherent safety for this book above, CCPS incl uded the following caveats (Ref 2.13 CCPS DHS): ISTs are relative: A technology can only be described as inherently safer when compared to a different technology, including a description of the hazard or set of hazards being considered, their location, and the

Piping and Instrumentation Diagram Development 120 manual throttling valve. This arrangement is common for control valves and is called control valve station (Figure 7.19c). The basic arrangement of a control station is a control valve with two isolation valves on each side of control valve, and the bypass pipe is outside of the isolation valves with a throttling valve on it (Figure 7.20). Whenever there is a problem in control valve wherein the control valve should be pulled out from the operation and inspected or be sent to the workshop for maintenance, the isolation valves around the control valve are closed, and an operator stands up beside the bypass throttling valve and puts his hand on it while watching the sensor reading, the sensor that is used to send order to the‐under‐the‐maintenance control valve and adjust the opening percentage of the manual valve to mitigate the system. Here basically we used another control loop in the absence of the main control loop. However, this fake control loop is handled by an operator. Definitely an operator cannot work as a “control loop” for long period of time because it is very boring and tiring job. However, it is acceptable to ask an operator to take care of the con­trol valve duty for short period of time when the main control valve is under maintenance. It is obvious that the manual bypass valve is fully closed in majority of time during the life of a process plant. We know that throttling valves are not very reliable as a tight shutoff device. Therefore, in critical services, another valve could be placed upstream of the bypass valve to work as a dependable tight shutoff blocking valve. The example of such critical services could be toxic services, aggressive chemical services, or high pressure steams. It will be discussed in Chapter  8 that venting and draining of the control valve station is important before performing any inspection or maintenance on it. There­fore, the operator should fully drain and vent the system. The important question is how to drain and/or vent this part of pipe to make sure there is no chance of liquid splash­ing or gas pushing during inspection and maintenance. In plenty of cases the control valve flange sizes are smaller than the pipe sizes. Therefore reducer and enlarger are needed in the size of the control valve. If the pipe size is large, it is a good idea to use eccentric reducer and enlarger instead of less expensive concentric reducer/enlarger. This trick provides full drainage for the system. Full drainage of the system is important when the service is liquid or contained liquid or an aggressive fluid and Type vs. Block valveBlock valve Bypass valve If tight shutoff is requiredControl valve D DFigure 7.20 Details of c ontrol valve station.Throttlingmanual valve Spare Operating2×100%(a) (b) (c) Figure 7.19 (a–c) Diff erent arrangements for control valves to provide reliability.

4 day-to-day process risk management strategies. Chapter 11 describes the relationship between the fo ur main IS strategies, i.e., Substitution, Minimization, Moderation, and Simplification and each element of a PSM/Risk-Based Process Safety (RBP S) program. Chapter 12 addresses tools available to assist with i mplementing these IS strategies. Chapter 13, entitled “Inherently Safe r Design Conflicts,” describes the conflicts that often develop between the various attributes of safety, operability, cost, as well as othe r risk parameters and the ways to understand and make decisions consid ering those constraints. With the advent of regulations requiring in herent safety consideration or implementation, Chapter 14 was writ ten to help guide regulators and industry through the various cons iderations and challenges of IS. Chapter 15 contains worked examples of IS study methods and case studies to show a step-wise process that can be followed for an IS evaluation. It also gives practical examples of successful implementation. Lastly, Chapter 16 describes potential future IS initia tives, including needs, research, expected practice issues, and regulatory issues. 1.4 HISTORY OF INHERENT SAFETY Inherent Safety is a modern term for an age-old concept: to progress towards eliminating or reducing ha zards rather than accepting and managing them. This concept goes back to prehistoric times. For example, building villages near a river on high ground, rather than managing flood risk with dikes and walls, is an inherently safer design concept. There are many examples of milestones in the application of inherently safer design. For example, in the United States back in 1866, following a series of explosions invo lving the handling of nitroglycerine, which was being shipped to California for use in mines and construction, state authorities quickly passed laws forbidding its transportation through San Francisco and Sacramento. This action made it virtually impossible to use the material in the construction of the Central Pacific Railroad.

4 • Process Shutdowns 67 has led) to unexpected and significant incidents (sometimes years later). Examples of different materials of construction related incidents have been reported elsewhere [2 0, p. Chapters 16 and 28]. These include using a titanium flange on a line carrying dry chlorine, using carbon steel instead of stainless steel or suitable alloy, adding acid to an aluminum tanker assuming it was made out of similar-looking stainless steel, and using off-speci fication bolts. These materials could have been specified incorrectly du ring the design, been purchased incorrectly because they were less expensive, or installed incorrectly since they were not located in dedi cated warehouse space, or simply installed incorrectly since they lo oked similar. For these reasons, a facility should verify that the correct design, purchase, acquisition, and installation of the component or eq uipment occurs during each group handover before start-u p of the process. Effective handovers can occur when there is a review between the Construction Stage 5, the “Startup” Stage 6 [in this guideline, Start-up is used], and the Operations Stage 7 before resuming operations (Figure 4.2). The first set of topics that a gatekeeper could cover with the review team, for the handover fr om the construction stage to the commissioning and start-up stage, is li sted in Table 4.1. The next set covered, before the operations gr oup starts the equipment back up (Stage 7), is the handover from th e commissioning and start-up stage to the operations stage (Table 4.2). These tables present typical issues that a review team may address as a part of their protocol when covering each stage, thus helping re duce the likelihood of an incident upon restart.

Piping and Instrumentation Diagram Development 286 You may have two different control loops, each with a control valve that competes for control. You can merge the two control valves into one and then rank each con-trolled variable as a basis for override control. In this context, override control is basically an attempt to control a system when the single loop generates inter - ference. In such cases, we generally that say the solution is to implement “model predictive control” (MPC), but override can be used in very simple cases. Override control is used primarily for optimization or automatic start‐up of units. Sometimes it is implemented in units to upgrade and improve them if the units show frequent process upset, equipment failure, trip, or PSV actions. Override control is also used for cases where the mag- nitude and/or extent of the disturbance cannot be pre-dicted accurately. Some example are flows coming from underground (oil and gas extraction) or flows in utility networks with multiple users. Actually, the second example is a very common exam ple of the application of override control: where a utility consumer is suspected to be using more of the utility than expected and it may drop the pressure of util-ity stream (and impact several other utility users), an override control may be implemented to isolate the “badly behaved utility user” from the utility network and protect the other utility users as soon as its usage goes beyond a reasonable level. The other examples of using override control are listed in the Table 14.9. 14.8.3.2 Limit Con trol Limit control can be considered to be another form of override control. However, the override signal doesn’t come from the plant; it is manually set at a particular value inside the controller. If that value is reached then limit control is activated. A limit signal is a fixed and wise signal, which takes care of the control duty when the primary signal goes to its crazy side. It is not used in most plants, but its main application is when start‐up and shutdown of plant is automated, or in the normal operation of highly auto-mated plants.Limit control is used mainly for plant safety and opti- mization, but it can be replaced by an SIS. Figure 14.28 shows an example of limit control. In this example, we have a temperature control loop for a fur - nace. The temperature controller adjusts a control valve on the air feedline to the furnace. A limit control value of 180 °C is s et on the temperature controller to make sure the internal tubes won’t burn out. 14.8.4 Split Range and P arallel Control We learned that to design a control loop we need to find one stream as a manipulated stream to put a control valve on. There could be cases where more than one flow rate affects a single “process variable” (PV). In such cases, we are normally able to select the one that affects the PV most strongly and eliminate the others from considera-tion. However, where you have two streams that are equally influential, you need to implement split‐range or parallel control through control valves on both process variables (both streams). Split‐range or parallel control is introduced where you have more than one manipulated stream/one control valve that you need to control. This is most prevalent in pressure loops on gas streams, but it has been imple-mented on liquid streams as well. In parallel control, both control valves work simulta- neously together while in split control, one control valve starts its work when the other control valve starts to reach its end point (closed or open), and doesn’t move when the second control valve is functioning. For both parallel and split control, there are two modes: straight and reverse. In straight mode, both control valves move toward opening or closing (simultaneously or Table 14.9 Pr ocess cases that call for the application of override control. To prevent process upset To maintain equipment integrity ●Drop in utility network pressure (for big users) ●Fouling ●Creation of two‐phase flow ●Drop of NPSHA and cavitation ●Fire heater tube burnout AirTTTCTY180°C > Figure 14.28 Example of limit con trol in a furnace.

Appendix 216 Process Safety Culture Compliance with Standards Process Safety Competency Workforce Involvement Stakeholder Outreach Process Knowledge Management Hazard Identification and Risk Analysis Operating Procedures Safe Work Practices Asset Integrity and Reliability Contractor Management Training and Perform. Assurance Management of Change Operational Readiness Conduct of Operations Emergency Management Incident Investigation Measurement and Metrics Auditing Management Review and Contin. Improv. 5 3 % 4 7 % 12345 67 89 1 0 1 1 1 2 1 3 1 4 1 5 1 6 1 7 1 8 1 9 2 0 5% 3% 6% 1% 2% 9% 15% 10% 3% 12% 1% 3% 8% 2% 3% 8% 4% 1% 1% 2% Year 35 31 16 12 20 3 6 30 52 35 10 40 5 12 27 8 11 29 15 4 3 7 See C7.6.3-1Millard Refrig. Sys. (CSB 2015a) C4.7.1-1DCRC Flash Fire (CSB 2015b)2015 1 11 1 1 (C4.7.1) (A.4-1)Replacing Oil and Valves (Sanders 2015; p. 153)Not Known1 11 1 1 (C4.7.1) (A.4-1)Tank Cleaning (Sanders 2015; p. 155)Not Known1 1 1 1 1 1 (C.4.7.1) (A.4-1)Vacuum Incidents, Cleaning (Beacon 2002)Not Known1 11 (C4.7.1) (A.4-1)Replacing Valves (Sanders 2015; p. 156)Not Known1 11 1 C4.7.2-1Maintenance Replacement (Sanders 2015)Not Known11 1 1 1 1 (C4.7.2) (A.4-1)Repaired, Cold Furnace (Kletz 2009; p. 327)Not Known11 1 1 1 1 (C4.7.2) (A.4-1)Water Pump Explosions (Beacon 2013a)Not Known11 1 1 (Shut-down) Not discussed in Chapter 4 (Start-up) Not discussed in Chapter 4 Pillar IV Learn from ExperienceIncident Elements Identi fied as "weak" (See Figure 10.3) No. of Identified RBPS Causes Chapter 1 Introduction Chapter 4 - Table 1.1 Modes 3, 4 Planned Shutdowns 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 incidents dur ing the transient operating mode.

296 11.12 EMERGENCY MANAGEMENT In recent years the scope, complexit y, and shear length of emergency response plans at many facilities ha s grown substantially. The reason for this growth is a combination of several factors, including the consolidation of separate emergency plans required by different regulations, e.g., process safety inci dents, environmental release events such as oil pollution, hazardous wa ste operations, naturally occurring events such as tornadoes, hurricane s, and other severe weather events, transportation events such as train derailments. Also, security-related events such as bomb threats and phys ical/cyber security breaches have been added to many emergency response plans. Additionally, supplemental procedures related to em ergency response, such as crisis management, business continuity, an d event recovery have sometimes been made part of the same plans. Non-mandat ory federal government guidance has contributed to this growth, with the wide-spread use of the federal “One Plan” guidance published in 1996 (Ref 11.14 Fed Reg). As with operating procedures , reducing the complexity of emergency response and emergency action plans where possible will make these vital plans easier to use, particularly during upset/emergency situations where th e tension is higher. This is an important application of Simplification . The review of modification of emergency response plans is also an opportunity to challenge and possibly eliminate/minimize some of the hazards addressed in the plans. For ex ample, review of the fire fighting and response procedures (if the fa cility responds to fires) is an opportunity to employ the IS strategy of Minimization t o r e v i e w a n d possibly eliminate ignition sources, or to use distance to the maximum extent and segregate fire hazards if possible. It is also an opportunity to review the fixed and mobile equipme nt that supports the emergency response plan. For example, can the ma terials of construction of the fire main be upgraded from cast iron to more modern materials to resist corrosion (i.e., Substitution ), can explosion barriers be used, or occupied building structures be upgraded to withstand higher pressures (i.e., Moderation), or can the response equipment be simplified or storage locations improved to increase the speed of access (Simplification).

491 Table E.3. Material Releas e Threshold Quantities (reformatted from API RP 754) Threshold Release Category Material Hazard Classification Option 1 Material Hazard Classification Option 2 Threshold Quantity (outdoor) Threshold Quantity (indoorb) TRC-1 TIH Zone A Materials H330 Fatal if inhaled, Acute toxicity, inhalation (cat 1) Tier 1: ≥ 5 kg (11 lb) Tier 2: ≥ 0.5 kg (1.1 lb) Tier 1: ≥ 0.5 kg (1.1 lb) Tier 2: ≥ 0.25 kg (0.55 lb) TRC-2 TIH Zone B Materials H330 Fatal if inhaled, Acute toxicity, inhalation (cat 2) Tier 1: ≥ 25 kg (55 lb) Tier 2: ≥ 2.5 kg (5.5 lb) Tier 1: ≥ 2.5 kg (5.5 lb) Tier 2: ≥ 1.25 kg (2.75 lb) TRC-3 TIH Zone C Materials H331 Toxic if inhaled, Acute toxicity, inhalation (cat 3) Tier 1: ≥ 100 kg (220 lb) Tier 2: ≥ 10 kg (22 lb) Tier 1: ≥ 10 kg (22 lb) Tier 2: ≥ 5 kg (11 lb) TRC-4 TIH Zone D Materials H332 Harmful if inhaled, Acute toxicity, inhalation (cat 4) Tier 1: ≥ 200 kg (440 lb) Tier 2: ≥ 20 kg (44 lb) Tier 1: ≥ 20 kg (44 lb) Tier 2: ≥ 10 kg (22 lb) APPENDIX E - CLASSIFYING PROCESS SAFETY EVENTS USING API RP 754 3RD EDITION

316 | Appendix E Process Safety Culture Case Histories foster mutual trust the cause of the observed poor workforce involvement, communication, m anagement oversight, and training? Understand and Act Upon Hazards/Risks. E.30 What We Have Here is a Failure to Com m unicate A plant producing an ingestible product from non-hazardous raw materials ruined a significant quantity of product by accidentally contam inating the product with facility wastewater. The error was detected while the product was still in the warehouse, so no custom er was harmed, but if not caught m any people could have been injured. The process was implemented in equipment originally built for another process. The process tank had an overflow line that discharged to the facility sewer below the water level, which helped m inim ize odors related to the old process. Because the new process used vacuum charging of raw materials, the MOC process and the PHA identified that the overflow line needed to be removed. The m inim al modification was perform ed via a simple m aintenance work order. After the process had been running for some tim e, the plant engineering department conducted an equipment audit and noticed that the overflow line, which was still shown on the P& ID, was missing. Over a weekend shutdown, they brought in contractors to replace the line. On M onday m orning, the operator noticed im mediately that the line had been replaced. He halted production until the line could be removed again by m aintenance. Several months later, Engineering reinstalled the line again, and the operator again noticed it and had it rem oved. Unfortunately, the third time the line was reinstalled, there was a new operator who did not recognize the change. When he pulled vacuum to charge the raw m aterials, he also siphoned wastewater B ased on Real Situations

202 GUIDELINES FOR MANAGING ABNORMAL SITUATIONS Figure 7.8 Schematic Drawing of the Separation Stages The batch distillation took place in a horizontal steel drum called “60 Still Base,” as shown in the schematic drawing in (HSE UK 1994). This 45 m3 vessel was 7.9m (26 ft) long, 2.7m (9 ft) diameter, and was fitted with three steam heater batteries (series of tubes) towards the bottom, with the lower battery about 230mm (9”) above the tank bottom and the othe r two some 430mm (17”) above the lower coil. The heater batteries were heated with steam that came from a 27.6 barg (400 psig) supply pressure, via a regulator that reduced the pressure below 6.9 barg (100 psig), wi th a relief valve set to lift at 100 psig. Figure 7.9 Schematic Drawing of 60 Still Base

136 Guidelines for Revalidating a Process Hazard Analysis Typically, for a facility siting checklist Update , these discussions include asking questions such as: • Were there changes to the process or process structures that have: o Adversely affected unit layout or spacing between process components? o Adversely affected the location of large inventories of hazardous materials (e.g., changes in normal inventory levels; receipt of material in larger containers; temporary storage of material in totes, drum s, trucks, or railcars)? o Introduced any new hazardous materials? o Adversely affected Motor Control Centers? o Adversely affected the location or construction of the control room or any other temporarily or permanently occupied buildings (e.g., engineering, maintenance, laboratory, administration)? o Resulted in any new hazardous events, or significant changes to previously identified hazardous events, with the potential for impacting occupied buildin gs? (within the PHA boundary, within the facility, or outside the fence line) o Added likely sources of ignition? o Introduced new electrical area classification issues? o Impaired the coverage of the firewater system or its activation? o Adversely affected the adequacy of drains, spill basins, dikes, or sewers? o Impaired access to, or warranted, installation of additional, emergency stations (e.g., showers, respirators, personal protective equipment [PPE])? o Created new facility siting conc erns that should be updated into the existing facility siting study for the entire facility? o Created the need for additional passive fire protection or impaired existing protection (e.g., for support structures)? • Have there been any changes outside the PHA boundary that could adversely affect facility siting of this process? (e.g., new construction of other operating units, new construction near the fence line, new release points, or discharge locations) • Has operating experience since the prior PHA identified any new hazardous events with the potential for impacting occupied buildings?

9 • Other Transition Time Considerations 184 Risk management system weaknesses: LL1) Recommendations from the in cident investigation included completely emptying equipment if being “mothballed” (out of service longer than usual), testing any residual material in the equipment before reusing it, and prevent deterioration of any residual residues with a protective barrier (in this case water or other solvent). Relevant RBPS Elements Process Knowledge Management Hazard Identification and Risk Analysis Operating Procedures Operational Readiness 9.7.2 Incidents when executing decommissioning efforts on mothballed equipment At the time of this guideline’s pu blication, no incidents had been identified which occurred during this transient operating mode. 9.8 Decommissioning considerations Decommission consideration will depe nd on the scope of the project and facility-specific factors. As mentioned in Section 9.2, robust handovers, especially those involv ing contractors, are essential to prevent incidents (also refer to Table 9.2). This includes the demolition activities, which are typically carried out by an outside demolition contractor. These specialized cont ractors should be qualified to operate the required demolition eq uipment, and, most importantly, they should have training on the hazards of the chemicals associated with the decommissioned equipment and the knowledge and ability to handle them (especially if the equi pment was not adequately cleaned).

14.9 References | 189 While Wai-Kee was deep in these thoughts, Mei and Andrew were discussing the first responders’ competition, which they called “Mission Not Impossible,” a play on the American movies starring Tom Cruise. Mei said, “I wonder who’s going to win this year’s competition?” Andrew replied, “I’m not really rooting for any particular team. It’s just fun to watch how these teams think so creatively about how to handle a situation. I think this competition has been a great success with all of the engagement.” Mei smiled and thought, “I couldn’t agree with you more. At the beginning, there were only four teams. Now, there are more than ten teams involved. It’s only going to get bigger as we continue to hold this event.” 14.9 References If so indicated when each incident described in this section was introduced, the incident has been included in the Index of Publicly Evaluated Incidents, presented in the Appendix. Other references are listed below. 14.1 ARIA (2001). Explosion of ammonium nitrate in a fertilizer plant. www.aria.developpement-durable.gouv.fr/fiche_detaillee/ 21329_en/ ?lang=en. (accessed June 2020). 14.2 CCPS (2018). Guidelines for Siting and Layout of Facilities. Hoboken, NJ: AIChE/Wiley. 14.3 CSB (2013). West Fertilizer Explosion and Fire [Video]. 14.4 Mattson, S. (2007) 28 die as dynamite truck explodes in Mexico. San Antonio Express-News (10 September). 14.5 Rddad, Y. and Snyder, J., (2019). Driver killed after fertilizer truck explodes in South Arkansas; Area evacuated after blast that was heard miles away. Arkansas Democrat Gazette (27 March). 14.6 Tremblay, J.-F. (2016) Chinese investigators identify cause of Tianjin explosion. C&EN (8 February).

OPERATING PROCEDURES, SAFE WORK PRACTICES, 409 CONDUCT OF OPERATIONS, AND OPERATIONAL DISCIPLINE operating practices and procedures throughout a range of industries. A few of the topics include the following. (IChemE) Hazards of Steam, Fourth edition, 2004 Safe Furnace and Boiler Firing, Fifth edition, 2012 Hazards of Trapped Pressure an d Vacuum, Third edition, 2009 Confined Space Entry, First edition, 2005 Control of Work, Second edition, 2007 Safe Tank Farms and (Un)Loading Operations, Fourth edition, 2008 Safe Ups and Downs for Process Units, Seventh edition, 2009 Occupational Safety and Health Administration publications. The OSHA mission is safe working conditions. In addition to standards, OSHA publishes booklets, Fact Sheets, and QuickCards on safe work practices. Figure 19.6 is the OSHA QuickCard on permit-required confined space. (OSHA) Summary Process safety depends on the day-to-day ability of the organization to rigorously conduct operations correctly every time. The failure of one person in completing a job task correctly one time can, unfortunately, lead to serious inju ries and process safety incidents. The conduct of operations includes operators taking action s in the facility, engineers checking data, and managers planning work. They should all have the operational discipline to conduct work correctly, every time. Work can be performed correctly and consistently by following operating practices, safe work practices, and other practices and pr ocedures supporting communication, process oversight, and business performance. These pr actices and procedures should reflect the current facility design and operations. If they do not, a management of change process should be used to update them. Operating procedures provide step-by-step gu idance on operating the process including what should be done to keep it performing inside of operating limits and what to do if it exceeds those limits. Operating procedures sh ould address all phases of the operation. Safe work practices address maintenance, inspection, and other work that can defeat safeguards (such as breaking containment) and introduce hazards (such as hot work). Many other practices and procedures support safe operations including those that support communications across shifts such as shift handover and shift operating notes. As many tasks take place over mult iple shifts, having a clear record and understanding of what has happened and what should happen next is important to support safe operations.

Containers 157 of a container below the nozzle. It is important to know that the location of the nozzle and the location of the tip of down comer doesn’t change anything about the hydraulics of a system. The hydraulics of a system is dictated by the pressure or liquid levels in the source and destination. Down comers can be used in some liquid containers with dirty service liquid. A down comer in this case helps us to remove all the sludges from the bottom of a container. Riser: risers can be used as the internal portion of noz- zles if we intend to remove fluid from a specific portion of a container above the nozzle. For example, one common application of risers is for overflow nozzles. Generally speaking overflow nozzles on containers should be placed at the top of the container at a point that is higher than the HHLL. However, this location is not the best location from an operations point of view. It is very high and could have limited accessibility. One option is installing over - flow nozzles on the lower side of the container and put - ting an internal riser where its tip goes up to the HHLL. Extended nozzle: nozzle extensions can be used for the cases where, for whatever reason, the incoming fluid is not supposed to be in contact with the container walls before getting enough homogeneity with the rest of fluid in the tank. Elbowed nozzles are used for different reasons. It could be to dissipate the energy of the stream. The other reason could be to provide a good flow pattern in the container. Elbowed nozzles are not generally used for storage tanks. The above are only a few types of nozzle internals. The nozzle internals could be more complicated systems too. One example is a nozzle with a floating pipe (Figure 9.20). This system guarantees always taking liquid from the surface of the liquid bulk. 9.9.7 Nozzle Ex ternals The nozzles on containers could be connected to long pipes or instruments. These are not generally considered as “nozzle externals” per se.However, there are some cases where a short piece of pipe is connected to the nozzle from outside of the nozzle. One example is a gooseneck on vent nozzles. 9.10 Overflow Nozzles Overflow nozzles are obviously only for containers with liquid in them. Overflow nozzles are basically “safety sys - tems” to protect containers against filling out. Not every single container is equipped with an overflow nozzle. In fact, overflow nozzles are more common for tanks than vessels. However, even for some tanks the overflow noz-zle is not considered! There could be cases that we cannot afford to see overflowing liquid outside of them. For example it could be the case that it is very risky for a very flammable liquid to overflow into the dyke around the tank because of the chance of fire. In such cases we may decide to overlook the overflow nozzle and instead imple-ment a more reliable tripping system to protect the tank. The convention and simplest types of overflow nozzle are shown in Figure 9.21a. Later it was found in opera-tion that overflow systems create a lot splashing, which prevents operators from coming close to the overflowing tank. Then in the updated versions of overflow systems, a vertical pipe connected to the nozzle (Figure 9.21b). After a while they realized that in tanks that frequently overflowed, the overflow stream washes out of the tank foundation. An elbow is connected to the end to solve this problem (Figure 9.21c). Skimmed oil To 1100-P- 00201100-P&ID-002-03Figure 9.20 Nozzle with an in ternal floater.(a) (b) (c) Figure 9.21 Con ventional overflow system.

328 INVESTIGATING PROCESS SAFETY INCIDENTS Table 15.1. Requirement Compliance Checklist (USA OSH A/ EPA) (cont.) Requirement Statement Compliance? Five-Year Accident History (Additional EPA Requirements) Yes No (1) Date, time, and approximate duration of the release. (2) Chemical(s) released. (3) Estimated quantity released in pounds and, for mixtures containing regulated toxic substances, percentage concentration by weight of the released regulated toxic substance in the liquid mixture. (4) Five- or six-digit NAICS code that most closely corresponds to the process. (5) The type of release event and its source. (6) Weather conditions, if known. (7) On-site impacts. (8) Known off-site impacts. (9) Initiating event and contributing factors if known. (10) Whether offsite responders were notified if known. (11) Operational or process changes that resulted from investigation of the release and that have been made by the time this information is submitted in accordance with §68.168. (12) Level of accuracy. Numerical estima tes may be provided to two significant digits. In the UK onshore indust ry, the reporting of inci dents falls under RIDDOR (Reporting of Injuries, Diseases an d Dangerous Occurr ence Regulations 2013). These regulations clearly define th e types of incidents that should be reported and the records that must be kept, but does not cover the scope of the investigation. For major incidents involving processes that are covered by the COMAH (Control of Major Acciden t Hazards) regulati ons, regulation 26 (COMAH, 2015) provides instructions and high-level guidance on the investigation to be carried out by th e competent authority and supported by the facility owner/ operator. The UK HS E (Health and Safety Executive) has also published a guide for investigat ing incidents and accidents, (HSE, HSG 245, 2004) which includes a series of ta bles that could be used as a measure of compliance with recommended practice.

TOOLS AND METHODS FOR MANAGING ABNORMAL SITUATIONS 137 The skill of the control panel and fiel d operators is directly related to the effectiveness of their response to most abnormal situations. An example is provided by the Example Incident – Tower Flooding that was presented in Section 3.4.2.2 (see Exampl e Incident 3.10, Ch. 3). In that incident, most if not all the appropri ate information was available to the control panel operator, but the in stinctive response only served to aggravate the situation. Several step s might have been taken in advance to prevent the problem, such as: Educating the control panel operators on the principles of distillation columns, including flooding of this type. Providing additional information to the operator – for example, a pressure differential reading betw een the top and bottom of the tower that could indicate too mu ch vapor flow in the tower. Using process simulators to enhance operator training. When developing the training scop e and approach, introduction of abnormal situation management can be added as a training module. The concept of ASM® should be introduced as well as demonstrated with a few examples inherent to the chemic al process that is familiar to the personnel. Then ASM®, in addition to issues id entified from HIRAs, can be interwoven into frequently used and often recognized training tools and methods such as: Formal training manuals that address fundamentals such as process engineering parameters, basic process operating design, chemistry, chemical, and fire risks. Tabletop training exercises with desired responses on abnormal situations and scenarios. Emergency situations and response drills. Alarm response training. Process simulation of the process control systems. E-module training. One-on-one training.

94 Human Factors Handbook It is important to identify and consider all possible users during the design process. 9.3 Major accident example 9.3.1 What happened? The explosion at the Buncefield fuel st orage facility (2005) in Hemel Hempstead (North of London) was one of the biggest in peacetime Europe. It measured 2.4 on the Richter scale and was audible in Belgium, France, and the Netherlands. The fire engulfed over 20 fuel tanks, and the resulting smoke plume was visible from over 60 miles (97 kilometers) away. The devastation was enormous with many nearby properties damaged. Remarkably, there we re no fatalities, although 40 people were injured. During a gasoline filling operation to ta nk 912, several safety controls failed that should have prevented the tank being overfilled. Eventually large quantities of petrol overflowed from the top of one of the storage tanks. A vapor cloud formed, which ignited and caused a massive explosion and a fire that lasted five days. Figure 9-1: The Buncefield fuel storage facility before and after (reproduced from HSE [38]) The accident report is available from the UK Health and Safety Executive [38]. 9.3.2 Why did this happen? Many factors contributed to the explosion including organizational failure, and issues with design and maintenance of the overfill protection and containment system. One issue was the poor design of the interfaces and screens used by control room staff to monitor tank levels and the filling process. The United Kingdom’s Health and Safety Executive report [38] into the accident found that: • Only one visual display screen was av ailable. This meant that the status of only one tank could be viewed at a time, with information on other tanks stacked behind them. On the night of the accident, the display for tank 912 was at or near the back of the stack of tank displays.

Plant Process Control 309 By doing this type of control, flow surges created upstream of this control should be taken care of by – for example – placing a surge tank upstream. Better stream splitting can be done using a parallel‐range control strategy. This is shown in Figure  15.29. In  this arrangement, more flow goes through stream A than stream B. The flow loop can be replaced with a level loop on the upstream container. If you choose to split the stream based on the concept of “preferred destination, ” you can use the same schematic as in Figure  15.29, but using “split‐range control” rather than “parallel‐range control. ” The next example is a very important example of stream splitting, which is used commonly. In this arrangement, one stream is with level loop control from the upstream container, and the other stream is with a flow loop (Figure 15.30). The most important point here is that the flow control should be on the branch with the smaller flow and never the other way around. This is because it makes the con-trol far more accurate.One famous example of this technique is on the top of distillation towers. In Figure  15.31, you can see two arrangements for splitting condensate from the bottom of a condensate drum. In a typical distillation tower, the vapor coming off the top goes through a condenser and then to a drum. From the drum, the liquid is split into a reflux stream and another stream that goes to downstream equipment. The flow in each split stream will depend on the reflux ratio desired. The arrangements for either high or low reflux ratio must con-form to the rule that flow control should be on the stream with the smaller flow, and level control should be on the other. In Figure 15.31, the higher flow rate stream is shown with thicker lines for clarification. Figure 15.32 shows flow splitting of the stream based on the “preferred destination” concept. In this arrangement, the flow preferably goes to stream A. When there is a problem downstream that would make it difficult for stream A to take the flow, this manifests itself as an increase in pressure upstream. The  higher pressure is picked up by the pressure loop on  stream B, and in response the control valve on stream B starts to open to allow some flow to go through to stream B. In some designs, the control valve on stream B is replaced with a switching valve (on/off action). 15.7 Fluid Mover Control System There are two main classes of fluid movers: liquid movers and gas movers. The common control component for all fluid movers is capacity control. Capacity control means controlling the flow rate generated by the fluid mover. Fluid movers may have more control components in addition to capacity control. Stream AFC FC Stream B Figure 15.28 Flo w splitting: independent branch control. LC FC Smaller flowStream A Stream B Figure 15.30 Flo w splitting from a tank. Parallel Stream A Stream BFC Figure 15.29 Flo w splitting with parallel control.

218 Human Factors Handbook Figure 18-3: Factors contributing to error Cognitive bias Biases that impair cognitive processes, and prevent people making objective judgments about the situation and possible consequences. These include tunnel vision, confirmatory bias, similarity bias, and escalation of commitment. Positional authority/ Authority Bias This is based on title or role. Senior members of the team or organization make decisions or initiate actions that, in some cases, may not be correct. Subordinates are afraid to challenge the individuals in power, due to fear of repercussions, or due to cultural norms (e.g., do not challenge individuals in authority). Hofstede’s [74] cultural dimension of ‘power distance’ (i.e., the degree to which the less powerful members of institution accept and expect that power is distributed unequally) refers to the concept of positional authorit y. Individuals from cultures that value high power distance are very de ferential to figures of authority and generally accept an unequal distribution of power. Individuals from cultures demonstrating a low power distance readily question authority and expect to participate in decisions that affect them. Task focus Fixation on a task often leads to displaced situation awareness. Individuals miss additional clues fr om the environment, outside of the task in hand. Task focus often prevents errors being corrected. Time pressures Working under tight time schedules causes individuals to use procedural shortcuts and workarounds to meet these deadlines, which often leads to error. Stress High levels of stress impair individuals’ cognitive processes, such as not thinking clearly, not making objective judgments, and experiencing memory lapses. This may result in individuals failing to detect and correct errors. Limited self- scrutiny When individuals have completed the same task successfully on repeated occasions, they become complacent, believing everything will be fine, as it typically has been. 18.4 Coaching people to recognize risk of making errors 18.4.1 Training in error capture An appropriate level and type of training or coaching can improve an individual’s self-awareness. It can also decrease errors and helps to improve recovery from error. High hazard industries (including process industries) have started implementing or integrating error traini ng and coaching within their training programs. The success of error coaching depends upon the attitude of the recipient. A defensive response can shut down the error reduction effort.

14 PROCESS SAFETY IN UPSTREAM OIL & GAS The upstream industry exploration and production systems are diverse and complex, with many different designs to accommodate well conditions, variation in well fluid characteristics, processing conditions and water depth, if offshore. The scope of the upstream industry that falls within the scope of this book is mentioned previously in Figure 1-2 in Chapter 1. Wells both onshore and offshore produce oil, gas and/or condensate along with associated non-hydrocarbon gases, water and sand. The composition affects the well design, the surface operations, and the potential for proce ss safety events. Conventional reservoirs require a source rock containing hydrocarbons which migrate through porous rock or through fracture planes to a trapped area with an impervious seal rock where they accumulate . Absence of an impervious seal rock means any oil and gas dissipate and no reservoir forms. Good descriptions of important reservoir characteristics are provided in Franchi and Christiansen (2016). Gas and condensate are light and often free flowing to the surface. Oil is heavier and may be free flowing or require artif icial lift for its recovery. Due to high pressures in the well, gas and condensate may remain dissolved in the oil phase until the pressure drops near the surface. Crude oils are typically characterized by API gravity, which relates to specific gravity (SG). 2.1.2 Types of Upstream Facilities Upstream operations begin with exploration wells, usually after seismic surveys and interpretation of prior drilling at nearby wells, if available. If exploration drilling identifies a commercial prospect, then the well may be converted to a production well. Alternatively, it may be permanently abandoned, and a new production well drilled with specific casing diameters and lengths to match the production profile of the better understood reservoir charact eristics and anticipated flowrates. There are a wide variety of drilling and production facilities located onshore and offshore. ABB (2013) provides some useful examples. Large onshore fields may have many wellheads with outlet product flowlines that are then manifolded into gathering lines which merge with other lines and then flow into a large central product treatment facility (details on tr eatment are provided later in this section). Onshore facilities in warmer climates may be in the open, but if located in cold climates (e.g., Alaska), then these are typically housed in large buildings to provide weather protection for personnel and equipment. Smaller onshore fields are often of limited scale and are located on pads containing multiple wells. This is particularly true with wells that require fracturing. These may be connected by gathering lines to larger sites. There may be basic treatment facilities with further processing to occur at other sites. There may also be local storage tanks for liquid products awa iting export by truck, train or pipeline. Offshore facilities vary greatly in design due to water depth and the harshness of the climate. The Petroleum Engineering Handbook (2007) and ABB (2013) both

CONSEQUENCE ANALYSIS 281 Figure 13.7. Atmospheric stability effects (CCPS 1999) (Slade 1968) Table 13.5. Input and output for neutral and positively buoyant plume and puff models Input for puff model Input for plume model • total quantity of material released • atmospheric conditions • height of the release above ground • distance from the release • rate of release • atmospheric conditions • height of the release above ground, • distance from the release Output Time averaged concentration at specific locations (in the three spatial coordinates: x, y, z) downwind of the source

GLOSSARY 417 Causal Factor —A major unplanned, unintended co ntributor to an incident (a negative event or undesirable condition), that if eliminated would have either prevented the occurrence of th e incident, or reduced its severity or frequency. (Also known as a critical causal factor or contributing cause.) Cause —An event, situation, condition that results, or could result, directly or indirectly in an accident or incident. Chemical Process Quantitative Risk Assessment (CPQRA or QRA) —The quantitative evaluation of expected risk from potential incident scenarios. It examines both cons equences and frequencies, and how they combine into an overall measure of risk. The CPQRA process is always preceded by a qualitative syst ematic identification of process hazards. The CPQRA results may be used to make decisions, particularly when mitigation of risk is considered. Common Cause or Common M ode Failure —Failure, which is the result of one or more events, causing coincident failures in multiple systems or on two or more separate channels in a multiple channel system, leading to system failure. The source of the common cause failure may be either internal or external to the systems affected. Common cause failure can involve the initiating ev ent and one or more safeguards, or the interaction of several safeguards. Consequence —The undesirable result of a loss event, usually measured in health and safety effects, environmental impacts, loss of property, and business interruption costs. Consequence Analysis —The analysis of the ex pected effects of incident outcome cases, independent of frequency or probability. Deductive Approach —Reasoning from the general to the specific. By postulating that a system or process has failed in a certain way, an attempt is made to determine what modes of system, component, operator, or organizational beha vior contributed to the failure. Enabling Event —An event that makes another event possible. Sometimes used for enabling condition. The term enabling condition is preferred, since enabling conditions are not gener ally events but rather conditional states. Episodic Event —An unplanned event of limited duration. Event —An occurrence involving the process caused by equipment performance, human action, or by an occurrence external to the process.

136 PROCESS SAFETY FOR ENGINEERS: AN INTRODUCTION Types of Hazards (Beyond Chemical Hazards) The hazards encountered in the workplace incl ude more than solely chemical hazards (as discussed in Chapters 4 through 7). Recall the defi nition of process safety presented in Chapter 1. It also includes hazards related to energy. Thinking more broadly, the list of potential ha zards includes the following topics which will be discussed in this chapter. Kinetic energy Potential energy Electrical energy Meteorological and geological Health Kinetic energy hazards are those due to the motion of equipment. Many workplaces include large pieces of rotating equipment such as turbines. If these fail, they often fail catastrophically throwing debris which can caus e harm and destruction. A noteworthy incident is the failure of a turbine in the Sayano-Shush enskaya hydroelectric plant in Russia as shown in Figure 8.4 which resulted from heavy vibrat ion. (IWP&D 2010) Kinetic energy hazards also include moving vehicles which can impact equi pment and piping or trains that can derail. Process Safety - A disciplined framework for managing the integrity of operating systems and processes handling hazardous substances by applying good design principles, engineering, and operating practices. Note: Process safety focuses on efforts to reduce process safety risks associated with processes handling hazardous materials and energies . Process safety efforts help reduce the frequency and consequences of potential incidents. These incidents include toxic or flammable material releases (loss events), resulting in to xic effects, fires, or explosions. The incident impact includes harm to pe ople (injuries, fatalities), harm to the environment, property damage , production losses, and adverse business publicity. (CCPS Glossary)

287 thus not obvious to those who did not participate in the weighing of design options and therefore the design could be compromised, and the process safety risk increased by la ter changes where the rationale for the original design is unknown. Figure 11.3 Inherent Safety-Based Management of Change Process

6 • Recovery 112 design was insufficient to prevent overheating, resulting in the runaway reaction and overpres surization of the reactor. Relevant RBPS Elements Process Knowledge Management Hazard Identification and Risk Analysis Management of Change Operating Procedures LL2) Personnel and management accepted routine failure of the cooling system since the recovery efforts without a shut -down had always been successful and kept operations running within its normal operating conditions. Relevant RBPS Elements Process Safety Culture Asset Integrity and Reliability Conduct of Operations

144 | 11 REAL Model Scenario: Culture Regression equipment maintenance schedule and doesn’t have as much time to make sure that all the proper permits are filed when there is maintenance work. 11.1 Focus Charlotte was a recent chemical engineering graduate who just started her new job as a chemical process engineer on the offshore rig. Charlotte had always been adventurous and not one to shy away from challenges. She knew that working on a male-dominated rig had its problems, but this was her dream job. Charlotte was also vocal and used to sticking up for herself and her beliefs. One of her responsibilities was conducting risk assessments under the guidance of her boss, Lucas, the rig’s safety officer. Charlotte walked into Lucas’s office and said, “We’ve talked about conducting risk assessments, but they have primarily been focused on equipment failures. How would you feel about conducting one on human fatigue?” Lucas looked at her warily. He was pleased with her performance so far, but he felt she was always pushing him to do more. He said, “And what do you want to get out of this?” Charlotte replied, “The rotation is pretty tough on the crew, and from what I learned in university, human factors is often overlooked when it comes to safety. Fatigue results in slow reactions, reduced ability to process information, decreased awareness. Bottom line is that it can result in accidents.” Lucas thought to himself, “She has a point, and maybe this will help me figure out the cause of the recent mistakes a couple of the crew have made.” Lucas told Charlotte that first, he would have to check with Oliver, the production and maintenance supervisor. After all, Charlotte would need to conduct some personnel interviews with Oliver’s crew. Plus, he needed to get the approval of the rig manager, Mason. Lucas brought up the request with Oliver, who eventually agreed, but not without some pushback. “My crew already has a lot on their plates,” Oliver said. “Taking time out for interviews is just another thing for them to do.” Lucas replied, “We’re all under pressure. But think of it this way. Fatigue has been a common factor in major accidents. Last thing we want on our hands is something big to happen on the rig. Let’s make a compromise. Instead of interviews, let’s do it as a survey. We’ll set it up so that it doesn’t take more than fifteen minutes to complete.” Oliver said, “Fine, but I’ll hold you to it. No more than fifteen minutes.”

230 Human Factors Handbook 18.6.4 Recovery from error It is not possible to prevent all errors. Some tips on how to recover from error in an operational setting are shown in Table 18-5. Table 18-5: Examples of error recovery techniques Error recovery technique Practical execution Challenge perceptions The team takes an objective look at what is happening (based on evidence) and possible actions to take. Reset the team Team reset is required to: • Diagnose the current state of the team. For example, identify team conflict and team characteristics such as risk taking. • Address the elements that act as a barrier to effective teamwork, and that prevent the team from resetting. For example, conflict or blame culture. • Create team norms – set up new behavioral norms, such as peer checking of each safety critical activity. • Set team goals – create new goals for safety standards. For example, strive for 0% workplace injuries. Recognize the consequences of error Discuss consequences of error from various perspectives. For example, consequences for the employees, the company, customers, and the wider society. Encourage collaboration Ask team members to discuss what could have been done differently, and how another person’s contribution could have helped with detecting the error in a role. 18.6.5 Task verification Error prevention requires application of a risk-based approach to determine task verification needs. “Risk-based” means prioritizing or categorizing based on risk level. In this case, the task verification approach should be appropriate for the risk level. Task verification supports timely identification and correction of human failures. Some examples of task ver ification are shown in Table 18-6. It is important to ask an individual who is not part of the team or not involved in the task to conduct the task verification (e.g., independent verification). Inclusion of an independent person in task verification reduces bias and subjective judgment.

OVERVIEW OF RISK BASED PROCESS SAFETY 41 RBPS Element 5: Stakeholder Outreach This element covers activities with the community, contractors and nearby facilities to help neighbors, outside responders, regulators, and the general public understand the asset’s hazards and potential emergency scenarios. This activity may be coordinated by local regulators for smaller onshore upstream facilities, but usually is driven by the company for larger integrated facilities. It also applies offshore, but the stakeholders may be different. Large onshore facilities may be covered by national regulations and these require sharing of the possible hazard zones and emergency plans. Where there are joint owners, the outreach can resolve any differences in owner process safety policies. Example Topic: Stakeh older Communications A shale development in Colorado was located very close to suburban housing and caused major concerns to the community. The local regulator, in that case the City and County of Broomfield, worked with the company to hold a series of public meetings to address community concerns. A set of safety and environmental management best practices was agreed with the company, and this formed the basis of regulatory contro ls. The communication process was better than the company could have achieved on its own due to the strong emotions raised in the community and the need for a neutral party to facilitate contacts. RBPS Application Stakeholder Outreach directly addresses this issue and allows for alternative means for communication. 3.2.2 Pillar: Understand Hazards and Risk This pillar addresses process knowledge management and the identification of hazards and management of risks. Process knowledge must be readily available and kept up to date. Risk management is an extension of some regulatory requirements, but in practice most companies conduct h azard identification and add risk ranking as part of the analysis. Inherently safer design considerations are applied here. RBPS also notes that companies may choose to go beyond qualitative risk ranking to some form of quantitative risk assessment (LOPA or QRA). RBPS Element 6: Process Knowledge Management Process knowledge management is the assembly and management of information needed to perform process safety activities. It includes verification of the accuracy of this information and confirmation that this information is kept up to date. This information must be readily available to those who need it to safely perform their jobs. Upstream facilities, whether onshore or offshore, are often remote and knowledge and documentation of importance to personnel may be located in central offices. Access to this knowledge and do cumentation should be arranged and involve regular meetings or communication links. An incident where the importance

APPLICATION OF PROCESS SAFETY TO ONSHORE PRODUCTION 105 ●Sprinkler systems ●Fixed and semifixed foam systems ●Foam storage and application systems ●Portable fire extinguishers ●Special fixed extinguishing systems (e.g., FM200, INERGEN) Additional guidance is available in CCPS (2003) and in NFPA and API standards. Fire systems involving logic controllers are often defined as safety instrumented systems and guidance for thes e is available in IEC 61508 and 61511. If not covered by these standards, the equipment should be independent of the plant distributed control system. Fire Hazard Planning Fire protection at small unmanned production sites is often modest compared to large facilities, matching their needs econ omically. Larger facilities may use fire codes to define their systems. Another approach is to use Fire Hazard Analysis rather than simply to apply code recommendati ons. FHA is a scenario-based approach where potential leak events are assessed, a nd their impacts calculated. Each scenario should be developed and tested to verify that the planned fire responses are adequate and tested through tabletop and field drills. Some issues addressed in an FHA include the following. ●Calculating firewater required to apply onto the fire and simultaneously to cool adjacent exposures effectively ●Where fire hose is required, determining if firefighters can deploy and connect all the hose required in a specified time ●Determining if the drainage is sufficient to control spills of burning hydrocarbons, etc. ●Determining if firewater monitors have the range to effectively reach the fire and if remote operation is appropriate

279 Figure 11.1 Plug Valve Operator This incident demonstrated not only an important human factors issue, but also a weakness in the MO C program at the refinery where it occurred. Allowing a valve operator to be installed in the opposite manner that convention would require for this type of valve invited an incident to occur. When the valv e was re-configure d the MOC should have revealed this serious (and obvi ous) error but did not. Additionally, in this situation facilit ies should only procure and install quarter-turn ball/plug valves (a common valve application) where the valve is designed so that the valve operator can only be installed in the normal convention. Internal engineering specif ications for valves should reflect this convention (see Compliance with Standards, section 11.2 above). Another example of the importance of human factors in process safety is demonstrated by the fire at another refinery. This incident also involved valve operators. In this ca se, a gearbox on a pump suction in hydrocarbon service was malfunctionin g, which did not allow the valve to be opened. To correct the malfunctioning gearbox the device was partially disassembled in accord ance with refinery maintenance procedures. However, this particular valve was an older design (15 of 500 similar valves in the refinery) where a part called a top-cap was part of the pressure retaining boundary of th e valve and the support bracket for the gearbox also held the top-cap in place. On newer design valve gearboxes this was not the case. When the bolts were loosened on the top-cap and valve stem operated with a pipe wrench, a large release of hydrocarbon resulted, which ignited due to a welding machine nearby. The fire seriously injured four pers onnel in the area. Figure 11.2 shows the old and new designs and how th e removal of the support bracket helped precipitate the inci dent (Ref 11.18 CSB, 2016) In the past 10-15 years there has been a significant amount of research into developing quantitati ve methods for analyzing inherent safety. Amyotte and Kletz have summa rized efforts made by different organizations to develop these quanti tative inherent safety evaluation tools. These tools allow relative comparisons of the hazards and the ensuing risk from different processi ng options. See Chapter 12 for a description of these quantitative IS tools.

Piping and Instrumentation Diagram Development 134 You probably realized that “blinds” are the vital compo- nent of the different types of isolations. There are at least two types of blinds: spectacle blinds and spade blinds. Figure 8.5 shows a spectacle blind when it is in use. In the figure the pipe is obviously NOT blinded since the solid portion of the blind is up. Table  8.3 summarizes the different features of these blinds. It should be noted that in Table 8.3 the position of the blinds in the open and close positions are shown but in P&ID we only show the “open” position of blinds. A decision needs to be made about the right type of isolation method for each application. The isolation method depends on the type of item that needs to be iso-lated, the fluid type, and the pressure of the fluid in the item. The first parameter is the type of process element. The process element could be so small that even the hands of the operator cannot get into it. Such items need the least positive isolation. If the item is head‐in or worse, walk‐in, it needs a stronger isolation system. Walk‐in equipment is a piece of equipment that is large enough that a person can walk in it, but it is not designed for human to live in it. The technical name for such spaces is “confined space. ” “Confined space” is a concept that has a technical mean- ing and in different jurisdictions also has a legal meaning. A utility network or a pressure gauge can be isolated with a simple root valve as they are not voluminous. The second parameter is fluid type. If the fluid is non‐ innocent the more positive isolation system should be used. In technical terms, fluids like steam, acids, and toxic materials need a strong isolation system. Figure 8.5 Spectacle blind: outside of the pipe and in the pipe . Table 8.3 Fea tures of different types of blinds. Spectacle blind (see Figure 8.8) Spade blind (spacer) Real shape Open Close P&ID symbol Open CloseOpen Close Pros It can be seen by the operator whether a pipe is blinded or notNot heavy; easy to handle Cons Sometimes too heavy, especially for large pipe sizes with large wall thicknessThe operator cannot see whether the pipe is blinded or not; easy to misplace the mate Applications Only for smaller pipe sizes, typically less than 12″ Generally for larger pipe sizes, typically larger than 10″ When area 200 can run without area 300 (for t period) When area 200 can NOT run without area 300 (for t peri od) Area 300 pipe ra ck pipe lineArea 300 pipe ra ck pipe line Area 200Area 200TP-23 123 Figure 8.6 Isolation f or utility streams across different areas.

3 • Normal Operations 36 During abnormal operations, expected process deviations above or below the aim may approach or exceed the upper or lower safety limits . However, if process control is not recovered, these deviations may approach or exceed the upper or lower design limit of the equipment. When the operating conditions exceed these equipment design limits, catastrophic equipmen t failure may occur, releasing the equipment’s contents (a “loss of co ntainment” event). Such loss events on equipment handling hazardous ma terials and energies could lead to fatalities, injuries, environmental harm, and property damage. These responses to abnormal operat ions, introduced with the flow chart illustrated in Figure 1.3, is disc ussed in detail in Part II of this guideline, Abnormal and Emergency Operations. The normal operations timeline in Fi gure 3.2 is illustrated with the pressure control points in Figure 3.5. These specific upper and lower safety limits influence the pr ocess and equipment design and operation [1] [14] [21]. Figure 3.2 will be used in subsequent chapters as each of the following transient operating modes listed in Table 1.1 are discussed: Shut-down for a process shutdo wn (Type 3, in Chapter 4) Start-up after a process shutdown (Type 4, in Chapter 4) Shut-down for a facility shutdo wn (Type 5, in Chapter 5) Start-up after a facility shut down (Type 6, in Chapter 5) Shut-down for an unsc heduled shutdown (Type 7, in Chapter 7) Start-up after an unscheduled shut down (Type 8, in Chapter 7) Shut-down for an emergency shut down (Type 9, in Chapter 8), and Start-up after an emergency shut down (Type 10, in Chapter 8).

70 PROCESS SAFETY IN UPSTREAM OIL & GAS wide pressure variations in close proximity can result in losing fluid and taking a kick from the same completion zone. Workovers and intervention activities may also use explosive charges to rejuvenate a well, similar to completions. Improperly grounded systems can result in premature detonation of the explosive charges with damage to the wellbore systems or to personnel if accident ally triggered at the surface. Key Process Safety Measure(s) Safe Work Practices – same as Section 4.2.6 4.2.8 Depleted Wells Risks The SINTEF data reported in Section 4.2.2 shows 1% of blowout events occur during abandonment. Plugging wells in depleted reservoirs can involve multiple barriers. These barriers can fail allowing hy drocarbons to flow and eventually to exit the well or into the surrounding soil. Dusseault et al (2000) discuss reasons for loss of integrity and sustained casing pressure in oil wells. Key Process Safety Measure(s) Asset Integrity and Reliability : As with all equipment, barriers must be properly designed and maintained to ensure their long-term performance, with critical systems also providing a means to warn operators of impending or actual failure. 4.2.9 Mud Room / Shale Shakers, Well Fluid Handling and Treatment Locations Risks Release of flammable and sometimes toxic gas from the well mud returns or treatment systems is possible. Gas clouds can result where there is insufficient ventilation and fires and explosions can result if ignition occurs. Large scale onshore facilities which are fully enclosed have well developed ventilation systems for mud handling and treating systems. Onshore dr illing and production facilities are usually well separated. Offshore, some facilities conduct both well construction and production simultaneously and a SIMOPS plan is needed to address potential risks (see Section 5.2.5 for further details). Key Process Safety Measure(s) Hazard Identification and Risk Analysis : Companies with offshore deepwater or large onshore production facilities often carry out detailed fire and gas safety studies. This may be less common at smaller onshore facilities or unmanned shallow water installations. The HIRA studies apply to all aspects of well construction, not just surface handling. Provision of flammable and, if appropriat e, sour gas detection is standard, with alarms to the drill cabin as well as to the control room.

Table C-1 continued HF Competency Performance/ Knowledge Criteria Level 1 - Operator Level 2 - Supervisor* Level 3 - Manager** Overarching concepts, principles, and knowledge Human Performance and error Understands factors influencing human performance Can identify various factors influencing human performance Can recognize how specific factors (e.g., environment, task or environment related factors) contribute to/influence human performance Able to assess human performance and is able to review and mitigate impact of performance influencing factors Understands link between human performance and error Can identify various types of error (slips and lapses) and mistakes Can identify and mitigate impact of factors contributing to error Able to assess effectiveness of error mitigating strategies Cognition, memory bias, attention spans Cognition, memory bias, attention spans Has skills to minimize potential cognitive bias. Can recognize where tasks exceed human cognitive, memory or attention spans Can contribute to design of tasks to minimize potential cognitive bias and exceeding cognitive, memory or attention spans Can help develop new ways to minimize potential cognitive bias and to minimize exceeding cognitive, memory or attention spans

5.3 Process-Related Element Grouping |185 that drives this quest for continual im provement can be susceptible to normalization of deviance . MOC is essential to manage ongoing process improvement safely. M OC m akes sure that changes do not introduce new hazards, that the process continues to m eet the com pany’s engineering standards and risk criteria, and that all elem ents of the PSM S are updated to reflect the changes. In a strong process safety culture, MOC should be considered an essential and valuable activity. MOC helps ensure that changes to processes and personnel do not inadvertently introduce new hazards or increase process safety risk. M OC includes a formal review and authorization process that evaluates changes to equipment, process conditions, procedures, and organization, addresses any needs to improve safeguards, and assures that procedures, asset integrity, and process knowledge is updated to reflect the change. The MOC procedure should be applied for all changes that are not replacements-in-kind. A replacement-in-kind is a new com ponent, m aterial, or person that m eets the sam e specifications as the original. Com panies with a strong process safety culture will nonetheless consider replacements-in-kind carefully, to m ake sure that the original specification was sufficiently com plete. Sometimes, new raw m aterials and process com ponents m ay contain defects or im purities that are not addressed in the specification, but can cause process failures. For exam ple, incidents have resulted from castings sourced from countries with emerging economies that were identified as replacements-in-kind based on metallurgy but contained voids that led to premature failure. Leaders should design and enforce use of the MOC procedure as part of the imperative for process safety . This should be done in such a way that all employees come to appreciate the im portance of MOC and perform it with an appropriate sense of vulnerability ,

1 • Introduction 13 However, process upsets are anti cipated and accounted for with engineering and administrative controls designed and implemented to monitor and respond to these upsets. These abnormal operations can have successful or unsuccessful ou tcomes, either recovering the process back to its normal operatin g conditions and resuming normal operations or resulting in an unsch eduled process shut-down. As is depicted in Figure 1.3, the equipm ent and process may need to shut- down when the recovery efforts are unsuccessful, transitioning the process from its normal operating state to the shutdown mode. It is important to note at th is point that there is a temporary operating time, often designated as a stan dby mode, which provides time for process upsets to be diagnosed be fore either completely recovering the process conditions or sh utting the process down. As is depicted in Figure 1.3, there are two general types of shut- downs during abnormal operatio ns: normal shut-down procedures and emergency shut-down procedur es. Note that normal shut-down procedures can be used for both planned and unplanned shut-downs. Emergency shut-down procedures are, by definition, unplanned events and emergency situations. If a significant loss event occurs, the emergency shut-down procedures ar e activated, potentially resulting in impact to people, the environment, and assets. Part II of this guideline co vers the abnormal and emergency operations modes: Chapter 6 disc usses the successful response to abnormal operations—the recovery; Chapter 7, the shut-downs and start-ups afterward associated with unscheduled shutdowns; and Chapter 8, the shut-downs and sta rt-ups afterward associated with emergency shutdowns.

148 Guidelines for Revalidating a Process Hazard Analysis Example 9 – Defending Prior Work Study leaders should try to engage everyone in the team deliberations. This is particularly true in a revalidation beca use some of the team members may have a personal investment in the prior work (PHA, MOCs, incident investigations, etc.) that is being reviewed. For example, th e engineer on the revalidation team may have been the person responsible for an MOC being examined. They naturally know more about the details of the MOC and its conclusions than anyone else does, so the team will let them speak fi rst. However, the leader cannot let that explanation dominate or eliminate any furt her discussion. If no one else speaks up, the leader should direct specific questions to other team members, such as asking the operator whether the change resulted in any alterations to the startup procedure or asking the maintenance expert whether the change affected how the equipment is isolated and drained. Revalidation leaders should also intervene if they sense that team members know that the recommendations needed to reduce risk will not be well received or that they do not want to face the work that the recommendations will generate. Challenging prior risk judgme nts may provoke a confrontation with other team members who made those ju dgments and damage team cohesion. Team leaders should immediately redirect any questions about individual competence or motives in the past to the current factual issue(s), considering risk in light of what is known today. To move the revalidation forward, the leader may suggest that a contentious issue be resolved using a company-approved method such as LOPA. If LOPA confirms the risk is elevated with respect to the organization’s current risk tolerance, th e leader can then solicit team ideas for risk reduction. Example 10 – Assuming Similar Proc ess Equipment Has Identical Risks Where multiple pieces of equipment or multiple trains are nominally the same, a common PHA practice is to analyze one component or train as typical of all. That may have been true in the original design. For example, all the reactors may have been equipped with identical high pressure/temperature alarms. However, the alarm setting may have subsequently been changed for operational reasons, and the changes were not all the same. When the facilitator encounters nodes that were previously analyzed as “typic al” of similar equipment, the facilitator should closely question the revalidation team to ensure that the equipment has no differences that would significantly affect the risk judgments.

Piping and Instrumentation Diagram Development 274 14.7 Multi‐Loop C ontrol Architectures Up to now we have learned the architecture of a single loop control. In this section we will focus on the archi-tectures of different multi‐loop control systems. 14.7.1 Cascade C ontrol There may be some cases where we have to deviate from a simple, single control loop and use cascade control. Figure 14.2 shows a simple loop for level control on a tank, and how we could improve this by moving to cas - cade control. In the left‐hand schematic, there is a simple level loop on a container. Because of some “reasons, ” we need to move away from a simple control loop to a complicated cascade loop. In a cascade loop, the level controller out - put (and not the level sensor output) is used as the set point for a flow control loop on the pipe. To be able to talk about different features of cascade control we need to learn some terminologies of cascade control. Let’s look at (Figure 14.3).Another name for cascade control is reset control. The main loop is called the primary, or master, loop. The sec - ond loop is called the secondary, or slave, loop. We refer to the control architecture for this tank as level‐to‐flow cascade control. The “reasons” that forced us to move away from a sim- ple control loop can be lumped into two groups: some of them located in the container (group 1 problems) and some of them located on the controlled pipe (group 2 problems). These reasons are listed here. First four of these rea- sons lie in the area of the sensor and one of them (the last one, number 5) lies in the area of the control valve. 1) Lar ge fluctuations. The controlled variable in the pri- mary loop may have large fluctuations, making it dif - ficult to control. 2) Tig ht control. You may need to keep tight control over a narrow band of operation for the controlled parameter. One example could be control in a small vessel. Irrespective of the process, the level control of a small vessel needs to be tight. If tight control is needed, we may be forced to use cascade control. 3) Slow r esponse. The response to control action on the parameter may be slow. A parameter could be inherently slow, like temperature, or it could be slow in a specific application. For example, temperature sensors in large spaces are very slow, or, when a sen-sor is very far from the control valve the loop would be slow. Figure 14.4 shows the relative speed of parameters. If we are dealing with slow loops, we may have to use cascade control rather than single‐loop control. In such an arrangement, the slow parameter should always be in the master loop, and the faster parameter will be in the slave loop. 4) Resp onse rate. In some cases, we may want to limit the rate at which the controlled variable increases or decreases. 5) Se vere fluctuations in flow in the secondary loop. There could be a case where the control valve is placed on a pipe with severe flow fluctuations. In such cases, using cascade control is inevitable. One famous reason that a pipe may experience severe fluctuations is when that pipe is a header or SPSP LC FCLC 1 2 Figure 14.2 Mo ving from a simple loop to cascade control. RSPSP LC FC Secondary (Slave) LoopPrimary (Master) Loop FC Figure 14.3 Cascade c ontrol terminology.SlowestComposition < Temperature < Level < Flow MasterFastest Slave Figure 14.4 Sluggish c ontrol system.

200 | Appendix: Index of Publicly Evaluated Incidents Section 1. RBPS Elements (Continued) Operating Procedures—Primary Findings A1 C12, C20, C24, C28, C37, C40, C59, C70, C76 D20, D42 HB6 J11, J16, J18, J20, J30, J38, J39, J43, J43, J45, J48, J49, J50, J58, J65, J70, J78, J81, J103, J108, J120, J123, J124, J125, J127, J134, J146, J147, J149, J176, J181, J187, J200, J206, J208, J216, J217, J234, J235, J252, J255, J256, J257, J261, J263, J271 S8, S11, S14 Operating Procedures—Secondary Findings C2, C11, C26, C39, C57, C58 HB9 J15, J28, J31, J52, J53, J54, J56, J57, J69, J74, J87, J98, J107, J110, J130, J131, J133, J158, J171, J174, J177, J180, J182, J188, J197, J228, J237, J246, J254, J264, J268, J270 Safe Work Processes—Primary Findings A2 C8, C9, C10, C20, C22, C27, C44, C51, C54, C55, C67, C68, C71, C77 D18, D20, D32, D42 HA6 J7, J18, J37, J39, J40, J41, J42, J43, J46, J49, J50, J59, J64, J67, J79, J82, J89, J109, J114, J122, J148, J162, J165, J167, J174, J177, J183, J185, J186, J188, J190, J195, J197, J198, J199, J201, J203, J209, J223, J229, J242, J247, J253, J259 S1 Safe Work Processes—Secondary Findings A1, A4, A5, A7, A10 C16, C31, C32, C39, C42, C43 HB6 J15, J17, J19, J20, J23, J25, J26, J28, J34, J47, J51, J76, J97, J101, J111, J118, J128, J143, J144, J146, J176, J228, J234, J237, J252, J258, J261, J267 S4, S7, S15 Asset Integrity and Reliability—Primary Findings A1, A4, A6, A7 C8, C11, C14, C15, C16, C17, C21, C23, C25, C28, C29, C35, C51, C52, C69 D19, D25, D33 HA3, HA6, HB3, HB4, HB5, HB7, HB9 J3, J4, J16, J17, J18, J22, J27, J29, J33, J40, J44, J56, J62, J73, J77, J80, J84, J91,

13 Process risk management is the term given to collective efforts to manage process risks through a wide variety of strategies, techniques, procedures, policies, and systems th at can reduce the hazard of a process, the probability of an accident , or both. In general, the strategy for reducing process risk, whethe r directed toward reducing the frequency or the consequences of pote ntial accidents, can be classified into five categories: 1. Inherent - Minimizing and possibly eliminating the hazard by using materials and process conditions that are non-hazardous; i.e., substituting water for a flammable solvent; 2. Spatial - The use of distance to reduce the effects of hazards, such as siting the process as far as po ssible from occupied buildings and outside receptors. 3. Passive - Minimizing the hazard through process and equipment design features that reduce either the frequency or consequence of the hazard without the active functi oning of any device or requiring human input; i.e., providing a dike wall around a storage tank of flammable liquids; 4. Active - Using SCAI (safety controls , alarms, and interlocks), and mitigation systems to detect and respond to process deviation from normal operation; e.g., a pump whic h is shut off by a high-level switch in the downstream tank when the tank is 90% full. These systems are commonly referred to as engineering controls or safeguards. 5. Procedural - Using policies, operating procedures, training, administrative checks, emergency response, and other management approaches to prevent incidents, or to minimize the effects of an incident; i.e., hot work procedures and permits. These approaches are commonly referred to as admi nistrative controls and require humans that interface with the pr ocess (primarily operators) to recognize and react to abnormal or adverse events in a prompt and correct manner as directed by the approved procedures and the training on those procedures.

9. Human Factors in equipment design 113 9.9 Key learning points from this Chapter Key learning points include: • Poorly designed equipment can cause people to make mistakes, especially if they are fatigued, time pressured, stressed, or inexperienced. • It is important to be able to identify the signs of poor design and understand how they impact on performance. • Use Human Factors principles in the design of equipment. • It is important to include users of equipment when designing and testing new equipment. Users can also provide key information on learning from experience.

Application of Control Architectures 291 In some cases another pressure gauge is installed on the suction side of the centrifugal pump. This could be the case when there is a high chance of cavitation in the pump. If the pumping liquid is very hot and/or very volatile there is a high potential for a cavitation phe-nomenon, which is detrimental for the pump. In such cases, it is a good idea to put another pressure gauge in the suction of the centrifugal pump to be monitored by a field operator and make sure that there is enough pressure in the suction of pump to prevent cavitation. Sometimes the temperature and pressure gauges are installed on the suction side of the centrifugal pump instead. There are some cases that a third pressure gauge is placed around a centrifugal pump and that the pressure gauge is the one upstream of the strainer. If the pumping liquid is very dirty and there is a high chance of plugging the strainer it is a very good idea to see one pressure gauge upstream of the strainer and one other pressure gauge downstream of that. This provision helps the field operator to check these two pressures and make sure that the strainer is not plugged. The case for PD pumps is not much different from centrifugal pumps. The only difference could be elimina-tion of a pressure gauge on the suction side of the pumps as they are not sensitive toward cavitation. 14.9.3 Heat Ex changer Sensors We know that almost always the outlet temperature of the target stream should be equipped with a temperature control loop. In addition to this, all the streams around a heat exchanger are equipped with a temperature sensor and a pressure sensor. The reason for the temperature sensor is very obvious, as the main duty of a heat exchanger is changing the temperature. However, it could be questioned why there is a need for installing pressure gauges on a heat exchanger. The main reason for installing pressure gauges on the pipes around a heat exchanger is to identify if there is a leakage inside of the heat exchanger. As a heat exchanger consists of two enclosures, one for cold fluid, the other one for hot fluid, and are in contact with each other while they are not always visible, it is a good idea to monitor the pressure of the fluids to make sure no leakage or rup-ture happens. In the case of leakage or rupture the fluid migrates from the high pressure side to the low pressure side and the pressure raises the low pressure fluid. Even though it was mentioned that pressure gauges and temperature gauges are installed on the streams around the heat exchanger, there are some exceptions and deviations.For example, if the heat exchanger is a utility heat exchanger, it is not very common to see a temperature gauge on the utility stream. The reason is that the util-ity stream is intended for heat transfer. Its inlet tem-perature is fairly constant and the outlet temperature is  not very important as it goes to a “temperature com pensation system” such as a fired heater or an air cooler. When two or more heat exchangers are stacked up on each other possibly no pressure gauge or temperature gauge are installed on the space between the two stacked heat exchangers. The reason is lack of room for these instruments. If it is really necessary to monitor the tem-perature and/or pressure of each single heat exchanger in a stacked arrangement a thermowell instead of tem-perature gauge and a pressure point instead of pressure gauge could be installed. A thermowell and a pressure point (PP) is installed and operators can use them to check the temperature and pressure of the stream using their portable temperature gauge and portable pressure gauge. Sometimes, instead of installing a pressure gauge on the outlet side of a utility stream, only a PP is installed to save some money (Figure 14.35). 14.9.4 Fir ed Heater Sensors Because of the criticality of fired heaters, they are equipped with a BPCS, SIS and monitoring systems. Generally speaking all flows and all temperatures are monitored in a well‐equipped fired heater. On the process fluid side, the total flow and the flow of each pass are monitored. On firing system flows and the flows of fuel and air are monitored. On the flue gas side the pressure of the stack is monitored, which gives a sense of the flue gas flow rate too. The temperature of the outlet process fluid, on each pass, needs to be monitored. Vent DrainNo gauges in between, only inst. location Figure 14.35 Monit oring of stacked heat exchangers.

306 guidewords as “deviations,” using, for example, the Guideword Matrix provided in Table 12.1. Recommende d inherent safety guidewords and their descriptions are shown in Table 12.2 (Ref 12.7 Goraya). As previously mentioned, these guid ewords are simply the four most general and widely applicable prin ciples of inherent safety. The description for each is purposely br ief and is focused on materials, process routes, equipment and procedures. Use of these guidewords as “mind triggers” for each section of a process, as shown in Table 12.1, or during a particular process safety activity, such as management of change or incident investigation, as subsequently desc ribed, will help ensure that the concepts of inherent safety are visible within the process or activity. These guidewords are in tended as a supplement to existing tools that may already be in use within a specific process safety protocol. As with the use of HAZOP methods for conducting regular PHAs, it can be used at any level of subdivision of the process, including node-by-node. While thorough, this can al so increase the amount of time required to conduct the study. It al so requires a degree of brainstorming by a team that understands IS concepts well and can apply them to a process based on the P&ID and knowle dge of the operation. Appendix C provides an example of a combination HAZOP/checklist worksheet that includes review of potential IS opport unities applicable to the particular node and its hazardous materials. Table 12.1: Guideword Matrix (Contra Costa County, CA, Industri al Safety Ordinance Guidance Document, Attachment D, June 15, 2011) Minimization Substitute/ Eliminate Moderate Simplify Raw Material In-process Storage Product Inventory

106 PROCESS SAFETY FOR ENGINEERS: AN INTRODUCTION Figure 6.1. Overview map of the Bhopal vicinity (C&EN 1984) Detailed Description MIC is a flammable and highly toxic liquid. It is also water reactive, with the reaction being very exothermic. Water entered the MIC storage tank , and an exothermic reaction did occur. Investigators are not certain how the water got into the tank. Several theories have been developed about exactly what happened. One is that water entered the tank through a common vent line from a source over a hundre d meters away. Another is that water was deliberately introduced by a disgruntled employee. Other theories include the idea that water entered from the scrubber over time because th e tank was not pressurized or that the hose connections for water and nitrogen were confused with each other (Macleod 2014). Whatever the initial source of the water cont amination, several systems failures that could have mitigated the consequences of the event occu rred. See Figure 6.2. In the figure RVVH is relief valve vent header and PVH is process vent header. Pressure gauges and a high temperature alarm which could have warned of an exothermic reaction initiation. A refrigeration system that cooled the liq uid MIC was shut down and the refrigerant sold to save money. This could have remo ved heat from the reaction to prevent or reduce the amount of MIC that boiled up. The relief vents of the MIC tank were directed to a scrubber that could have absorbed a portion of the MIC; however, the vent gas scrubber was turned off.

Plant Process Control 305 where we don’t really need to control a flow. For example, a stream that goes to a large storage tank possibly doesn’t need to be flow controlled; an intermittent flow may not need to be flow controlled either. If the flow of a stream needs to be controlled then the second question is how?Flow in a pipe is a function of the pressure at both ends, P A and P B. In the schematic in Figure  15.17, if we increase the pressure at point A the flow (F A‐B) will increase; if we increase the pressure at point B the flow will decrease. This means the flow rate, F A‐B, doesn’t stay at the value specified in the mass balance table. The flow rate is actually a function of pressure at points A and B. Therefore, if you try to fix the flow rate in a pipe, you need two pressure loops at either end (Figure 15.18). Here what we are trying to say is if the intention is to control flow in a piece of pipe, two control valves are needed; one at each end of the pipe! Even though this is theoretically a requirement, in reality there are not many cases where we need to install two control valves to be able to control the flow rate. We only need to install two control valves if the pressures at both ends of the pipe are fluctuating. If somehow we can prove that in one specific case the pressure on one side of a pipe is fairly constant, we may be able to get rid of the control valve at that “fairly fixed pressure” point and use a single control valve arrangement. Table  15.7 shows the points where we can assume a fixed pressure, and points where the pressure is most likely fluctuating. Option A Option BPC PT PE PT PC PE Figure 15.16 Con trol of pressure in a pipe. PAFA-B FB Figure 15.17 Con trol of flow. PC PC Figure 15.18 “Fixing ” flow rate.Table 15.7 Fix ed pressure points and fluctuating pressure points. These points on pipes can be considered as “fairly fixed pressure” pointsThese points on pipes cannot be considered as “fairly fixed pressure” points ●Flow point near storage tanks ●Flow point near a vessel with a constant level (level controlled) ●Flow point near a vessel with a constant pressure (pressure controlled) ●Flow point on a network with multiple users ●Flow point near a container with limited surge capacity ●Flow point far from fixed pressure points

5 • Facility Shutdowns 93 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 & Perform. Assurance Management of Change Operational Readiness Conduct of Operations Emergency Management Incident Investigation Measurement and Metrics Auditing Management Review and Contain. Improv. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 23 1 4 2250 8 35 32 612072491 2 2000 1 21 1 1 121 1 0 2 211 2111 1 0 3 21 1 0 3 1 11 0 4 21 1 0 21 1 0 5 21 1 21121 1 0 6 21 1 0 31 1 1 0 7 11 0 3 1 11 0"Points to Remember" (IChemE BP Loss Prevention Series) CCPS Risk Based Process Safety Element BP Texas City Incident - CCPS RBPS Sum Identify hazards and safeguards listed in the procedure. Follow written procedures/c hecklists step by step. Make sure all valves/blinds /locks are in the proper position.3% Start-up/shutdowns are rare, so refresher training may be needed. Make sure all critical instrumentation/ equipment are functional and that certifications are current. Make sure that all work permits have been closed and the equipment is approved for use. Pillar IV Learn from Experience49% Talk through the procedures as a team before each shift. Pillar I Commit to Process Safety Pillar II Understand Haz. and Risks Pillar III Manage Risk35% 12% Table 5.3 Points to remember when startin g up after a facility shutdown time

A polymerization process conduc ted using a gradual addition batch process required a large qu antity of organic solvent to keep the system viscosity low en ough for effective mixing and heat transfer. In the event of an uncontrolled (runaway) polymerization, a large quantity of flammable and toxic material would be ejected through the re actor rupture disk. Instead of relying on an expensive and elaborate emergency relief discharge system to control th e potential hazard, the basic process chemistry was re-considered. It was found that it was possible to make the product usin g a suspension polymerization process in water. Because of th e significantly lower amount of solvent required in the new process, most of the material released in a runaway reaction would be water, with just a small amount of solvent and unreacted monomer. Reaction runaway was also less likely because of the higher heat capacity of water, a l l o w i n g i t t o a b s o r b m o r e o f t h e h e a t o f r e a c t i o n d u r i n g a process upset (KA). Dow had been using benzene as an azeotroping agent in one of its oxide derivatives units but replaced the benzene with a less hazardous chemical (OK). Dow replaced chlorine gas with so dium hypochlorite (bleach) as a water purification chemical at selected non-chlorine producing sites, and replaced Varsol in a process with another, less hazardous solvent (OK). 15.8.3 Moderate Use less hazardous conditions, a less hazardous form of a material, or facilities which minimize the impact of a release to hazardous material or energy. In the 1930s, ammonia plants typically operated at pressures up to 600 bar (9000 psia). Over the years, improved understanding of the chemistry has resulted in a downward trend in the operating pressure of ammonia plants; in the 1980s, plants operating in the range of 100-150 bar (1500 – 2250 psia) were common. The newer, low-pressure plants are inherently safer, cheaper, and more efficien t than their high-pressure counterparts (KA). 427

4.2 Can passive leak-limiting technology be used to limit potential loss of containment? • Blowout resistant gaskets (e.g., spiral wound) • Increasing wall strength of piping and equipment • Maximize use of all-welded pipe • Using fewer pipe seams and joints • Providing extra corrosion/erosion allowance (e.g., Sch. 80 vs. 40) • Reducing or eliminating vibration (e.g., through vibration dampening or equipment balancing) • Minimizing the use of open-ended (bleed or vent), quick-opening valves (for example, quarter-turn ball or plug valves) • Eliminating open-ended (bleed or vent), quick-opening valves (for example, quarter-turn ball or plug valves) in hazardous service • Using incompatible hose conne ctions to prevent mis-connection (e.g., air/ nitrogen, raw materials) • Use of round valve handles for op en-ended quarter-turn valves to minimize potential for bumping open • Improving valve seating reliability (e.g., using system pressure to seal valve seats where possible, using valve seat geometry, valve operations, and flow to eliminate or reduce seat damage) • Eliminating unnecessary expans ion joints, hoses, and rupture disks • Use of articulated arms instead of hoses for loading/unloading of hazardous materials • Eliminating unnecessary sight glasses/glass rotameters; use high- pressure/armored sight glasses as needed • Eliminate use of glass, plastic or other brittle material as material of construction • Use of seal-less pumps (e.g., canned, magnetic drive) • Use of top-unloading vessels/sto rage tanks; minimize number of bottom connections/ fittings • Minimizing the number of different gaskets, nuts, bolts, etc. used to reduce potential for error 4.3 Has attention to control system human factors been addressed through: 452

76 PROCESS SAFETY FOR ENGINEERS: AN INTRODUCTION segments of the oil and natural gas industry. Some of these recommended practices address process safety and fire protection. Of note are the following. (API) API 520 “Sizing, Selection, and Installa tion of Pressure-Relieving Devices” API 521 “Pressure-Relieving an d Depressuring Systems” API RP 752 “Management of Hazards Associ ated with Location of Process Plant Permanent Buildings” API RP 753 “Management of Hazards Associ ated with Location of Process Plant Portable Buildings” API RP 756 “Management of Hazards Associ ated with Location of Process Plant Tents” API RP 2001 “Fire Protection in Refineries” API RP 2003 “Protections Against Ignitions Ar ising Out of Static, Lightning and Stray Currents” CCPS Guidelines for Combustible Dust Hazard Analysis . This book describes how to conduct a Combustible Dust Hazard Analysis (CDHA) for processes handling combustible solids. The book explains how to do a dust hazard anal ysis by using either an approach based on compliance with existing consensus standards, or by using a risk-based approach. Worked examples in the book help the user understand how to do a combustible dust hazards analysis. (CCPS 2017) CCPS Guidelines for Consequence Analysis of Chemical Releases . This Guidelines book provides technical information on how to cond uct a consequence analysis to satisfy your company's needs and the U.S. EPA rules. It covers quantifying the size of a release, dispersion of vapor clouds to an endpoint concentration, outcomes for various types of explosions and fires, and the effect of the release on people and structures. (CCPS 1995) CCPS Guidelines for Evaluating Process Plant Buildings for External Explosions, Fires, and Toxic Releases, 2nd Editio n. Siting of permanent and tempor ary buildings in process areas requires careful consideration of potential e ffects of explosions and fires arising from accidental release of flammable materials. This book, which updates the 1996 edition, provides a single-source reference that ex plains the American Petroleum Institute (API) permanent (752) and temporary (753) building recommended prac tices and details how to implement them. New coverage on toxicity and updated standard s are also highlighted. Practical and easy-to- use, this reliable guide is a must-have for im plementing safe building practices. (CCPS 2012) CCPS Guidelines for Siting and Layout of Facilities , 2nd Edition . This book has been written to address many of the developments since the 1st Edition which have improved how companies survey and select new sites, evaluate acquisitions , or expand their existing facilities. This book updates the appendices containing both th e recommended separation distances and the checklists to help the teams obtain the informatio n they need when locating the facility within a community, when arranging the processes wi thin the facility, and when arranging the equipment within the process units. (CCPS 2018)

340 Human Factors Handbook relevant key lessons learned from other errors, incidents, and near misses. These lessons should be communicated to all individuals and teams that do similar work and/or are exposed to similar error traps. Efforts should be made to ensure that the lessons learned result in improvements, and that these improvements are fully applied. This means that behavioral and/or system changes must be evident. The effectiveness of any change should be assessed. 26.4 Learning in high performing teams Even high performing teams make errors . They are viewed as “high performing” because they learn from errors. That is, they identify root causes of error, seek out relevant lessons learned, and apply appropriate solutions. Errors and mistakes are “windows into reality” and offer a unique opportunity for learning and application of learning into the working environment. Lessons should be drawn from both negative events (accidents) and positive events (what went well). Self-directed learning behavior is an attribute of high performing teams, as shown in Table 26-1. .

3 • Normal Operations 30 Normal day-to-day operations in the chemical process industries can be continuous, batch, or a combination of the two. Process safety efforts are essential for every day-to-day operation, no matter what type of process it is. For a continuous process, an open, steady-state system, the process conditions during the normal operations time, such as the flow rates, temperatures , and pressures, are not changing over time; the process is at the normal operating conditions and within its standard operating limits. On the other hand, in a batch process, a closed, unsteady-state system, the no rmal operations time is when the process conditions, such as temperatures, pressures, and concentrations, are changing over time; the process is at the normal, but transient, operating conditions and within its standard operating limits. These operating limits are boun ded by safety limits, such that deviations from these limits (abno rmal operations) must be addressed before a loss event occurs. Whether continuous or batch, these processes have a normal start-up and a normal shut-down associated with them: either the process is running or not. These processes have normal start-up and shut-down procedures associated with them, as well [19]. This guideline will focus on continuous or batch operations, recognizing that a combination of these guides will apply to processes with have elements of both continuous and batch processes (e.g., a Continuous Stirred Tank Reactor (CSTR)). This chapter specifically addresses the normal operations mode with normal shut-downs using the procedures dedicated for normal shut-downs and normal start-ups afterwards—transient operating modes Type 1 and Type 2 listed in Table 1.1. Thus the scope of the transient operating modes in this chapter are shown in Figure 3.1. Chapters 4 and 5 cover the special, additional shutdown-related preparations for projects or maintenance-related procedures and

Principles of P&ID Development 67 (Continued)Case P&ID Showing the pump SIS or alarming system to protect the pump from an abnormal condition (such as seal leakage monitoring system) FC FT FT FEPM MCC I115 HS HSH/O/AS/SS/S command Run status Common trouble alarm L/R status stop Shutdown command FV FEFC FV FOFO M436 Maintenance and inspection Adding a pressure gauge in discharge or suction FC FT FT FEPGPM MCC I115 HS HSH/O/AS/SS/S command Run status Common trouble alarm L/R status stop Shutdown command FV FEFC FV FOFO M436 PG Adding a block valve in suction and discharge lines (such as a gate valve) to isolate the pump during the maintenance FC FT FT FEPGPM MCC I115 HS HSH/O/AS/SS/S command Run status Common trouble alarm L/R status stop Shutdown command FV FEFE FV FOFO M436 PGTable 5.11 (C ontinued)

TOXIC HAZARDS 115 Cameo Chemicals Database contains basic chemical descriptions, ID numbers, potential hazards, response recommendatio ns, physical properties, and regulatory information. It is available at https://cameochemicals.noaa.gov . (NOAA) Emergency Response Planning Guidelin e (ERPG) values can be found at: https://www.aiha.org/get-involved /aiha-guideline-foundation/erpgs Protective Action Criteria - A database co ntaining AEGL, ERPG, and TEEL values for more than 3000 chemicals can be found at: https://energy.gov/ehss/protective- action-criteria-pac-aegls-erpgs-t eels-rev-29-chemical s-concern-may-2016 . (PAC) BASF Medical Guideline provides medical ca re information for exposure to a large number of chemicals. It is intended for use by emergency responders and medical professionals and can be found at https://collaboration.basf.com/portal /basf/en/dt.jsp?setCursor=1_1032616. Methods to address toxic impacts in risk assessment are described in the following. Crowl and Louvar, Chemical Process Safe ty Fundamentals wi th Applications, 4th Edition . This book provides a compilation of many process safety topics. It links academic concepts to industrial process safety. (Crowl 2019) CCPS Guidelines for Chemical Process Quantitative Risk Analysis . This guideline describes the application of quantitative risk analysis in pr ocess safety, including toxic risk analysis. It describes the identification of incident scenarios, evaluation of the risk by defining probability of failure and the consequence impacts. In understanding the risk, then risk reduction strategies may be evaluated. (CCPS 1999) Summary Chemicals can pose toxic hazards in addition to the fire and reactive hazards described in previous chapters. Various organizations have established lists of toxic chemicals and exposure limits for those chemicals. This assi sts in understanding what level concentration may impact people and the environment. Other Incidents This chapter began with a detailed description of the Bhopal incident. Other incidents involving toxic chemicals include the following. ICMESA Dioxin Release, Seveso, Italy, 1976 Union Carbide MIC release, Bhopal, India, 1984 Marathon Oil Refinery HF Releas e, Texas City, Texas, US, 1987 Motiva Enterprises Sulfuric Acid Tank Fa ilure, Delaware City, Delaware, US, 2001 Georgia Pacific Hydrogen Sulfide, Pennington, Alabama, US, 2002 CITGO HF Release, Corpus Christi, Texas, US, 2009 DuPont Phosgene Release, Belle, West Virginia, US, 2010 Millard Refrigeration Ammonia Release, Theodore, Alabama, US, 2010 DPC Chlorine Release, Festus, Missouri, US, 2012

6.3 Improving the Process Safety Culture of an Organization |233 sensitivity what they can do to remove barriers to workflow between the two groups. This is particularly important for broad PSMS elements such as AI/MI and Safe Work Practices where m any functions need to coordinate, and the roles between groups differ widely. Finally, like any change, culture change can be a difficult transition for many. Increasing communication of all kinds can help people weather the transition. Regardless of their initial acceptance of the change, maintaining a steady stream of com munication and encouraging inter-silo comm unication at least m akes it clear why the change is needed and where it is going. Measurement/Metrics As mentioned above, the use of metrics to show progress of the culture improvement effort is important both to help leadership remain focused on the culture change and to show workers that progress is being m ade. General PSMS m etrics, as discussed in section 5.1, are also critical both for operation and improvement of the PSMS. Both kinds of metrics should be discussed in leadership m eetings, and comm unicated across the organization. It is important to focus m etrics collection on what is essential, and m ake the reporting, collation, and interpretation of metrics as easy as possible. Like documentation, above, metrics collection can be subject to normalization of deviance, especially if those reporting the metrics view it as a burden. Enhancing Com munication As discussed previously, upward communication regarding observed issues serves a critical role in enhancing both the process safety culture and the PSMS. Leaders can do 4 things to enhance upward com munication (Ref 6.12):

234 PROCESS SAFETY FOR ENGINEERS: AN INTRODUCTION CCPS a, Process Safety Beacon, “The Seal that Didn’t Perform”, July 2002, http://sache.org/beacon/files/ 2002/07/en/read/2002-07%20Beacon-s.pdf . CCPS b, Process Safety Beacon, “It’s a Bird , It’s a Plane, It’s a Pump”, October 2002, http://sache.org/beacon/files/ 2002/10/en/read/2002-10%20Beacon-s.pdf . CCPS c, CCPS Process Safety Beacon, “I nterlocked for a Reason”, June 2003, http://sache.org/beacon/files/ 2003/06/en/read/2003-06%20Beacon-s.pdf . CCPS d, CCPS Process Safety Beacon, “Avoid Im proper Fuel to Air Mixtures”, January 2004, http://sache.org/beacon/files/ 2004/01/en/read/2004-01%20Beacon-s.pdf . CCPS e, CCPS 2007, Process Safety Beacon, “The Great Boston Molasses Flood of 1919”, May 2007, http://sache.org/beacon/files/2007/05/en/read/2007-05-Beacon-s.pdf . CCPS f, CCPS Process Safety Beacon, “What if You Load the Wrong Material Into a Tank?”, April 2012, http://sache.org/beacon/files/ 2012/04/en/read/2012-04-Beacon-s.pdf . CCPS g, CCPS Process Safety Beacon, “Vacuum is a Powerful Force!”, February 2002, http://sache.org/beacon/files/ 2002/02/en/read/2002-02-Beacon-s.pdf . CCPS h, CCPS 2011, Process Safety Beacon, “Understand the Reactivity of Your Heat Transfer Fluid”, February 2011, http://sache.org/beacon/files/2011 /02/en/read/2011-02-Beacon-s.pdf . CCPS Glossary, “CCPS Process Safety Glossary”, Center for Chemical Process Safety, https://www.aiche.org/ccps/resources/glossary . CCPS 1995, Guidelines for Safe Oper ations and Maintenance , Center for Chemical Process Safety, John Wiley & Sons, Hoboken, N.J. CCPS 1998, Guidelines for Improving Plant Reliabilit y through Data Collection and Analysis , Center for Chemical Process Safety , John Wiley & Sons, Hoboken, N.J. CCPS 2005, Guidelines for Safe Handling of Powders and Bulks Solids , Center for Chemical Process Safety, John Wiley & Sons, Hoboken, N.J. CCPS 2006, Guidelines for Mechanic al Integrity Systems , Center for Chemical Process Safety, John Wiley & Sons, Hoboken, N.J. CCPS 2012, Guidelines for Engineerin g Design for Process Safety 2nd Edition , Center for Chemical Process Safety, John Wiley & Sons, Hoboken, N.J. CCPS 2016, Guidelines for Asset Integrity Management , Center for Chemical Process Safety, John Wiley & Sons, Hoboken, N.J. CCPS 2020, Process Safety Beacon, “Not all vi brations in process equipment are ‘good vibrations’”, November, www.aiche. org/ccps/process-safety-beacon. Colfax Fluid Handling, https://www.colfaxcorp.com/ CSB, “Static Sparks Explosion in Kansas”, Case Study No. 2007-06-I-KS, U.S. Chemical Safety and Hazard Investigation Board, Washington, D.C. CSB 2009, “Investigation Report T2 Laboratories , Inc. Runaway Reaction”, Report No. 2008-3-I- FL, U.S. Chemical Safety and Hazard Investigation Board, Washington, D.C.

86 PROCESS SAFETY IN UPSTREAM OIL & GAS BSEE (2013) has published its policy covering nine characteristics contributing towards a positive safety culture, based on similar guidance for the US nuclear industry. The nine charact eristics are as follows. 1.Leadership commitment to safety values and actions 2.Hazard identification and risk management 3.Personal accountability for process and occupational safety 4.Work processes address safety and environment 5.Continuous improvement for safety and environment 6.Positive environment for raising concerns 7.Effective safety and envi ronmental communication 8.Respectful work environment 9.Inquiring attitude avoiding complacency COS (2018) has issued guidance on offshore safety culture and has a SEMS maturity self-assessment tool. The COS safety culture good practice covers six of the characteristics above, as the other thre e topics (hazard identification and risk management, work processes, and continual improvement) are covered sufficiently through API RP 75.

488 PROCESS SAFETY FOR ENGINEERS: AN INTRODUCTION Table E.1. Tier 1 Level and Tier 2 Level Consequences (CCPS 2018) Consequence Tier 1 Tier 2 Injury OSHA “days away” Hospitalization Fatality OSHA Recordable Fire/Explosion damage > $100,000 USD > $2,500 and < $100,000 USD Community impact Community evacuation or shelter-in-place An engineered pressure relief or upset emission from a permitted or regulated source Discharge amount, time and rainout Unsafe location Onsite shelter-in-place Public protective measures Discharge amount, time and rainout Unsafe location Onsite shelter-in-place Public protective measures Acute release Above materi al threshold quantity E.2.1 Release Criterion Release from process. Tier 1 and 2 incidents involve an LOPC from process. Process refers to equipment, storage tanks, active warehouses , ancillary support areas, on-site remediation facilities, and distribution piping under control of the company used in the manufacture of petrochemical and petroleum refining products. Unplanned and Uncontrolled. Intent of the release is a criterion in Tier determination. PSEs must be unplanned or uncontrolled. It is possible to plan a safe release from primary containment. An example is a safe de-invento ry of process in preparation for equipment maintenance. As a planned activity, this type of LOPC is not considered a Tier 1 or 2 indicator. E.2.2 Outcome Criterion If the discharge contains one of the four conseq uences described here, the tier of the release is determined by the release quan tify described in Section E.2.3. Injury. An LOPC of any material in any amount that results in an injury requiring treatment beyond first aid is considered for Tier 1 or 2 cl assification. If the LOPC results in an injury classification of OSHA recordable, then the PSE cl assification is Tier 2. Tier 1 classification results from more serious injury such as an OSHA “days away” case, or hospital admission and/or fatality. OSHA Recordable Injury or Illness - Any work-related injury or illness requiring medical treatment beyond first aid. (adapted from OSHA) Days Away from Work - An OSHA recordable injury or illness resulting in one or more days away from work, rest ricted work, or transfer to another job. (adapted from OSHA)

360 PROCESS SAFETY FOR ENGINEERS: AN INTRODUCTION Each of these areas relates to human performance in specific ways. People: Human anthropometry (size and shape) impa cts a person’s ability to do different types of work Humans have physical needs of food, water, air, and rest The 5 human senses provide informatio n, but the senses can be degraded The human brain processes information, but it too has limitations Humans have environmental limits, e.g. temper ature, noise, light as well as boring or stressful situations Facilities & Equipment: This is the human ma chine interface (HMI), i.e. matching the human and the hardware in the workplace. Management Systems: Operating Procedures affect performance; they should be accessible and clear Work handover communication impacts th e workers understanding of the process state Training should focus on correct operations and process safety risks Work motivations should be consistent with process safety goals Putting all of this together, a workplace ca n be designed to support successful human performance. For example, An operator can follow a procedure, identi fy the correctly labeled valve, and reach the handle to easily turn it. An operator can safely complete a job begun on a previous shift supported by a clear shift handover, documented work process, and management support in providing adequate time to complete the job. The control panel design is such that it draws the operator’s attention to a process parameter that is exceeding normal operatin g limits in time for the operator to take corrective action based on training. Human Factors is not identified as a stand-alone element in the management systems identified in Chapter 2 or the regulations identified in Chapter 3. Instead, it can be seen as a part of these Risk Based Proce ss Safety elements. (Bridges 2010) Process Safety Culture Workforce Involvement Operating Procedures Training and Performance Assurance Operating Readiness Conduct of Operations Although the term ‘human factors’ may not be commonly used in guidance, it is frequently cited as a cause of process safety incidents. Companies and organizations such as CCPS have

Pressure Relief Devices 233 12.16.2 Gas/Vapor Disposal T he first choice for relieving gas/vapor is disposing of it in the atmosphere because it is the least expensive choice! These days this choice is allowable only if the reliving gas/vapor is innocent. Disposing to atmosphere is generally regulated in dif - ferent jurisdictions. The regulatory bodies allow release to atmosphere if the release doesn’t jeopardize the health and safety of people and doesn’t generate pollution in the environment. The famous examples of releasing to atmosphere are releasing air or steam to the environment. Wherever it is allowed to release to atmosphere a note is place on the P&ID as “to atmosphere. ” The other note, which could be confusing, is “to safe location. ” Although a P&ID developer may not care about the meaning of this note, he needs to know this as there are some cases that finding a “safe location” is not easy! Each company may have their own interpretation of “safe location for PRD release” and generally their inter - pretation is outlined somewhere in their guidelines. However, where there is no guideline the “safe location” can be assumed as: vertical and upward release so that the release point is higher than a minimum 3 m fr om the ground (or platform) and a minimum 2 m fr om the top of all equipment with a radius of 7.5 m around t he release point(Figure 12.30). The outlet of the PRD to atmosphere may need to be equipped with a “bird screen” to prevent bugs and larger animals from entering. Bird screens are tagged as an SP item on P&IDs.Table 12.12 Ultimat e destination. Liquid Gas/vapor ●System relieving ●Open drain ●Closed drain ●Small vessel on the ground ●Ground ●Atmosphere (if innocent) ●Flare (if combustible) ●Emergency scrubber or quench pool (if contains absorbable and non‐innocents) ●VRU (if recoverable) ●System relieving (a) (d) (b)(c) PSV Or PSVFigure 12.28 Differ ent liquid disposal systems. Pipeline Figure 12.29 Sy stem relieving for thermal expansion PSVs on a pipeline.

68 | 5 Learning Models 5.10 Peters, T. and Waterman, Jr., R.H. (2004). In Search of Excellence: Lessons from America's Best-Run Companies. New York: Harper Business. 5.11 Willyerd, K., Grünwald, A, Brown, K. et al. (2016). A new model for corporate learning. D!gitalist Magazine (11 March 2016).

84 Guidelines for Revalidating a Process Hazard Analysis These examples illustrate how recent operating experience affects the revalidation approach and preparation. Example 1 – Significant Incidents: Background. A refinery crude unit is due for PHA revalidation. A few changes have been implemented since the prior PHA, an d the PHA includes LOPA for selected higher consequence scenarios. However, be cause of a fire associated with the unit heater, a large project has been comp leted to install a burner management system (BMS), convert from naphtha fuel to natural gas, and convert the heater from forced draft to natural draft. Review of Operating Experience. As part of this review, the process safety coordinator for the facility decides that most of the unit PHA can be Updated . Few changes or incidents have occurred in the unit feed and fractionation sections, and unit mechanical integrity repo rts show no issues for this portion of the process. However, a decision is made to Redo all sections of the PHA associated with the heater, because ma ny complex changes have occurred that were not analyzed by the prior PHA team. In addition, the existing LOPA scenarios associated with the heater should be Redone to properly account for the IPLs added or affected by the BMS pr oject. This information is given to the current revalidation leader and team prior to the meeting. Example 2 – Organizational Changes: Background. Company A has been purchased by Company B and a significant turnover has occurred in the operations, engineering, and management departments. No personnel who were involved in the prior PHA are available. Review of Operating Experience. This situation is challenging. A significant amount of operating experience has been lost. Company B should probably Redo the PHA for several reasons, including consistency with corporate risk tolerances/procedures and development of process knowledge for newer staff. However, in cases such as this, the prior PHA should not be disregarded and should be used during the Redo to help the revalidation team understand the original design and current intentions of the process.

384 Human Factors Handbook the tower bottom section. At an ea rly morning shift meeting, personnel decided that tower start-up could not begin because the storage tanks were thought to be full. They did not share this decision with the ISOM operations team. • A miscommunication between operators meant that the level control valve was closed and light raffinate wa s directed into the heavy raffinate line. The level control valve was manually maintained at closed to keep the level high, to protect downstream equipment. Experience was that if the level was set at 50%, the level could drop below 50% and shut down the process. The control valve should have been set to automatic and at 50% to establish outflow to storage. • The feed pump was started. A defective indicator incorrectly showed raffinate leaving the tower through the closed level control valve. Burners were then started to heat the raffinate entering the tower. This increased the liquid level and the pressure in the tower. • Operators thought that the pressure rise was due to overheating in the tower bottoms, compressing nitrogen in the tower, as this was a known issue. Therefore, they opened a valve to relieve the pressure to a blowdown drum. • The tower outlet valve was then open ed. However, heat from outflowing raffinate was being transferred to incoming raffinate by a heat exchanger. This caused the liquid in the tower to expand and the level to rise until it entered the overhead vapor line and flowed into a pressure relief system. The relief valves opened and released raffinate to a blowdown drum and stack. From there it vented to the atmosphere. • The operator noticed the pressure spike. He fully opened the tower level control valve and turned off the furnaces but did not stop the raffinate feed into the tower. However, this wa s too late. The raffinate overflowed the stack and the vapor was likely ignited by a nearby truck engine.

5 • Facility Shutdowns 94 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 & Perform. Assurance Management of Change Operational Readiness Conduct of Operations Emergency Management Incident Investigation Measurement and Metrics Auditing Management Review and Contain. Improv. 12345 67 89 1 0 1 1 1 2 1 3 1 4 1 5 1 6 1 7 1 8 1 9 2 0 23 1 4 2250 8 35 32 612072491 2 2000 8 211 0 31 1 1 0 9 21 1 0 3 1 1 1 0 10 11 1 13 111 0 11 11 0 1 1 11 12 11 0 2 1 1 11 13 21 1 0 2 1 1 0 14 11 21121 10"Points to Remember" (IChemE BP Loss Prevention Series) CCPS Risk Based Process Safety Element BP Texas City Incident - CCPS RBPS Sum Report any deviations/anomaly. Keep an open eye for "weak signals." Do not be satisfied by sh ort term fixes ("symptom treatment") but identify and treat the root causes of incidents. Ask questions and get help with operations which you are not familiar with. If unsure, or if multiple parameters seem to deviate from the usual ones indicate d in the procedure, shut- down the process and al ert potentially exposed personnel.Never go to the next step of the procedure before all conditions from previous steps aren't fully completed with satisfactory results. Check equipment setup thoroughly and monitor initial conditions in the field, maintain excellent communication with control room. Use Management of Change reviews before modifying any start-up procedures.3%Pillar IV Learn from Experience49%Pillar I Commit to Process Safety Pillar II Understand Haz. and Risks Pillar III Manage Risk35% 12% Table 5.3 Points to remember when starting up after a facility shutdown time (Continued)

INTRODUCTION 7 twins” and E-Learning, and some computer-based trai ning materials developed in conjunct ion with this book. The primary target populations for this information are front line operators, operations managers, plant engineers, process safety engineers, technical experts, and others who co uld benefit from it. This chapter provides guidance to organizations in the process industries about organizing and structuring training on management of abnormal situations, including advice on trainer comp etencies, training programs, and assessment of required skills. Chapter 5 – Tools and Methods for Managing Abnormal Situations Chapter 5 provides a brief summary of the relevant management tools that are currently available for recognition and management of abnormal situations; specifically, tools that illustrate the principles described in Chapter 3, and some worked examples of how to use them. Hazard Identification and Risk Analysis is also addressed, including a list of scenarios that Process Safety Engineers can use during HIRAs to address abnormal situations. Chapter 6 – Continuous Improvement for Managing Abnormal Situations Chapter 6 includes detailed descriptions of the available metrics related to managing abnormal situations that can be used for improvement, as well as the impact on incident invest igations related to complex system issues, including lower-level incident s such as near-misses. The chapter also provides guidance on identify ing the system and conditions that caused the incident and how to fix th e system in order to prevent future incidents. Chapter 7 – Case Studies/Lessons Learned Throughout the book, brief example incidents are used to illustrate specific issues in managing abnormal situations. Chapter 7 focuses on three case studies in more detail to illustrate to pics discussed in earlier chapters. The case studies address general as well as specific information regarding abnormal situations that have occurred both within and outside the process industry and how these events were handled from the perspective of management of abnormal situations. Both positive and negative examples are included in concise summaries.

9 • Other Transition Time Considerations 172 system and then introducing th e Nitrogen, pressurizing and depressurizing or sweeping with Nitrogen. Simultaneous Operations (SIMOPS), if applicable, should be addressed during the steps of the commissioning and initial start-up efforts. Note that the plans for safe and efficient operations and maintenance once the equipment and proc esses have been commissioned should address the facility’s ongoing Environmental, Occupational Health and Saf ety (EHS), and process safety management system, as well. Th ese management systems should include policies and procedures based on the RBPS elements, such as [14]: Process knowledge and technologies, including new or updated equipment files, Piping and Instrumentation Diagrams (P&IDs), Process Flow Diagrams (PFDs), etc. Operating procedures, includi ng normal start-ups and shut- downs; Safe work practices for maintenance-related and specific, non-routine activities; An asset integrity and reliability program, such as an Inspection, Testing, and Prev entive Maintenance (ITPM) program for the new equipmen t (including purchasing and warehousing support personnel); A contractor management program to ensure that qualified contractors execute their activities safely and competently; Training programs to ensure performance goals are achieved (including safely operating an d maintaining the equipment); Managing change procedures, to ensure that the hazards and risks associated with changes to raw materials, equipment, processes, and personnel are re viewed and approved before being implemented; Emergency management systems to ensure safe emergency response and to help reduce the consequences of loss events, including internal emer gency response teams and

110 Guidelines for Revalidating a Process Hazard Analysis brainstorming during the revalidation. Si milarly, some organizations limit the number of previous PHA participants on the revalidation team. While it is very beneficial to have some previous particip ants on the team, there is also a huge benefit to having some new team members who are more likely to raise new issues or challenge previous judgments. When preparing for the revalidation, the study leader should solicit appropriate individuals to fulfill the skill and experience requirements, and they must commit their time to participate. Any identified deficiencies or other concer ns regarding the staffing of the team should be resolved with the responsible member(s) of management. Another consideration is selection of the PHA revalidation leader. The leader drives the study to ensure competent application of the analysis method in conformance with the revalidation appr oach, and compliance with regulations. If the previous PHA was found to be defici ent in those aspects, site management should select a different PHA facilitator with a fresh perspective and skills that are more appropriate. Once the team composition is establis hed, the study leader should assess what training those individuals need for effective team participation and determine the most effective method to provide that training, as discussed in Section 7.2.2. Given the other constraints (e.g., normal work locations, calendar deadlines, team meeting hours, project budget ), it may be difficult to identify the most efficient way to provide the necessary training. The study leader should first decide what training requirements are specific to the individuals versus those that address general needs of the team. For example, everyone needs a general understanding of process operations and the facility layout. If most team members already have at least a basic familiarity with the process and equipment, then the individuals who need a tour and/or process review can schedule it at their convenience before the revalidation begins. On the other hand, if several participants need such training, then it will be more efficient to begin the first meet ing with a review of the process and a brief tour of the facility. The revalidatio n meeting dynamics are further discussed in Chapter 7. 6.1.3 Estimating Schedule, Time, and Resources During review of the prior PHA and rece nt operating experience, the revalidation leader should develop good insight into the resources that will be required for the revalidation. If the Redo approach is chosen, the revalidation schedule should approximate the schedule of the in itial PHA for a new unit of similar size and complexity. Note that this may be significantly longer than the original PHA

F.2 Culture Assessment Protocol |353 m anagem ent? Is critical, safety-related news that circumvents official channels welcomed? 75. Are com munications altered, with the message softened, as they m ove up the m anagement chain? Is there a “bad news filter” along the comm unications chain? 76. Do m anagement messages on the importance of safety get altered as they m ove down the m anagement chain? Do m anagem ent ideals get reinterpreted in the context of day- to-day production and schedule realities? 77. Are those bearing negative safety-related news required to “prove it is unsafe?” 78. Has any “intimidation” factor in communications been elim inated? Can anyone speak freely, to anyone else, about their honest safety concerns, without fear of career reprisals? 79. Do mechanism s exist that effectively promote and facilitate two-way communication between managers and all relevant stakeholders? 80. Is there a process to review the effectiveness of safety com mittees in promoting process safety and as a means to develop and execute a plan to improve such effectiveness? 81. Is the internal sharing of inform ation that will reduce safety risks occur without fear of punishment? 82. Is there a strong em phasis on prom ptly recognizing and reporting nonstandard conditions to perm it the timely detection of “weak signals” that m ight foretell safety issues? This issue is closely related to the normalization of deviance. 83. In general, do personnel not bother to report minor process- related incidents, accidents, or near misses? 84. Do several channels exist to communicate process safety critical inform ation and to ensure that expertise can be accessed in a timely m anner, especially in emergencies? 85. Does the organization interact with outside stakeholders regarding their hazards/risks? 86. B ecause of the legal concept of co-employment have host organizations consciously separated contractors, even

178 | 5 Aligning Culture with PSMS Elements reporting near m isses will be welcomed and acted upon, and should ensure open and frank communications to remove barriers to reporting. Certainly, investigations of incidents and near m isses consume time and resources, and may delay the restart of operations. However, if the investigation finds root causes in the m anagement system and culture that can be corrected, future incidents – with all accom panying costs and delays – can be avoided. Ultimately, investigating and then following up on findings and recommendations will strengthen the management system and promote learning to advance the culture. As of this writing, there rem ain com panies, industry sectors, and countries that view incident investigation in ways that are negative to process safety culture. One harm ful view is that an incident investigation should be only cursory or else it will identify shortcomings that could be targeted by regulators and attorneys. This certainly can happen. However, regulators and opposing attorneys can also conduct their own investigations and draw their own conclusions. In other words, there is no real benefit to this approach, and real opportunities to advance the culture and the m anagement system are lost. Another harmful view is of incident investigation as a tool to assess blame on the operator or mechanic whose error caused the incident. B y doing so, management m otivates workers to hide near-m isses and cover-up incidents. Comm unication of bad news up the chain of comm and becom es stifled and opportunities to prevent future incidents are lost. Finding blam e should never be the objective of an incident investigation. That is not to say that a proper investigation will never find that a worker or manager acted counter to the com pany’s perform ance policy or broke a law. This is discovered from time to time, and when it does, the imperative for process safety requires that the individual receive the appropriate discipline.

2 Human performance and error 2.1 Learning objectives of this Chapter Understanding human performance is import ant to support people to successfully complete tasks. Understanding human perf ormance will also help people reduce the likelihood of errors and mistakes. Learni ng from errors and mistakes is part of the journey to high performance. This chapter provides some Human Factors principles that will help people to reduce the likelihood of errors and mistakes . By the end of this chapter, the reader should be able to: • Understand how to proactively and methodically support human performance, • Understand the many factors a ffecting human performance, and • Understand solutions used to he lp improve human performance. These principles will be addressed again in later Chapters. 2.2 An example of success ful human performance 2.2.1 What happened? The “Miracle on the Hudson” happened on January 15, 2009, when a bird strike occurred shortly after US Airways flight 1549 took off from New York’s LaGuardia airport [15]. The Airbus struck a flock of Canada geese while on the climb from the airport. The Captain, Chesley Sullenberger, and First Officer Jeff Skiles decided to ditch (emergency water landing) the aircra ft in the Hudson River, saving all on board. This famous event was portrayed in a 2016 film ( Sully ) starring Tom Hanks as Chesley “Sully” Sullenberger. The successful ditching of an unpowered passenger airline onto the Hudson River, within six minutes of the bird strike, when both engines had failed, is an example of skilled and knowledgeable human performance. Following the bird strike, a very short peri od of time was available for the pilots to determine what had happened, enac t a Mayday, determine they could not return to the airport, decide they had to glide around and find an alternative landing site (the Hudson River), and identify a new course. They achieved this and ditched on the Hudson River, after whic h the 150 passengers and five crew were rescued by nearby boats and ferries. The normal procedure for dual engine failure was to attempt to return to the airport. This turn back to the airport was not possible at the plane’s low altitude. It was also not possible to complete a “dua l engine failure” checklist due to the limited time available prior to ditching. Si mulator training did not cover ditching. Human Factors Handbook For Process Plant Operations: Improving Process Safety and System Performance CCPS. © 2022 CCPS. Published 2022 The American Institute of Chemical Engineers.

Piping and Instrumentation Diagram Development 382 18.2.4 Chemical Hazards C hemical hazards can be reduced by placing, again, some sort of guard on the areas where there is chance of exposure. For example for aggressive material, “splash guards” could be placed on flanges of piping of aggressive liquids. “Splash guards” are generally stated in a note on the P&ID.If there are several items that may leak or splash aggres - sive liquids, they can put in an area enclosed by a pony wall and a specific curtain. This is again is captured by a note on the P&ID. 18.2.5 Energy Hazar ds Two types of energy hazards that may have a significant appearance on P&IDs are noise hazards and burning hazards caused by contact with hot surfaces. The strategies to put barriers for these two hazards are discussed here. 18.2.5.1 Noise Barrier If the sound of a certain piece of equipment is higher than the limit set by the local regulatory body, action should be taken to meet the requirements of the local occupational health authority. First of all it needs to be checked whether a certain noise is harmful or not. Generally speaking if you are in a noisy space and you need to shout to talk to a person in your vicinity, the noise is harmful for your ears. Standards generally call for workplaces with less than 85 dBA noi se as a safe workplace. This 85 dBA i s measured over eight hours and taking the average reading (TWA: time weighted average). The noise control can be done by isolating the noisy equipment. Isolating the noise source can be done by sound proof insulation or acoustical insulation. For noisy equipment, the acoustical insulation can be placed by constructing a sound proof shelter or cabinet. This can be captured by a note in the P&ID. If it is intended to minimize the noisy flow, the pipe may need to be acoustically insulated. In Figure 18.1, an example of a sound proof symbol on the P&ID is shown. “ AI” stands for “acoustical insulation” and 10 dB ref ers to the value of decrease in sound level. 18.2.5.2 Burning Pr evention The way we prevent burning of personnel if they become in contact with hot surfaces is thermal insulation. However this insulation is only for the purpose of burning prevention therefore it is named: “personal protection insulation. ” In a personal protection (PP) arrangement, a thin layer of insulation around pipes, equipment, and containers is installed. The purpose of “PP insulation” is only to prevent burning the operator when their hands or body come into contact with the hot equipment for as short as one second. A metallic wall with a temperature above 60–70 °C is generally considered a burning surface. Different com panies have different procedures for using insula- tion for PP . Some of them mention in their process guidelines that insulation must be used if the tempera-ture of the pipe or equipment is higher than 60  °C, while some other companies use 65 °C or even 70 °C as the criterion. However, not all high‐temperature equipment and/or pipes need PP insulation. If they are in remote areas or in inaccessible locations, PP insulation can be avoided if the company’s guidelines allow it. Personal protection insulation for pipes and equipment is needed if the temperature is a burning temperature and if operators could be in the vicinity. Table 18.2 shows P&ID presentation of PP insulation for pipes, equipment and instruments. The insulation thickness for PP is about 1–2 in.Table 18.1 Hazard and injur y. Hazard Injury Before accident After accident Preventing the injury Mitigating the injuries’ impacts The hazards can be controlled through three different strategies:1) Reduc ing or eliminating the hazard through engineering passively. 2) Reduc ing the hazard actively by putting barriers between workers and hazards. The barrier could be on the equipment side or operator side. The operator‐side barrier is a type of “personal protection equipment” (PPE). 3) Reduc ing the hazard through establishing rules for work practices and implementing them.The injury consequences can be mitigated through different methods including: ●Providing first aid ●Providing safety showers ●Providing eye washers.10 dB Al Acoustic insulation Figure 18.1 Soundpr oof insulation.

Conducting PHA Revalidation Meetings 139 Conversely, it is equally important to document which core analysis items the supplemental risk assessments stem from. Using both Redo and Update approaches is common for supplemental risk assessments, and they typically follo w the approaches of the core study revalidation. Risk assessments of the Update portions are Updated , and risk assessments of the Redo portions are Redone . Alternatively, a conservative approach would be to Redo all of the supplemental risk assessments, using the previous assessments as a guide where practical. 7.2 FACILITATING EFFECTIVE REVALIDATION MEETINGS The skills that trained PHA facilitators acqu ire on the job are as applicable to the efficient conduct of a PHA revalidation as they are to the conduct of an initial PHA. PHA team facilitation is disc ussed in detail in the CCPS book Guidelines for Hazard Evaluation Procedures [2]. The leader, however, should consider some unique aspects of a revalidation session. Most fundamentally, the revalidation team has the previous PHA documentation to use in a manner consistent with the Update or Redo approach. At the very least, “holes” in the previous PHA documentation show where extra effort needs to be applied in the current revalidation. In addition, the revalidation team may have the benefit of several years of actual operational and maintenance experience in lieu of the original design team’s optimism. Attention to items discussed in the following sections will help the facilitator keep the sessions as productive as possible. 7.2.1 Team Composition Core members of the revalidation team, meeting all the regulatory and company policy requirements, should be selected prior to the study sessions, as was Inexperienced Team? Team makeup is often a management decision outside of the revalidation team scope. The revalidation leader, however, should make responsible managers aware of a particularly inexperienced team. Such a team could lead to a deficient revalidation or an artificially high number of recommendations involving the conduct of additional studies or evaluations on issues that should, more properly, be resolved by the revalidation team during the meetings. An experienced team is critical to the success of any revalidation.

GLOSSARY xxxi Failure Modes and Effects Analysis (FMEA) A hazard identification technique in which all known failure modes of components or features of a system are considered in turn, and undesired outcomes are noted. Fireball The atmospheric burning of a fuel-air cloud in which the energy is in the form of radiant and convective heat. The inner core of the fuel release consists of almost pure fuel whereas the outer layer in which ignition first occurs is a flammable fuel-air mixtur e. As buoyancy forces of the hot gases begin to dominate, the burning cloud rises and becomes more spherical in shape. Fitness for Service (FFS) A systematic approach for evaluating the current condition of a piece of equipment in order to determine if the equipment item is capable of operating at defined operating conditio ns (e.g., temperature, pressure). Flammability limits The range of gas or vapor amounts in air that will burn or explode if a flame or other ignition source is present. Importance: The range represents a gas or vapor mixture with air that may ignite or explode. Generally, the wider the range the greater the fire potential. See also Lower Explosive Limit / Lower Flammable Limit and Upper Explosive Limit / Upper Flammable Limit. Flammable liquids “An ignitable liquid that is classified as a Class I liquid. A Class I liquid is a liquid that has a closed-cup flash point below 100 °F (37.8 °C), as determined by the test procedures described in NFPA 30 and a Reid vapor pressure not exceeding 40 psia (2068.6 mm Hg) at 100 °F (37.8 °C), as determined by ASTM D323, Standard Me thod of Test for Vapor Pressure of Petroleum Products (Reid Method). Class IA liquids include those liquids that have flash points below 73 °F (22.8 °C) and boiling points below 100 °F (37.8 °C). Class IB liquids include those liquids that have flash points below 73 °F (22.8 °C) and boiling points at or above 100 °F (37.8 °C). Class IC liquids shall include those liquids that have flash points at or above 73 °F (22.8 °C), but below 100 °F (37.8 °C).” (adapted from NFPA 30). Flash fire A fire that spreads by means of a flame front rapidly through a diffuse fuel, such as a dust, gas, or the vapors of an ignitable liquid, without the production of damaging pressure. Flash point temperature The minimum temperature at which a liquid gives off sufficient vapor to form an ignitable mixture with air within the test vessel used (Methods: ASTM 502). The flash point is less than the fire point at which the liquid evolves vapor at a sufficient rate for indefinite burning. Frequency Number of occurrences of an event per unit time (e.g., 1 event in 1000 yr = 1 x 10-3 events/yr).

306 PROCESS SAFETY FOR ENGINEERS: AN INTRODUCTION Estimate the maximum distance for a6.9 kPa (1 psi) blast overpressure isobar from leaking a medium fuel reactivity vapor into a medium-congestion 2930 m3 (3832 yd3) enclosed process structure. Assume the conc entration within this confined space could exceed the lower flammable limit. Show your results. Estimate the distance to 6.9 kPa (1 psi) blast overpressure from a 30 kg/s (66 lb/s) outdoor isopropyl amine (molecular weight 59.1, medium reactivity fuel) airborne rate into a medium congestion process area. The distance to the lower flammable limit of 2 volume % is estimated as 140 m (459 ft) at a wind speed of 3 m/sec (9.8 ft/s) and averaging time of 18.75 sec. Show your results. Describe the effect on dispersion and downwind concentration by changing a) release height, b) wind speed), c) distance downwind , d) distance in the horizontal direction from centerline of plume, e) distance in the vertical direction from centerline of plume, and f) stability class. For each parameter given in exercise 16, which direction of change leads to the most conservative estimate of downwind concentration from a release. Why? References API STD 521, “Pressure-Relieving and Depressuri ng Systems”, American Petroleum Institute, Washington, D.C . ARCHIE, “Automated Resource for Chemical Hazard Incident Evaluation”, FEMA https://apps.usfa.fema.gov/thesaurus/main/termDetail . ASTM E1529, “Standard Test Methods for Dete rmining effects of Large Hydrocarbon Pool Fires on Structural Members and Assemblies”, ASTM International, https://www.astm.org/Standards/E1529.htm. Autoreagas, TNO, www.tno.nl. Baker 1973, Baker, W. E., Explosions in Air , University of Texas Press, Austin. Baker 2002, “Explosion Risk and Structural Damage Assessment Code (ERASDAC)”, 30th DOD Explosive Safety Seminar, U.S. Department of Defense Explosive Safety Board, Arlington, VA. BakerRisk 2005, A. Pierorazio et.al., “An Up date to the Baker-Strehlow-Tang Vapor Cloud Explosion Prediction Methodology Flame Speed Table”, Process Safety Progress, Volume 24, No. 1, American Institute of Chemical Engineers, New York, N.Y., March. CCPS 1987, Guidelines for Use of Vapor Cloud Dispersion Models , Center for Chemical Process Safety, John Wiley & Sons, Hoboken, N.J. CCPS, 1999, Guidelines for Chemical Processe s Quantitative Risk Analysis, Center for Chemical Process Safety, John Wiley & Sons, Hoboken, N.J. CCPS 1999, Guidelines for Consequence Analysis of Chemical Releases , Center for Chemical Process Safety, John Wiley & Sons, Hoboken, N.J. CCPS 2010, Guidelines for Vapor Cloud Explos ion, Overpressure Vessel Burst , BLEVE and Flash Fire Hazards, Center for Chemical Proce ss Safety, John Wiley & Sons, Hoboken, N.J.

| 287 APPEN DIX E: PROCESS SAFETY CULTURE CASE HISTORIES The following case histories dem onstrate the application, or failure in application, of the core principles of process safety culture. Some are taken from the public domain, som e from privately-shared experiences, and some are fictional but based on real situations. These exam ples are ideal for group discussion. In addition to discussing the thought-provoking questions, readers can ask “Does anything like this happen here?” and “Are there learnings we could use to im prove our culture? In nearly all the exam ples, the source did not analyze process safety culture impacts or did not analyze them fully. So, the actual weaknesses in cultural core principles associated with each case m ay not be known. Others showed strengths that should be emulated, but readers could potentially do even better. E.1 Minim alist PSM S A specialty chem ical com pany produces m aterials using highly toxic feedstocks such as phosgene, chlorine, and several others in com plex highly exotherm ic reactions. The processes also use several flamm able industrial solvents. The inventories of these feedstocks are relatively large, e.g., the chlorine is stored and fed to the process in 90-ton rail cars, of which there are always at least three onsite. The facility also changes or trials products and introduces new ones or variants of existing ones frequently. The EHS M anager of the facility, who is responsible for the PSMS has a very long tenure and firm ly believes that the best approach to com plying with applicable regulations is to meet the m inim um requirements and no more. He has successfully negotiated with regulatory inspectors over the years and has been successful in restricting inspections only to the specifically covered areas. Essential Practices for Creating, Strengthening , and Sustaining Process Safety Culture , First Edition. CCPS . © 2018 AIChE . Published 2018 by John Wiley & Sons, Inc.B ased on Real Situations

301 11.15 Hendershot, D.C., Tell Me Why , 8th Annual International Symposium, Mary Kay O’Connor Process Safety Center, Texas A&M University, October 2005. 11.16 Kletz, T., Amyotte, P., Pr ocess Plants – A Handbook for Inherently Safer Design, 2nd Ed., CRC Press, 2010. 11.17 U.S. Chemical Safety Board (CSB), Oil Refinery Fire and Explosion , Case Study 2004-08-I-NM, October 2005. 11.18 U.S. Chemical Safety Board., Key Lessons from the ExxonMobil Baton Rouge Refinery Is obutane Release and Fire, Final Safety Bulletin, November 2016.

14. Operational competency assessment 159 Some common causes of performance gaps include: • Ineffective training: o The training did not specifically target the required set of competencies. o Learning objectives were not clearly defined. o The method of learning was not sufficient to develop the required competency. • Skill fade linked to infrequent tasks. o Even with acquired learning, if there is not enough opportunity to put the new competency into practice, an individual’s ability to do the task to the required standards decreases over time. • Excessive workload (e.g., due to understaffing) affects personal and team performance. • 'Drift' over time. Even individuals who are experienced and knowledgeable start to 'do things in their own way'- over time. This behavior can drift into being unsafe and need to be recognized when carrying out reviews. • Poor assessment practices, such as: o Poorly defined competency assessment matrices. o Assessors lacking training and expertise in assessment. • Personal reasons and circumstances, such as: o Poor health. o Challenging family circumstances. • Work-related conditions including, such as: o Poor quality of tools used for the task. o Poor team cohesion. o Difficult or poor relationships with co-workers. 14.5.3 Principles for managing performance gaps Effective management of performance gaps is based on the following principles: • Performance gaps should be dealt with promptly. The aim would be to: o Demonstrate commitment to a high-reliability culture. o the issue from becoming more serious over time, and prevent it having negative safety-related outcomes (e.g., leading to an incident). • Welfare and confidentiality should be maintained by: o Providing people with relevant support. For example, having one-on-one discussions to investi gate issues which may impact employee performance. o Ensuring confidentiality of performance reviews and follow up discussions.

379 A.3 Five Principles of Human Performance The United States’ Department of Energy 2009 manual “Human Performance Improvement Handbook Volume 1: Concepts and Principles” [121] (pages 1-19 and 1-20) gives the following principles: 1. “People are fallible, and even the best people make mistakes. 2. Error-likely situations are predicta ble, manageable, and preventable. 3. Individual behavior is influenced by organizational processes and values. 4. People achieve high levels of performance because of the encouragement and reinforcement re ceived from leaders, peers, and subordinates. 5. Events can be avoided through an understanding of the reasons mistakes occur and application of the lessons learned from past events (or errors).” The DOE (page 1-19) state that: “Excellence in human performance can only be realized when individuals at all levels of the organization accept thes e principles and embrace concepts and practices that support them.” “Integrating these principles into management and leadership practices, worker practices, and the organization’s processes and values will be instrumental in developing a working philosophy and im plementing strategies for improving human performance within your organization.” A.4 Twelve Principles of Error Management Professor James Reason and Alan Hobbs, in their 2003 book “Managing Maintenance Error, A Practical Guide” [17] offer 12 Principles of Error Management. These twelve pr inciples are as follows: 1. “Human error is both universal & inevitable: Human error is not a moral issue. Human fallibility can be moderated but it can never be eliminated. 2. Errors are not intrinsically bad: Success and failure spring from the same psychological roots. Without them we could neither learn nor acquire the skills that are essential to safe and efficient work. Appendix A - Human error concepts

92 | 3 Leadership for Process Safety Culture Within the Organizational Structure foster check-the-box thinking by inadvertently implying that m anagement values the metrics over the perform ance. Use Monetary Incentives with Caution In line with the above caution regarding metrics, leaders should exercise great caution when considering monetary incentives for achieving process safety related key performance indicators. Consider whether the incentive might foster check-the- box thinking or contradict the core principles of process safety culture in some other way and therefore backfire. See section 4.2 for a discussion of process safety culture and com pensation. No Fines does not Mean Strong Process Safety Performance The absence of violations or findings from recent regulatory inspections does not guarantee that the PSMS is functioning as it should. Inspectors and auditors cannot know a facility’s technology and management system as well as facility personnel, they visit the site only occasionally. Moreover, many process safety hazards do not fall under process safety regulations, but still must be m anaged to protect the com pany’s assets, workers, neighbors, and reputation. Facilities should certainly seek to assure compliance, but gaps in the PSMS can exist even with the absence of citations and findings. Reconcile Culture and Budget Leaders should provide adequate financial and personnel resources to m anage the process safety hazards of the facility. While leaders should challenge their teams to deliver results with efficient use of resources, cutting blindly will eventually erode a good culture. Instead, strengthening the process safety culture will help apply resources more efficiently. Conversely, throwing m ore resources at a weak process safety program cannot fix it in the presence of a poor culture.

Pipes 101 Table 6.8 Rule of thumb for selec ting three‐way connections. IfSuitable three‐way connection P&ID representation Y/uni2033 X/uni2033Y is the same size of X Tee Y is the same is one size smaller than XReduced Tee Y is smaller than X by two or three sizesO‐let Y smaller than half of X Tee and then reducer Table 6.9 Differ ent methods of connecting pipes end to end. Component Application P&ID schematic Socket welding, thread connecting Pipe size less than 2″ Butt welding Pipe size more than 2″ Flange Table 6.10 Differ ent methods of ending pipes. Component Application P&ID schematic Screwed cap Pipe sizes less than 2″ Plug Pipe sizes less than 2″ Blind flange Pipe sizes more than 2″ Welded cap (with or without drain valve) Quick connection Tend to be used in small to medium bore size pipesSP

PROCESS SAFETY AND MANAGEMENT OF ABNORMAL SITUATIONS 25 their duties with alertness, due thought, full knowledge, sound judgment, pride, and accountability, the impact of abnormal situations should be minimized, and serious process safety incidents prevented or mitigated. This requires key personnel with leadership skills, and strong management systems, practical organizations, and communications systems. For example: Operating procedures (containing safe operating limits, consequences of deviation from sa fe limits, troubleshooting, and actions required to correct a deviation must be up-to-date, accurate, easy to use and understand, and readily accessible. Organizations should avoid complexity by having a limited number of vertical management layers and departments, to avoid interface issues and poor communication that could make it more difficult to react to a dynamic abnormal situation. Operators and their supervisors should be empowered to shut down the plant in the event of a serious abnormal situation, if insufficient time is available to escalate the problem within a hierarchical organization. Management should drive a ‘no blame’ culture within the plant and commend actions that prevent or mitigate potential process safety incidents. In summary, most abnormal situations that result in process safety incidents also involve safety cultur e influences, including COO and OD issues (CCPS 2007a, 2011b). Every country develops its own culture, and its people develop habits, norms, and values that differ from other countries. National culture is ofte n a good predictor of attitudes, behaviors, and performance in the workplace. For example, in some cultures, operators are likely to escal ate an abnormal situation problem to the senior manager or superinten dent, rather than take action, especially if it involves plant shutdo wn or stop work. However, in other cultures, workers are more likely to follow written policies, practices and procedures that may or may not satisfactorily address abnormal situations. Within Europe, for exam ple, there are various sub-cultures with different approaches to an abnormal situation (EU-OSHA 2013). Similarly, in the USA there are su b-culture differences between states. For example, in some facilities, oper ators may have a ‘can do’ attitude and may not often refer to written op erating procedures, while in other

ACRON YM S YYYJJJ TNO Nederlandse Organi satie voor Toegepast Natuurwetenschappelijk Onderzoek (TNO; English: Netherlands Organization for Applied Scientific 3FTFBSDI UEL Upper Explosive Limit UFL Upper Flammable Limit VCE Vapor Cloud Explosion VLE Vapor Liquid Equilibrium XV Remote Activated/Controlled Valve

6.2 Assess the Organization’s Pr ocess Safety Culture |209 and positions along the hierarchy of the organization. Even sim ple cultural questions such as “Has the facility lost a sense of vulnerability with respect to process safety hazards?” m ay result in widely differing views among positions and levels (Ref 6.4). In each interview dealing with cultural issues, the interviewer should attempt to ask questions that are purposefully indirect. For instance, the following questions m ight be used to probe the “sense of vulnerability” issue: Do you believe that a catastrophic release could happen at this plant? Could an incident at this plant cause damage or harm offsite? How would you compare the likelihood of a catastrophic release to the likelihood of a car accident? To an airplane crash? To being struck by lightning? The interviewer may need to change the line of questioning during the interview, depending on interviewee’s responses. If a single employee’s answers are inconsistent with known risks at the facility, then the interviewer m ay conclude that their sense of vulnerability is weak. Then by posing these questions across the organization, the interviewer can determ ine if, for exam ple, workers feel vulnerable while m anagement does not. Interviews can be in formal settings such as conference rooms or informal such as around the lunchroom table. While interviews should usually be planned and scheduled, valuable input m ay also be gained from informal conversations that arise as the interviewee goes about the site. In union facilities, hourly employees may request the presence of a union official during the interview. In this situation, the interviewer should seek to determ ine if the role of the union official is to put the interviewee at ease, or control their response. • • •

| 255 Essential Practices for Creating, Strengthening , and Sustaining Process Safety Culture , First Edition. CCPS . © 2018 AIChE . Published 2018 by John Wiley & Sons, Inc.APPEN DICES Appendix A: Echo Strategies White Paper – Organizational Culture Appendix B : Other Safety & Process Safety Culture Frameworks Appendix C: As Low as Reasonably Practicable (ALARP) Appendix D: High Reliability Organizations Appendix E: Process Safety Culture Case Histories Appendix F: Process Safety Culture Assessment Protocol Appendix G: Hum an Behavior & Culture

4 • Process Shutdowns 48 Figure 4.1 Transient operating modes asso ciated with process and facility shutdowns. (Adapted from [15]) 4.3 Projects requiring equipment or process unit shutdowns This section will briefly summarize how process safety risks can be addressed effectively when preparing for and executing all types of engineering projects. Since each proj ect is unique, each one requires a systematic and disciplined approa ch to manage their process safety risks. All successful projects have structure, an execution plan and involve some form and degree of proj ect-related, as well as process