Implementation Issues of PSM in a Fertilizer Plant: An Operations Engineer’s Point of View M. Imran Rashid,a Naveed Ramzan,b Tanveer Iqbal,b Saima Yasin,b and Sana Yousafb a Operation Engineer, Dawood Herculeus fertilizer Complex, Pakistan b Department of Chemical Engineering, University of Engineering and Technology, Lahore, Pakistan; ramzan50@hotmail.com (for correspondence) Published online 30 January 2013 in Wiley Online Library (wileyonlinelibrary.com). DOI 10.1002/prs.11553 In this article, guidelines are provided for implementation of the elements of process safety management (PSM) such as management of change (MOC), process hazard analysis (PHA), incident investigation, emergency planning, and response. The role of mechanical integrity, operating procedures, compliance audits, pre-start-up safety review, contractors, training, work permits, and process safety information for implementation of the PSM in a fertilizer plant are discussed. Implementation of MOC is an important step for the adoption of PSM standards and a necessary condition for internal and external audits of the plant. There are many issues linked with implementation of PHA like plant modifications, shut downs, and production losses as well as the behaviour of the design engineer whenever modifications are required. Consequence analysis, an evaluation of an incident in terms of its effects on environment, equipment, and people is of great help. Incident reporting for a company can be improved by ensuring confidentiality and not exposing the reporting person. Improvement in the operating procedures and preserving mechanical integrity of the process plant are necessary conditions for the implementation of the standards. Internal and external audits of the company are the most important part of the PSM implementation. In general, the PSM implementation requires much effort and time C 2013 American Institute but pays off well if implemented fully. V of Chemical Engineers Process Saf Prog 32: 59–65, 2013 Keywords: process safety management; process safety information; process hazard analysis; management of change INTRODUCTION The compliance period established by the occupational safety and health administration (OSHA) for implementation of the process safety management (PSM) standard [29 CFR 1910.119(e)(1)(i)-(iv)] was between May 26, 1992 and May 26, 1997. It was assumed that the number of plant accidents would be reduced significantly after implementation of the standard. But 46 full investigation reports (1998–2008) of plant facility accidents published by U.S. Chemical Safety and Hazard Investigation Board (CSB) indicate that accidents have not decreased as expected [1]. Accidents are continuously damaging the people, plant facilities, and environment. Half of the reported accidents occurred in the PSM imple- C 2013 American Institute of Chemical Engineers V Process Safety Progress (Vol.32, No.1) mented plants while other half in smaller plants that are not covered by PSM [2]. A process reactor vessel was destroyed in December 2007 by a runaway reaction at the T2 plant in Jacksonville, Florida, similar to the previous accidents [3]. The Canadian Chemical Producers Association 2004 processrelated incidents measure analysis report of 89 incidents has shown that six PSM elements contributed to 85% of occurred incidents. Namely, these elements were, “Process and equipment Integrity,” “Process knowledge and documentation,” “Process risk management,” “Human factors,” “Management of change,” “Capital project review and design procedures” [4]. A similar study was conducted by Blair (2004) on 21 chemical accident investigations of CSB to identify repeatedly occurring causes. His top five ranked categories include: “maintenance procedures,” “process hazard analysis (PHA),” “engineering design and review,” “management of change,” and “operation procedures” [5]. Detailed analysis of incident data has identified operational discipline as a key factor in nearly half of the PSM incidents and was a key factor in over 50% of process incidents in 2000 [6]. Contribution of operational discipline to incidents at one site has increased from 61% in 2003 to 86% in 2006 [7]. These reports clearly demonstrate that there are potential issues in implementation of regulations from OSHA. PSM regulation has excellent requirements that could and should prevent accidents if process plants follow the regulation as intended [2]. Now the need is to address the practical difficulties faced in implementing PSM elements and provide guidance for industries to overcome the issues. Many authors have provided recommendations for continuous improvements in PSM systems [2–4,8–11]. Actual implementation costs for the PSM regulation have been orders of magnitude higher than originally estimated [12]. PSM auditing costs are high, and people are doubtful about its effectiveness [8]. PSM documentation is very difficult and requires great effort. Also, good documentation is just beginning and proper utilization is again difficult [2]. One major pitfall in implementation of effective PSM is that many companies believe that the “Risk Assessment” is sole responsibility of HSE and risk specialists, instead of being carried out in a more pragmatic way [13]. This article provides the guidelines for implementation of the elements of PSM such as management of change (MOC), PHA, incident investigation, emergency planning, and response. The role of mechanical integrity, operating procedures, compliance audits, pre-start-up safety review (PSSR), March 2013 59 contractors, training, work permits, and process safety information for implementation of the PSM in a fertilizer plant are also discussed. MANAGEMENT OF CHANGE MOC is not only one of the basic elements of OSHA’S PSM System but also is required by U.S. EPA’s Risk Management Program Regulation [14]. With any idea or problem, first you are required to initiate an MOC, that is, provide the problem/purpose of change, a proposed solution, and/or justifications for the said change. Then MOC approval from different departments such as operations, technical, project, safety, and authorizations from the higher management is required. After approval of the MOC proposal, you need to complete a job safety analysis. After the job, updates to the P&IDs, operating procedures, maintenance procedures, risk assessments and its accommodation in PHAs, PSSRs, employees trainings (or involved contractor workers training) along with communication to all shifts or other staff in factory are required. Employees involved in operation and maintenance of the process affected by said change need to be trained as well. How difficult it is to implement all above requirements when several hundred MOCs are raised annually in big organizations as reported by Nir Keren after gathering data from 26 facilities [15]. Follow-up with MOC in emergency situations is also difficult. Sometimes conflict arises when two departments disagree whether the job will be covered under the MOC or work permit system. Hence, proper training for MOC originators, processors, approvers, record keepers, and auditors are required after scheduled intervals. Figure 1 shows the MOC cycle for all steps in MOC. Detailed check lists are important but every change does not have to use every check list and the MOC process may become so uncertain that potential hazards may be overlooked in assessments [16]. Inadequate MOC has resulted in accidents such as Flixborough [17], fire and explosion at Giant Industries’ Ciniza oil refinery [18], and hydrogen reformer furnace failure at Syncrude Canada Ltd [19]. As PSM regulations require a PSM audit at least every 3 years, proper records of MOCs and whether jobs have been done under MOCs are required. However, different internal audit schemes by plant personnel or by corporate audits may be established inside the company [20] but on the other hand these also require much effort and cost. One other suggestion is to develop a software system, on which every MOC description is present along with its status in technical study phase, management review phase or approved. But again you need people to update and maintain such a system and manpower requirements will increase. One other suggestion is a monthly MOC meeting in which people from different departments should involve to discuss progress of all generated MOCs. Further guidance is provided by Nir Keren in a questionnaire for bench marking MOC [15]. Figure 1. MOC cycle/hierarchy Processes installed before 1980s, when technology was not so advanced, have undergone a lot of modifications and P&IDs were not always updated. The first main problem faced during PHAs is P&IDs that do not match site installations. P&IDs do not show all flanges of process plant, so more and more field visits are required. If leaking flanges due to reduced life, wrong gasket materials, or insufficient tightening (control valve upstream flange, downstream flange, bonnet flange, by pass line flange, isolating valves flanges) is considered to develop scenarios then 100 and 1000 different scenarios are possible, which confuses the PHAs team and turn PHAs into boring exercises. Operation engineers work in shifts and are well aware of incidents/accidents occurred in their shifts but not about all previous accidents. Usually, incident investigation reports are not available. This makes it difficult to consider all accidents in PHAs. Also, guidance from OSHA is not clear about consideration of previous incidents. Should previous incidents be limited to only those that occurred in a particular process that is undergoing a PHA? Or should other incidents be included, such as ones that occurred in other processes in the same plant; similar processes operated by the company at different plants; or all known published incidents involving a similar process or a particular piece of equipment? [1]. It is also difficult to consider all plant modes of operation such as startup, shut down, emergency shut downs, reduced load operation of plant, or partial plant shut down, and partial running conditions. Identification of all hazards associated with different forms of energy as given in Table 3.1 in the DOE Handbook [21] also requires effort and time. It is also difficult to address all hazard scenarios such as tube leaks of all exchangers, leakage from all process lines, leakage from all process equipments, leakage from all flanges, 60 DOI 10.1002/prs PROCESS HAZARD ANALYSIS March 2013 Published on behalf of the AIChE Process Safety Progress (Vol.32, No.1) leakage due to wrong gasket materials, leakage due to possible damaged gaskets, overflowing of process vessels, upsets in all process vessel levels, falling materials, all pumps seal or gland leakages, overpressures, high or low temperatures, and variations in flows. Now consider the difficulty involved for the PHA team if the team has to made qualitative consequence estimates for all such scenarios to determine type, severity, and number of injuries. Then, the PHA team should also consider number of people who may be exposed, the duration of exposure, evacuation routes and ventilation, and the effects of released material on various people. Sometimes workers are not free for PHAs. Sometimes process design engineers feel that their design is being questioned. For running plants, managers believe that a PHA will cause plant modifications, shut downs, and production loss [22]. Costly recommendations are not always welcomed or sometimes these can only be implemented when plant is down, so you may have to wait for a year. PHA teams sometimes use less rigorous PHA methodologies (What-If, Checklist) rather than hazard and operability study (HAZOP)/ failure mode and effect analysis (FMEA) because of time limitations from higher management and the estimated time for each node [22] is ignored. Also many PHA teams have not a single member knowledgeable in the specific PHA methodology being employed as per OSHA recommendations. Sparing the staff for PHAs, scheduled training and generating incident databases is important. Integrating job safety analysis (JSA) with PHAs [23] and suggestions about PHAs from Mark Kaszniak [1] are helpful for improvement. INCIDENT INVESTIGATION On April 5, 2012 when an operation engineer in a fertilizer plant started the Ammonium Carbamate pump 3 KV motor during initial startup phase, it did not start; a minor sound from the motor and a huge voltage dip in the control room were observed. On checking the motor, it was found to be burnt. Incident investigation was done using the WhyTree Analysis as shown in Figure 2 and it yielded the astonishing result that all 3 KV and 11 KV motors had not had any inspection schedule since plant installation. Usually, people are not well aware of the type of incidents that have occurred so this causes difficulty in identifying the root causes of the incident. Sometimes problems appear when people do not report the minor incidents or near missess. Major hindrances in incident reporting are the fear of punishment, disclosing one’s identity, and loss of personal relationships [24]. So incident report forms should not mention name of the person but often management can determine the responsible person from the area in which work was carried out or from time of incident. Authors have experience that sometimes such incidents have even affected the career of persons involved. So development of confidence and trust among workers is required to make these reports true and worthy for preventing future incidents. Recurrence of incidents can also be avoided by utilizing these incident investigation reports. Crowl and Louvar have also reported on the significance of incident investigation [25]. Learning from accident root causes is worthless unless shared with all company employees. Again during its communication, some specific people or specific department staff may not feel good about it. This creates dissatisfaction or conflict among the employees. Some companies even cannot understand the root causes of multiple incidents because of unavailability of skilled employees who can use different models such as Domino Loss Causation Model and Kletz’s Layered Investigation Model etc. Also, inherent safetybased incident investigation methodologies [26] are not adopted by industries which lack in such expertise. One other problem is that recommendations generated after the investigations of the accidents may not be closed in a timely manner Process Safety Progress (Vol.32, No.1) because some can only be implemented when the unit is down and this sometime takes a year. Some experienced professionals feel bad that they are asked questions during incident investigations and people feel that telling the truth will affect their careers or others. EMERGENCY PLANNING AND RESPONSE Many process plants have emergency response teams and procedures. However, often only a general emergency plan exists and for each separate case, a separate plan does not exist. Similarly, a general assembly point exists where people have to go in case of emergency siren. Emergency plans for ammonia release, carbamate release, chlorine release, vapor cloud explosion, heat radiation, and explosion blast wave must be different from each other and accordingly assembly points or evacuation procedures. Plants usually lack in training of the people regarding behavior in an emergency and people get confused in emergencies. The use of consequence analysis in most industries is rare and without it, emergency response planning is hard to do accurately. Different software can be used for consequence analysis such as: ALOHA [27], Computational Fluid Dynamics (FLACS, Fluent, CFX, etc.), BP Cirrus, Shell FRED, Canary from Quest Consultant, PHAST from DNV, Slab from EPA, LNGFire3, Breeze LFG Fire/Risk, and DEGADIS etc. Models (e.g. SAFER) can also be used to calculate the areas impacted by fires, toxic chemical releases, explosions, etc. for events such as holes in pipes, stack releases, and spills. Using this analysis, emergency preparedness plans are made, and the level of community evacuation needed is judged. Many companies do not have software available or the required expertise to use such software and do not plan there emergency response accordingly. These limitations cause difficulty in implementations of PSM regulations. MECHANICAL INTEGRITY As per PSM recommendations, mechanical integrity of all process equipment, piping system, and process controls should be checked. The main problems observed are that main equipment is missed in the inspection schedules (as indicated by Figure 2 incident) and sometimes the inspection cannot be done because at the time the inspection is planned the machinery is running (pumps/ compressor). Also, some times when an inspection (e.g., vibration) is scheduled, then the equipment is not running and is on standby; therefore, the required inspection cannot be done. Also, inspection of low switches (e.g., lube oil low-pressure tripping) cannot be done in a running plant, because these cannot be isolated if machinery is running and some machines do not have stand by machines. So this inspection will cause a reduction in production and management usually does not allow such inspections. Similarly, some interlocks such as of conveyors can only be checked when they are stopped, so in continuous running plants you may have to wait a whole year for check interlock operation. Some things are difficult to inspect, for example, glands of valves and mechanical seals of pumps because these require a lot of effort for inspection. So valve glands and mechanical seals should have proper schedule for replacement, but again it is difficult to maintain records in fertilizer plants where many thousands of valves may be present. It is difficult to remove insulation of all lines or vessels for inspection, because when removed, insulation is damaged and you have to do it again. Sometimes inspection methods are not correct and many inspections occur without any useful output. It is also observed that in some plants relief valves/pressure safety valves (PSVs) calibrations were not done for more than 5 years because of short turn around durations and poor inspection management. Figure 2 also indicates that the absence of inspection schedules was one major cause for incidents. Why Published on behalf of the AIChE DOI 10.1002/prs March 2013 61 Figure 2. Incident investigation. wileyonlinelibrary.com.] [Color figure can be has not somebody developed inspection schedules for these motors? One reason may be the work load on the staff or the small number of staff available for maintenance. Hence, factories should have proper level of maintenance staff and staff must follow vendor recommendations for all machinery inspections and/or replacement of parts. viewed in the online issue, which is available at example, a change from chromate base treatment to phosphate base. But this change is usually not updated in operating procedures. Most of the companies have operating procedures. However, some common missing data and or activities identified in operating procedures are given in Table 1. This table also shows the efforts required to implement this PSM element. OPERATING PROCEDURES With passage of time and increasing production demands, organizations change systems may be due to technology changes or due to mitigation of hazardous chemicals, for COMPLIANCE AUDITS 62 DOI 10.1002/prs March 2013 Published on behalf of the AIChE The most important element of PSM is the audit-not infrequent official audits but many internal audits to check that Process Safety Progress (Vol.32, No.1) Table 1. Upgrades required for compliance with the OSHA PSM regulation. 1 2 3 4 5 6 7 8 9 10 11 12 Specific procedures for jobs covering both maintenance and operation aspects are not present. No specific procedures for specific jobs exist, e.g. specific procedure for motor painting or pump painting does not exist. Similarly for pump bearing oil replacement, alignment checking and for mechanical seal replacement different procedures are not present Detailed shut down procedures and emergencies procedures are missing Operating procedures do not mention the hazards of operation e.g. in order to stop a circulation pump for a vessel, the vessel level should be low, otherwise it can overflow and result in splashing Effect of parameter (Temp, pressure, flow, level) variations on plant and mitigating actions are missing Procedures for some activities such as pump, filter or lube oil cooler change over are missing Details of Plant startup after turn around are missing Hazards of specific materials such as insulation materials, diatomaceous earth and asbestos are missing Procedures for some activities are missing e.g. entry into empty vessel/reactor for inspection Actions in case of loss of electricity/cooling water or partial failure of electricity/cooling water Parameters/conditions at reduced plant operation and parameters upper and lower limits Replaced chemicals and lubricating oils are not mentioned Plant control system changed from PLC to DCS operation but related instrumentation and alarms data is not updated Table 2. Different things checked in daily safety audits. 1. 2. 3. 4. 5. 6. 7. 8. 9. Damaged earth (grounding) cables of motors/exchangers Damaged insulations of hot pipes or steam tracing lines Missing ladder cages/safety locks Missing motor fan covers Steam condensate leakages from height in the walkways Motors coupling guards not properly fixed Uncovered pits Hanging sheets (metallic or asbestos) from height Over-head cranes hanging chains not fastened properly etc. process is used as required [2]. Different type of internal audit schemes may be used to check the effectiveness of PSM elements. These are: daily safety audit, work permits audit, and behavior observation audit. In a daily safety audit, conditions against the safety are checked as listed in Table 2. Work permit audits are done to check the appropriateness of work permit system. The purpose of the behavior observation audit is to check safe and unsafe behaviors of people regarding the use of PPE, tools and equipment, and work methods. Figure 3 depicts the safe and unsafe behaviors in X Factory for 15 months. Some people resist audits as is shown in the trend for the category “reaction of people.” The factor is more common in industries in which highly experienced workers are present. Companies can form teams consisting of safety personnel, operations, and those involved with regulatory compliance. Such teams can schedule their own audit schemes or can conduct combined audits with the PSM audit. Such audits of different PSM elements including findings and corrective actions are given by Michael R. Green [28]. new facility because of overtime issues, shortage of manpower, or time limitations to run the facility. The other issue is that proper operating procedures, startup procedures, emergency handling procedures, and safety considerations are not always completed before startup of new facility as recommended by OSHA. CONTRACTORS A large number of contractors are available who provide the workforce for daily activities such as loading, unloading, and packing. The major issue with contractor workers occurs when contractor does not provide mandatory personal protective equipment (PPE) to their workers. So screening criteria for hiring contractors should be used which includes contractor health and safety records. It is also observed that contractors sometimes do not discuss safety with workers, for example, hazards involved with insulation materials, paints, or asbestos sheets. Contractors should be trained regarding the safety of its workforce. Also, companies should have spare mandatory PPE which can be issued to the contractor workers for required job. Usually, contractor workers are less educated, so they may not use PPE when they see no one is around. Plant operation or safety engineers should discuss with them how important PPE is to your health and safety, both for their own life and for their family, to convince them to use all PPE. Sometimes contractor workers perform nonroutine activities like vessel entry, welding, radiography, commissioning, or services in an annual turnaround. It has been observed that contractors do not inform the working shift engineers and start welding and other activities. Actually, they should always inform and consult the related shift engineers for guidance and supervision. Engineers should also discuss safety with them so that all safety precautions can be used during any nonroutine work authorizations performed by the contractors. PRE-START-UP SAFETY REVIEW PSSR is required after an MOC job, for new equipment or for startup after a turn around. One problem in implementation is when the PSSR is not conducted before the installation of facility and it becomes difficult to implement the PSSR recommendations after facility installation as the items have not been agreed upon at the time of contract. After installation of an ammonia storage tank in a fertilizer plant, when the PSSR was carried out, the following required items were detected missing: A safety ladder required to operate the valve installed at top of the storage tank, a platform, and a supplied air mask at top of the tank. The other difficulty in implementation is related to the employee training. Employees are usually not trained at the Process Safety Progress (Vol.32, No.1) TRAINING All employees must have proper training on the facilities they are operating. If implementation of a PSM system is in progress, then more training will be required regarding PHA, MOC, incident investigation, PSSR, job safety analysis, emergency response planning and preparedness, near miss reporting, mechanical integrity, and hot work permit. Newly hired employees should be provided an overview of the process, health, and safety hazards associated with job tasks, safe work practices, operating procedures, emergency operations, and emergency shutdown. Audits have identified that newly appointed employees were not always properly trained in operating procedures [28]. If the operation staff Published on behalf of the AIChE DOI 10.1002/prs March 2013 63 Figure 3. Safe and unsafe behaviors trend indicating safety progress. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.] works in rotating shifts, major problems have been observed regarding the schedule of training especially when limited number of spare staff is available. The other issue is that fire fighting training is only provided to the fire fighters, not to the all factory staff. However, a few oil refineries have training schedules for all working staff not just the fire fighters. Also, training covering all aspects such as vapor cloud explosion, fire, toxic chemicals (ammonia, ammonium carbamate, chlorine) release, heat radiation etc. is not always provided. Also, often less training is provided regarding the hazardous waste management and mitigation of toxic chemical spills. HOT WORK PERMIT The main issue observed in implementation is the unawareness of many supervisors regarding the hazards present which have resulted in accidents. A welding job was performed in a small furnace and after welding job a dye test was performed to check welding quality. The dye test detected flaws, and welding was resumed without any sampling of the inside atmosphere. This clearly indicates that the supervisor was not aware of the combustion hazards associated with dye check materials. It is clearly mentioned in OSHA standard 1910.252(a)(2)(xiv)(B) that the supervisor shall determine the combustible materials and protect combustibles from ignition [29]. The other main issue observed is that hot work jobs are carried out usually without inert gas (e.g., N2) purging as specified in OSHA standard 1910.252(a)(3)(ii). The main problem behind it seems to be that the operations engineer has to deal with many hot work permits at the same time. He will be involved with conducting a JSA for said job, coordinating lab analysis, ensuring proper isolations or dismantling of electricity connections, depressurizing or venting of equipment, and communications with management. Hence, inert gas purging costs, time requirements, and the opera64 March 2013 Published on behalf of the AIChE tions engineer’s tight job schedule are the main reasons behind hot work jobs being done without inert gas. PROCESS SAFETY INFORMATION Process safety information is the compilation of complete and accurate written information regarding process technology, process equipment, and process chemicals. All such information should be available to all factory staff whether working on PSM or not. The main drawback observed in material safety data sheet (MSDS) is that some chemicals are missing, for example, epoxy paints and their hazards are not mentioned. Exposure limits of various chemicals (e.g., ammonia) are given but concentrations which are dangerous for human beings for different time intervals (e.g., AEGL-1, AEGL-2, and AEGL-3) are not mentioned. Similarly, a chemical interaction matrix is not present in many companies and if present, it does not cover all the chemicals used/present in plants. Also, chemical interactions for all chemicals involved in a process factory are not covered. Toxic effects of different materials involved is not known, for example, toxic effects associated with insulating materials (glass wool, low density fiber glass insulation, foam glass insulation, fiber glass insulation), refractory materials (alkaline earth silicate fiber, refractory castable), construction materials (carbon steel, stainless steel during cutting/welding), different acids, hydrazine, catacarb solution, and diatomaceous earth. Usually, the effects of deviations from operating limits are not in written form. Also, the consequences of such deviations that affect the process as well as HSE are not available. LITERATURE CITED 1. M. Kaszniak, Oversights and omissions in process hazard analysis: lessons learned from csb investigations, Process Saf Prog 29 (2010), 264–269. DOI 10.1002/prs Process Safety Progress (Vol.32, No.1) 2. J.F. Louvar, Improving the effectiveness of process safety management in small companies, Process Saf Prog 27 (2008), 280–283. 3. B. Knegtering and H.J. Pasmsn, Safety of the process industries in the 21st century: A changing need of process safety management for a changing industry, J Loss Prevention Process Ind 22 (2009), 162–168. 4. P.R. Amyotte, A.U. Goraya, D.C. Hendershot, and F.I. Khan, Incorporation of inherent safety principles in process safety management, Process Saf Prog 26 (2007), 333–346. 5. H. Luo, The effectiveness of U.S. OSHA process safety management inspection—A preliminary quantitative evaluation, J Loss Prevention Process Ind 23 (2010), 455–461. 6. J.A. Klein, Operational discipline in the workplace, Process Saf Prog 24 (2005), 228–235. 7. J.A. Klein and B.K. Vaughen, A revised program for operational discipline, Process Saf Prog 27 (2007), 58–65. 8. S. Arendt, Continuously improving PSM effectiveness—A practical roadmap, Process Saf Prog 25 (2006), 86–93. 9. I.S. Sutton, Use root cause analysis to understand and improve process safety culture, Process Saf Prog 27 (2008), 274–279. 10. J. chosnek and R. Clifton, Improved process safety management and simple metrics, Process Saf Prog 27 (2008), 284–288. 11. J.A. Klein and S. Dharmavaram, Improving the performance of established PSM programs, Process Saf Prog 31 (2012), 261–265. 12. W.G. Bridges, The cost and benefits of process safety management: Industry survey results, Process Saf Prog 13 (1994), 23–29. 13. M. Pagnini, S. Milaneses and A.D. Little, Practical Implementation of “Risk Assessment” through Process Safety Management Systems, s.p.ACorsoSempione 66/A, 20154Milan, Italy. 14. Risk Management Programs for Chemical Accidental Release Prevention, Final Rule, 40 CFR Part 68, Environmental Protection Agency, Washington DC, Federal Register, 68, 120, pp 31667–31732, June 20, 1996. 15. N. Keren, H.H. West, and M. Sam Mannan, Bench marking MOC practices in the process industries, Process Saf Prog 21 (2002), 103–112. Process Safety Progress (Vol.32, No.1) 16. K. Hanchey and J.R. Thompson, The challenge to implement and maintain an effective PSM program, Process Saf Prog 30 (2011), 319–322. 17. R.E. Sanders, Designs that lacked inherent safety: Case studies, J Hazard Mater 104 (2003), 149–161. 18. CSB, Oil refinery and explosion (Giant Industries’ Ciniza oil refinery), Case study, US Chemical Safety and Hazard Investigation Board, Washington, DC, 2005. 19. M. Rogers, Lesson learned from an unusual hydrogen reformer furnace failure, Process Safety and Loss Management Symposium, 55th Canadian Chemical Engineering Conference, Canadian Society for Chemical Engineering, Toronto, ON, October 16–19, 2005. 20. R. Wayne Garland, Management of change auditing system, Process Saf Prog 30 (2011), 342–345. 21. DOE Handbook, Chemical Process Hazard Analysis by U.S. Department of Energy, Washington D.C 20585, DOEHDBK-100-2004, 2004, 13–14. 22. N. Hyatt, Guidelines for Process Hazard Analysis, Hazard Identification and Risk Analysis, Dyadem Press Canada, 3rd ed., 2003. 23. R.L. Collins, Integrating job safety analysis into process hazard analysis, Process Saf Prog 29 (2010), 242–246. 24. Safety Management System Toolkit, Joint Helicopter Safety Implementation Team of the International Helicopter Safety Team, The international Safety Symposium 2007 Montreal, Quebec, Canada. 25. D.A Crowl and J.F Louvar, Chemical process safety fundamentals with applications, 2nd Ed., Prentice Hall, Upper sadle River, NJ, 2002. 26. A. Goraya, P.R. Amyotee, and F.I. Khan, An inherent safety—based incident investigation methodology, Process Saf Prog 23 (2004), 197–205. 27. EPA, National Oceanic and Atmospheric Administration, Aloha User’s Manual, The CAMEO Software system, Washington DC, February 2007. 28. M.R. Green and R. Lide, Auditing global logistics operations—A process safety focus, Process Saf Prog 29 (2010), 127–132. 29. OSHA standard 29 CFR subpart Q-Welding, Cutting and Brazing, 1910.252, general requirements (a). Published on behalf of the AIChE DOI 10.1002/prs March 2013 65