4
Identification and Use of Process Safety Metrics

INTRODUCTION

This chapter describes recent efforts by the chemical and petroleum industries to define process safety metrics suited to the needs of their enterprises. It then discusses how such concepts can be leveraged and applied at the Pueblo Chemical Agent Destruction Pilot Plant (PCAPP) and the Blue Grass Chemical Agent Destruction Pilot Plant (BGCAPP).

Leading and lagging process safety metrics have been of great interest for at least the last 5 years. Process safety and the metrics for such safety have been of concern especially in the chemical and petroleum industries, because it is those sectors that handle or produce toxic and flammable materials, which if released into the environment have the potential to cause multiple fatalities or injuries and significant environmental damage. Since the mid-1990s, these industries have used major unplanned releases of chemicals or energy as a primary process safety metric. Trade associations, including the American Chemistry Council and the American Petroleum Institute (API), have developed and implemented a common definition of a process safety incident (PSI) and have used the PSI to measure the relative performance of the companies that are members of those associations. The definition developed in the mid-1990s included any unintended releases of hazardous chemicals that exceeded the threshold quantity given in 40 CFR 302.4, a listing of designated CERCLA1 hazardous substances, or events that resulted in serious injury or damages in excess of $25,000.

Although the PSI was used, as defined, for more than 15 years by many U.S. chemical and petroleum companies as a performance metric, it never achieved sufficient acceptance and utilization as a benchmark owing to the lack of broad consensus that it was a good indicator of performance. This was in part because the table of threshold quantities given in 40 CFR 302.4 was not itself viewed as a good indicator of equivalent risks of the chemicals listed nor did it include all chemicals. International acceptance was also hampered by the metric’s association with a U.S. regulation.

PROCESS SAFETY METRICS FROM INDUSTRY AND ORGANIZATIONS

American Institute of Chemical Engineers’ Center for Chemical Process Safety Metrics

In 2006, the Center for Chemical Process Safety (CCPS) launched a project to develop better leading and lagging process safety metrics.2 The aim of the CCPS Metrics Project was to establish definitions of lagging metrics that would be broadly accepted internationally and be useful for benchmarking relative performance and parallel the Occupational Safety and Health

1

CERCLA is the Comprehensive Environmental Response, Compensation, and Liability Act, commonly known as the Superfund. This document lists dangerous chemicals and gives threshold quantities for the purpose of defining a process safety incident.

2

“Leading metric” and “lagging metric” are defined in Appendix A.



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 26
4 Identification and Use of Process Safety Metrics INTRODUCTION that resulted in serious injury or damages in excess of $25,000. This chapter describes recent efforts by the chemi- Although the PSI was used, as defined, for more cal and petroleum industries to define process safety than 15 years by many U.S. chemical and petroleum metrics suited to the needs of their enterprises. It then companies as a performance metric, it never achieved discusses how such concepts can be leveraged and sufficient acceptance and utilization as a benchmark applied at the Pueblo Chemical Agent Destruction Pilot owing to the lack of broad consensus that it was a good Plant (PCAPP) and the Blue Grass Chemical Agent indicator of performance. This was in part because the Destruction Pilot Plant (BGCAPP). table of threshold quantities given in 40 CFR 302.4 was Leading and lagging process safety metrics have not itself viewed as a good indicator of equivalent risks been of great interest for at least the last 5 years. Pro- of the chemicals listed nor did it include all chemicals. cess safety and the metrics for such safety have been International acceptance was also hampered by the of concern especially in the chemical and petroleum metric’s association with a U.S. regulation. industries, because it is those sectors that handle or p roduce toxic and flammable materials, which if PROCESS SAFETY METRICS FROM INDUSTRY released into the environment have the potential to AND ORGANIZATIONS cause multiple fatalities or injuries and significant environmental damage. Since the mid-1990s, these American Institute of Chemical Engineers’ Center for industries have used major unplanned releases of Chemical Process Safety Metrics chemicals or energy as a primary process safety metric. Trade associations, including the American Chemistry In 2006, the Center for Chemical Process Safety Council and the American Petroleum Institute (API), (CCPS) launched a project to develop better leading have developed and implemented a common definition and lagging process safety metrics.2 The aim of the of a process safety incident (PSI) and have used the PSI CCPS Metrics Project was to establish definitions of to measure the relative performance of the companies lagging metrics that would be broadly accepted interna- that are members of those associations. The definition tionally and be useful for benchmarking relative perfor- developed in the mid-1990s included any unintended mance and parallel the Occupational Safety and Health releases of hazardous chemicals that exceeded the threshold quantity given in 40 CFR 302.4, a listing of designated CERCLA1 hazardous substances, or events Compensation, and Liability Act, commonly known as the Super- fund. This document lists dangerous chemicals and gives threshold quantities for the purpose of defining a process safety incident. 1CERCLA 2“Leading metric” and “lagging metric” are defined in Appendix A. is the Comprehensive Environmental Response, 26

OCR for page 26
27 IDENTIFICATION AND USE OF PROCESS SAFETY METRICS Administration’s (OSHA’s) injury/illness metric, which to select metrics that are specific or applicable to that has been utilized broadly to benchmark worker safety organization. It also included an appendix document performance. Another aim of the project was to iden- containing several hundred potential metric options tify leading metrics that would monitor management that an organization should consider, depending upon systems or other early indicators of necessary actions the areas of performance that are most important or in that had to be taken to avoid process safety incidents, need of strengthening for that organization. especially catastrophic incidents such as those at Union The relationship of the hierarchy of incident catego- Carbide in Bhopal, India; BP in Texas City, Texas; and ries to the James Reason barrier model is shown in Fig- Phillips in Pasadena, California. ure 4-1. The latter model illustrates independent layers Industry interest and participation in the CCPS Met- of protection, which can alternatively be illustrated by rics Project grew with the release of The Report of the the “Swiss cheese” model that is shown in Figure 4-2. As explained in the two previously noted CCPS docu- BP U.S. Refineries Independent Safety Review Panel (also known as the Baker panel) and the U.S. Chemical ments, there are typically multiple independent layers, Safety Board’s (CSB’s) Investigation Report: Refinery or barriers, that prevent an incident from occurring or Explosion and Fire, Report No. 2005-04-I-TX, follow- that limit the severity of an incident. When all process ing the incident at the BP Texas City refinery in 2005. safety barriers are in place, a single barrier can typically Both reports called upon industry to develop and imple- fail without significant consequences. However, when ment better leading and lagging metrics. The CCPS multiple barriers fail, the probability that an incident Metrics Project resulted in two publications, a process can occur is increased. Individual barrier failures may safety metric pamphlet in December 2007 that recom- often occur without being noticed until a second or mended specific metrics for industry benchmarking3 third barrier has failed. For this reason, metrics are and a book in 2009 titled Guidelines for Process Safety needed to ensure the integrity of all barriers. Metrics. The latter document recommended a process for companies to adhere to in selecting and implement- American Petroleum Institute Metrics ing other metrics appropriate for their facilities.4 The metrics in the CCPS publications were embraced by Following the completion of the CCPS Metrics many U.S. and international trade associations and Project, the CSB still desired that an American National became the basis for metrics collected by those orga- Standards Institute (ANSI) standard be developed to nizations. These publications also describe a hierarchy codify the recommendations on process safety metrics. of metrics, both lagging and leading. Lagging metrics It hoped in this way to ensure that all relevant compa- relate to events that actually occurred (e.g., unintended nies and stakeholders would support a common set of releases of chemicals). Leading metrics may include metrics. CSB requested that API work with the United near-miss events that did not result in an unintended Steelworkers’ Union to sponsor an ANSI standard release, management system failures (e.g., missed or project. A committee was organized, and an ANSI stan- overdue inspections), activation of safety systems, or dard, Process Safety Performance Indicators for the other events that might indicate areas requiring atten- Refining and Petrochemical Industries (ANSI/API RP tion to reduce the likelihood of a significant event. 754), was created. The vast majority of definitions and The hierarchy is illustrated as a pyramid in Figure concepts developed by the CCPS Metrics Project were retained in the ANSI/API RP 754 document.5 Since 4-1. There are typically more minor events than major events, more near misses than actual releases, and more this document was developed shortly after issuance of management system defects or other early indicators the original CCPS metric definitions, the API commit- than near misses. tee used the opportunity to make minor modifications The CCPS Guidelines for Process Safety Metrics to metric definitions based upon lessons learned from expanded the discussion of metrics by describing the early implementation by users of the original CCPS processes that a company or organization should use metrics. See Box 4-1 for definitions of Tier 1-4 process safety events from API RP 754. 3Available at http://www.aiche.org/uploadedFiles/CCPS/Met - rics/CCPS_metrics%205.16.08.pdf. 4Available at http://www.wiley.com/WileyCDA/WileyTitle/pro- 5See ductCd-0470572124.html. http://www.api.org/Standards/new/api-rp-754.cfm.

OCR for page 26
28 PROCESS SAFETY METRICS AT THE BLUE GRASS AND PUEBLO CHEMICAL AGENT DESTRUCTION PILOT PLANTS C O N H S A E Z Q A U Process R E D N safety C E incident C O N H S A Other incidents E Z Q A U (e.g., all other loss of primary R E D N containment or fires) C E C O N H S A Near miss (system failures E Z Q A which could have led to an incident) U R E D N C E C O N H Unsafe behaviors or insufficient S A E Z Q A operating discipline U R E D N C E Legend: Barriers OR Layer of Protection or layer of protection Mitigation and emergency & Emergency preparedness measures Preparedness Measures Failure of a barrier or mitigation measure FIGURE 4-1 Hierarchy of leading and lagging metrics illustrated by the James Reason barrier model (left) and the Pyramid model of incident categories (right). SOURCE: CCPS, 2008. Used with permission. Figure 4-1.eps • Protective Hazards are contained by multiple “barriers” protective barriers HAZARD • Barriers may have weaknesses or “holes” • Weaknesses When holes align, hazard energy or or “holes” chemical is released, resulting in the potential for harm • Barriers may be physical engineered containment or behavioral controls dependent on people Accident • Holes can be latent/incipient or actively opened by people Figure 4-2.eps FIGURE 4-2 Illustration of the Swiss cheese model. SOURCE: CCPS, 2008. Used with permission.

OCR for page 26
29 IDENTIFICATION AND USE OF PROCESS SAFETY METRICS BOX 4-1 Definitions of Tier 1-4 Process Safety Events from API Recommended Practice (RP) 754 A Tier 1 Process Safety Event (T-1 PSE) is a loss of primary containment (LOPC) with the greatest consequence as defined by this RP. A Tier 2 Process Safety Event (T-2 PSE) is an LOPC with lesser consequence. A T-2 PSE is an unplanned or uncontrolled release of any material, including non-toxic and non-flammable materials (e.g., steam, hot condensate, nitrogen, compressed CO2 or compressed air), from a process that results in one or more of the consequences listed below and is not reported in Tier 1: • An employee, contractor or subcontractor recordable injury; • A fire or explosion resulting in greater than or equal to $2,500 of direct cost to the Company; • pressure relief device (PRD) discharge to atmosphere whether directly or via a downstream destructive device that results in one or more of A the following four consequences: —Liquid carryover; —Discharge to a potentially unsafe location; —An on-site shelter-in-place; —Public protective measures (e.g. road closure); —And a PRD discharge quantity greater than the threshold quantity in Table 2 in any one-hour period; or • release of material greater than the threshold quantities described in Table 2 in any one-hour period. A A Tier 3 PSE typically represents a challenge to the barrier system that progressed along the path to harm, but is stopped short of a Tier 1 or Tier 2 LOPC consequence. Indicators at this level provide an additional opportunity to identify and correct weaknesses within the barrier system. Tier 4 indicators typically represent performance of individual components of the barrier system and are comprised of operating discipline and management system performance. Indicators at this level provide an opportunity to identify and correct isolated system weaknesses. Tier 4 indicators are indicative of process safety system weaknesses that may contribute to future Tier 1 or Tier 2 PSEs. In that sense, Tier 4 indicators may identify opportunities for both learning and systems improvement. Tier 4 indicators are intended for internal Company use and for local (site) reporting. SOURCE: API (2010). Note: Table number refers to table in API RP 754, not this report. United Kingdom Health and Safety Executive Metrics that process and that could be monitored to prevent and mitigate the occurrence of a major accident. For In 2006, the United Kingdom Health and Safety example, if the most significant hazard in a plant is the Executive (HSE) published a methodology for estab- overflow or overpressurization of a specific vessel, the lishing metrics, Developing Process Safety Indicators: HSG 254 approach to metrics might call for specific A Step-by-Step Guide for Chemical and Major Hazards lagging metrics that serve as a track record for keep- Industries (HSG 254), based on specific processes.6 ing that vessel within safe operating limits, or leading This document recommended that facilities examine metrics that serve as an track record for maintenance of their specific process details or their particular process the vessel’s instrumentation that measures the content’s unit and develop leading metrics that are specific to level or pressure. 6HSE HSG 254 is available from HSE at http://books.hse.gov.uk/ hse/public/home.jsf and is also included on a CD included with the CCPS book Guidelines for Process Safety Metrics.

OCR for page 26
30 PROCESS SAFETY METRICS AT THE BLUE GRASS AND PUEBLO CHEMICAL AGENT DESTRUCTION PILOT PLANTS APPLICABILITY OF PUBLISHED CHEMICAL AND PROCESS SAFETY METRICS DERIVED FROM PETROLEUM INDUSTRY METRICS TO PCAPP PRIOR OPERATING EXPERIENCE AT CHEMICAL AND BGCAPP AGENT DISPOSAL FACILITIES Elements in each of the CCPS, API, and HSE met- As was noted in Chapter 3, causal factors for process rics publications could be appropriate for the Assem- safety events at other chemical demilitarization facili- bled Chemical Weapons Alternatives (ACWA) program ties were not directly related to the specific process to consider using at PCAPP and BGCAPP. An example used for agent destruction, so it is reasonable to expect would be for the Program Manager for Assembled that the same types of causal factors will be associated Chemical Weapons Alternatives (PMACWA) to define with any events and incidents that may occur at ACWA what is meant by a “process safety incident.” This defi- demilitarization facilities. The causal factors identified nition would be used in developing metrics to measure in Chapter 3 provide an excellent basis for the ACWA the frequency and severity of process safety incidents. sites to develop process safety metrics that relate There are established chemical release thresholds directly to chemical weapons destruction experiences. and associated metrics for common industrial chemi- At 28.4 percent of the total, standard operating cals such as caustic and nitric acid that could be directly procedure (SOP) deficiencies was the most prevalent applied to process safety at ACWA sites.7 However, the causal factor identified. For PCAPP and BGCAPP, physical properties of chemical warfare agents may not metrics that enable early identification and avoidance align well with the release threshold quantities used in of SOP deficiencies should be a priority. Among the the API or CCPS definitions. Perceptions of the toxicity parameters that could be considered are these: of chemical warfare agents require that the Army treat any unintended release of agent as a process safety • ocumenting the percentage or number of pro- D incident.8 cess safety operations and maintenance proce- dures reviewed or revised as scheduled. Finding 4-1. At the present time, there is no defini- • racking revisions to SOPs and documenting the T tion of a process safety incident other than “release of communication of those revisions and training on agent” within the Assembled Chemical Weapons Alter- revised SOPs. • mplementing and documenting job cycle checks9 natives program. Establishing or adopting a common I definition for process safety incidents would improve to ensure that training in roles and responsibilities consistency of reporting and sharing of lessons learned is understood and implemented. Such reviews within the program. evaluate employees’ understanding of SOPs and assess the adequacy of SOPs. Recommendation 4-1. The Program Manager for Assembled Chemical Weapons Alternatives should Human factors, which include “human error,” “mind- adopt the definitions of Tier 1-4 process safety events set,” and “improper technique,” altogether accounted in Recommended Practice 754, Process Safety Perfor- for approximately 37 percent of the causal factors. Again, metrics derived from training activities and mance Indicators for the Refining and Petrochemical Industries, a joint recommendation of the American job cycle checks can be useful in developing actions National Standards Institute and the American Petro- to mitigate these causal factors and to identify areas leum Institute, with the exception that the reporting where regular annual, or more frequent, training should threshold for chemical agents should be defined as any be improved or changed. Other possible considerations unintended release. are these: 7Such metrics can be found in API and CCPS publications and 9Job cycle checks are a formal process whereby a supervisor or in DOT 49 CFR 173.2. his designee assesses an employee’s performance in the field in 8There are intended releases of agent as part of the normal relation to the training he has received on the tasks he will perform demilitarization process, such as when munitions are drained and during the course of his job. This assessment would include the washed out. Any release that is not part of a planned process is an pertinent operations and maintenance tasks. All written procedures “unintended release.” that the employee is asked to follow would also be reviewed.

OCR for page 26
31 IDENTIFICATION AND USE OF PROCESS SAFETY METRICS Finding 4-2. D eveloping metrics for the Pueblo • esults of periodic employee attitude or percep- R tion surveys. Chemical Agent Destruction Pilot Plant and the Blue • requency with which upper managers visit the F Grass Chemical Agent Destruction Pilot Plant based on worksite, or percentage of scheduled visits that operating experience at other chemical agent disposal actually take place. facilities would help to avoid failures that lead to pro- • umber of unresolved recommendations from N cess safety incidents. risk analyses, incident investigations, audits, and Recommendation 4-2. The Program Manager for safety suggestions. • ercentage of near misses and incidents identified P Assembled Chemical Weapons Alternatives should as being caused by unsafe acts or shortcuts. take into account the causal factors in past process safety incidents at chemical agent disposal facilities Equipment malfunction was the second most preva- when devising process safety metrics for the Pueblo lent causal factor (19 percent) noted at other chemi- Chemical Agent Destruction Pilot Plant and the Blue cal demilitarization facilities. While the definition of Grass Chemical Agent Destruction Pilot Plant. equipment malfunction did not include design deficien- cies, it should be noted that design deficiencies caused OTHER PROCESS SAFETY METRICS THAT MAY some equipment malfunctions and that equipment mal- BE RELEVANT TO PCAPP AND BGCAPP functions and design deficiencies were present in over 29 percent of the total number of incidents. At the two Other leading metrics recommended in the CCPS ACWA facilities, this factor could become even more and API documents could also be relevant to PCAPP pronounced because first-of-a-kind equipment that has and BGCAPP. The committee believes that the follow- never been used before will be installed. Conducting ing metrics could be utilized at ACWA sites. design audits and basing metrics on the results could assist in finding design deficiencies before they cause Process Safety Near-Miss Events an equipment malfunction or other process incidents or upsets. A system of process-safety-critical equipment Near-miss events are undesired events that, under inspections is key to minimizing equipment malfunc- slightly different circumstances, could have resulted tions. Metrics-based parameters such as the following in harm to people or damage to property, equipment, could be considered: or the environment. This metric includes the following: • afety-critical equipment inspections could be S • Critical operating limit excursions. This is a assessed, for example, by the percentage of these process parameter deviation that exceeds the inspections completed on time. This metric relies operating limits for critical steps in the process. on the prior identification of equipment deemed The operating limits may be different for the same to be critical to safety. Such equipment might equipment depending on the operating phase. include pressure vessels, storage tanks, piping For example, the required temperatures might systems, pressure relief devices, pumps, instru- depend on the pressure. Troubleshooting efforts ments, control systems, interlocks and emergency should end when the established operating limits shutdown systems, mitigation systems, and emer- are exceeded at critical points in the process, and gency response equipment. predetermined action should be taken to return the • cheduled and preventive maintenance activities. S process to a known safe state. • quipment repair logs. E • Demands on safety systems. This is a demand on a safety system that is designed to prevent a loss Communications deficiencies are not typically docu- of primary containment (LOPC) or to mitigate mented until after a failure but should be considered the consequences of an LOPC. The safety system as integral to the full complement of process safety being activated may be known as a “safety instru- metrics. Audits of communications systems (active and mented system” following the terminology in the passive) and documentation of communications fail- International Electrotechnical Commission stan- ures might be one metric for communication failures. dard Functional Safety—Safety Instrumented Sys- tems for the Process Industry Sector (IEC-61511).

OCR for page 26
32 PROCESS SAFETY METRICS AT THE BLUE GRASS AND PUEBLO CHEMICAL AGENT DESTRUCTION PILOT PLANTS actions would be sampled quarterly or biannually, • Other unanticipated LOPC events. Recognizing and the percent of sampled MOC actions that met all that leaking chemical munitions are a known requirements and quality standards would be deter- concern, the sites may wish to have separate met- mined (CCPS, 2008; API, 2010). rics for leaks or LOPCs that were known to exist Understanding and using MOC leading indicators before munition processing begins and those that requires that the staff operating a facility understand occur during the processing. This would serve to its current operations: the technology, the operational identify aspects of the overall disposal operation knowledge possessed by personnel, and the physical that are increasing risk (CCPS, 2008). specifications of equipment. Implementing a system to manage change must be preceded by a program to Action Item Closure train a facility’s staff in MOC. There must also be a strong process safety culture for MOC to be effective. This metric looks at the percentage and/or number of Particularly in the area of subtle change, the people process-safety-related actions that remain unresolved best positioned to recognize the impacts of any pro- past the date by which they were to have been resolved. posed changes are those directly involved in operating These might include outstanding action items from haz- a facility. ard evaluations, compliance audits, overdue training, or prior incident investigations or drills (API, 2010). Metrics Related to Other Management Systems Completion of Emergency Response Drills CCPS’s book Risk-Based Process Safety contains a number of process safety best practices. It also lists This metric pertains to the number of completed potential metrics for the implementation status of those emergency response drills that use a realistic failure practices. PCAPP and BGCAPP managers could work scenario, completed written records, and completed to identify common process safety management sys- identification and closure of identified deficiencies tem elements that may not be well implemented yet (API, 2010). are important to the safe and reliable operation of the facilities. In light of what they learn, they could review Management of Change the extensive listing of CCPS metrics to find metrics that could be used at their facilities.10 While PCAPP In the area of process safety management, manage- and BGCAPP might not adopt many of these metrics, ment of change (MOC) refers to a specific system that, a number of them could be beneficial to their process prior to the implementation of a change, identifies, safety efforts. reviews, and approves any change to (1) equipment, (2) personnel assigned to the area, (3) raw materials, Finding 4-3. Many process safety metrics that could be or (4) the process technology or operating conditions. used by the Pueblo Chemical Agent Destruction Pilot Another aspect of MOC that must be recognized is that Plant and the Blue Grass Chemical Agent Destruction some changes are subtle. A subtle change might be one Pilot Plant are available to the public, including those involving the supplier of a raw material or a chemical or in the list of metrics in the Center for Chemical Process the rerouting of a pipeline to a different elevation. As an Safety publication Guidelines for Process Safety Met- example of the former, a new material might meet basic rics. These metrics could complement process-specific technical specifications but contain a contaminant that metrics developed at the respective sites. has not been reported to the purchaser but might cause a reaction in the process, with undesirable results. These Recommendation 4-3. The Pueblo Chemical Agent “subtle” changes are often labeled as “not replacement Destruction Pilot Plant and the Blue Grass Chemical in kind.” Agent Destruction Pilot Plant should adopt the met- Each change should be reviewed and assessed for rics listed below and develop process-specific leading its impact on operations and on safety, health, and and lagging metrics. The ACWA program should also the environment. The review should be documented and approved by management and should include any actions needed to move forward, specify responsible 10The CCPS metrics are listed in Appendix I of Guidelines for parties, and set closing dates for action items. MOC Process Safety Metrics (CCPS, 2009).

OCR for page 26
33 IDENTIFICATION AND USE OF PROCESS SAFETY METRICS consider a metric associated with emergency planning • ercent of sampled management of change P and response as well as published lists of process safety instances that met all requirements and quality metrics and should adopt those that appear to be of standards. value to these sites. EXAMPLES OF ACWA PROCESS-SPECIFIC • ount of process safety near-miss events. C METRICS • raining records such as validation of job cycle T checks and completion of training, including The United Kingdom HSE’s publication HSG 254, refresher training. Developing Process Safety Indicators: A Step-by-Step • ob procedures: J Guide for Chemical and Major Hazard Industries, —Statistics on whether a procedure was used and, could be readily applied to PCAPP and BGCAPP if it was, was the procedure the correct one? processes and operations. For example, one hazard at —Validation that procedures are current and PCAPP would be the premature release of contents accurate. from the agent neutralization system before complete • tatistics on the closure of action items. S neutralization (see Figure 4-3). This could be caused • ercent of inspections of safety-critical equip- P by the incorrect sequencing of the neutralization steps ment completed on time. Agent and MPT offgas energetics treatment access Washout water Drained agent Agent Washout storage water Caustic ANR Hot water Sampling tank Agent hydrolysate storage Agent-laden fluids Agent hydrolysate Offgas FIGURE 4-3 Process flow diagram for agent neutralization. SOURCE: Chris Haynes, Blue Grass Design Build Manager, Bechtel Parsons, “BGCAPP MDB intermediate design,” presentation to the BGCAPP intermediate design review executive overview, February 15, 2005. figure 4 -3

OCR for page 26
34 PROCESS SAFETY METRICS AT THE BLUE GRASS AND PUEBLO CHEMICAL AGENT DESTRUCTION PILOT PLANTS or inaccurate readings from instruments, among other Lagging metrics for this portion of the process could things. include records of any near-miss events or process Process-specific leading metrics for this portion of safety incident associated with the agent neutralization the process could therefore include the following: system before complete neutralization—for example, the premature opening of valves, faulty instruments, or • alibration records for analysis equipment asso- C the failure of caustic valves to open. ciated with the agent neutralization system, Another example of a possible equipment-specific including any performance deviations or drift in process safety event is failure of one of the energet- calibration; ics batch hydrolyzers (EBHs) that are to be used at • aintenance records associated with the agent M BGCAPP (see Figure 4-4). The committee has identi- neutralization reactor agitator, including any fied the sudden failure of the drive train, bearings, or changes in vibration or current; and any other aspect of the rotating drum either when filled • raining validation and job cycle checks (see T with energetics that have not been neutralized, or after above) for all operators involved in the operation neutralization when there are still metal parts that need of the agent neutralization system. to be removed, as a process safety risk associated with Vent air to offgas treatment system CCTV Spray bar Sheared rocket parts Vent air Helical flights Face shroud Unload Solids residue transport conveyor Gravity Solids rinsing damper spray bar Liquid level Liquid Solids Drum rotator Solids residue Liquid transport tank (simulated) FIGURE 4-4 Diagram of EBH. SOURCE: John Ursillo, Bechtel, “Process design overview (Blue Grass),” presentation to the Committee to Assess Designs for the Pueblo and Blue Grass Chemical Agent Destruction Pilot Plants, September 22, 2004. Figure 4-4

OCR for page 26
35 IDENTIFICATION AND USE OF PROCESS SAFETY METRICS this piece of equipment. Specific leading metrics for must set performance milestones and must review the this portion of the process could include the following: operation’s performance against those milestones with the organization’s top managers at least quarterly. If • aintenance records associated with the rotating M performance does not meet expectations or the goals drum, including any changes in vibration readings that have been set, the organization must develop an for the main bearing or in vibration or current action plan to rectify the situation so that goals can be readings for the drive train motor(s); achieved. Such performance reviews may suggest that • raining validation and job cycle checks (see T additional or different metrics are needed to help the above) for all technicians involved in the mainte- organization strive for and achieve continuous improve- nance of this equipment; and ment and operational excellence. Reviews of operations • aving an SOP for emptying the EBH if the unit H should be conducted not only at the site leadership level does not function or turn or tip to be emptied. but also above the plant level, including reviews by ACWA off-site leadership. Lagging metrics for this portion of the process could Finding 4-5. A formalized mechanism for a periodic include records for any near-miss events or process safety incident associated with the EBH—for instance, review of process safety metrics by management is bearing failures, motor failures, or drum leaks. an established best practice in industry to verify that These two examples illustrate the concept of devel- management is involved and can drive continuous oping process-specific leading and lagging process improvement. safety metrics. PCAPP and BGCAPP could utilize Recommendation 4-5. The Program Manager for a similar approach to develop such metrics for other equipment that poses the greatest potential for process Assembled Chemical Weapons Alternatives and site safety events. management should perform periodic reviews of pro- cess safety metrics utilized at PCAPP and BGCAPP Finding 4-4. The United Kingdom Health and Safety and implement action plans as appropriate to drive Executive’s Health and Safety Guidance 254 (UK HSE continuous improvements. HSG 254) provides a methodology to develop process- specific leading and lagging metrics. PROCESS SAFETY COMPETENCY Recommendation 4-4. Given that the two facilities The chemical, petroleum, and related industries have are pilot facilities and make extensive use of first-of-a- learned that maintaining a staff of trained process safety kind equipment, the Pueblo Chemical Agent Destruc- professionals is vital to the avoidance of process inci- tion Pilot Plant and the Blue Grass Chemical Agent dents. The focus on this area started in the mid-1950s, Destruction Pilot Plant should review their hazard when the chemical industry experienced a number of assessment documents to identify and consider imple- process safety incidents. Following the serious inci- menting leading or lagging metrics specific to each dents at Bhopal and elsewhere, the chemical industry piece of equipment or area of the plant. These efforts invested heavily in developing process safety expertise should follow the approach outlined in the United in its companies. Following the BP Texas City incident Kingdom’s Health and Safety Executive Health and in 2005, a second wave of hiring safety experts and Safety Guidance 254 (UK HSE HSG 254), Developing building safety competency occurred in many petro- leum and chemical companies. Process Safety Indicators: A Step-by-Step Guide for Chemical and Major Hazard Industries. Such process safety professionals, partnering with senior management, can educate staff and track the performance of key process safety programs such as MANAGEMENT OF BEST PRACTICES OF management of change, the generation and use of good PROCESS SAFETY METRICS IN INDUSTRY SOPs, incident investigations and corrective actions, Managerial leadership encompasses responsibility reporting of near misses and incidents, and process for setting the tone and performance expectations in safety training programs for operators. They review an organization. When process safety metrics are set all these findings with management and propose and for an organization, the operation’s line leadership develop programs as required to address issues. Part-

OCR for page 26
36 PROCESS SAFETY METRICS AT THE BLUE GRASS AND PUEBLO CHEMICAL AGENT DESTRUCTION PILOT PLANTS Recommendation 4-6. The Program Manager for nership with and support from higher levels of manage- ment is essential for the success of these professionals. Assembled Chemical Weapons Alternatives should Examples of metrics for process safety competency are maintain process safety expertise at the programmatic available from CCPS and can be tracked. In addition level to ensure effective implementation of process to tracking the proportion of positions key to process safety metrics. To be successful, process safety experts safety that are currently staffed, other metrics, such as must partner with and be supported by management. completed process safety training and the enhancement Finding 4-7. There are a number of resources that the of process safety competence for relevant personnel, such as managers, supervisors, and technical staff, can Program Manager for Assembled Chemical Weapons be utilized as well. Alternatives can use to learn about best practices for Although the practice of having safety profession- process safety management in the chemical and petro- als is not as widespread outside the chemical and leum industries. Process safety technology conferences petroleum industries, several other industries (nuclear such as the American Institute of Chemical Engineers’ power is one) and government facilities engaged in annual Global Congress of Process Safety and others hazardous processes have also been hiring full-time hosted by organizations such as the Center for Chemi- staff members to develop and monitor their process cal Process Safety and the Mary Kay O’Connor Pro- safety programs. These individuals often participate in cess Safety Center provide ongoing programming on industry conferences in order to learn about the best process safety and the identification of best practices. practices being implemented by other companies, with Recommendation 4-7. The Program Manager for the aim of sharing them with their own facilities and possibly implementing them there. Assembled Chemical Weapons Alternatives should Process safety technology conferences such as the undertake a review of best practices in process safety American Institute of Chemical Engineers’ annual management, especially in the chemical and petroleum Global Congress of Process Safety and other meetings industries. These practices are described in the Center hosted by organizations such as the CCPS and the Mary for Chemical Process Safety book Guidelines for Risk Kay O’Connor Process Safety Center provide ongoing Based Process Safety. Those that are applicable should programming on process safety and the identification be incorporated into the Pueblo and Blue Grass Chemi- of best practices. The CCPS, the API, and the U.K. cal Agent Destruction Pilot Plants. HSE documents discussed above would also provide a starting point for learning about industry best practices REFERENCES for process safety. API (American Petroleum Institute). 2010. Process Safety Performance Indicators for the Refining and Petrochemical Industries. American Finding 4-6. The chemical and petroleum industries Petroleum Institute: Washington, D.C. have found it very beneficial to have employees on CCPS (Center for Chemical and Process Safety). 2008. Process Safety Lead- staff with process safety expertise. These individuals ing and Lagging Metrics: You Don’t Improve What You Don’t Measure. New York, N.Y.: American Institute of Chemical Engineers. partner with senior management and are accountable CCPS. 2009. Guidelines for Process Safety Metrics. Hoboken, N.J.: John for monitoring industry best practices in process safety Wiley & Sons, Inc. and for implementing those that are applicable within UK HSE (United Kingdom’s Health and Safety Executive). 2006. Develop- ing Process Safety Indicators: A Step-by-Step Guide for Chemical and their facilities. These individuals are also tasked with Major Hazard Industries. Health and Safety Guidance 254. Available assisting in embedding process safety into the organi- online at http://www.hse.gov.uk/pubns/priced/hsg254.pdf. Last accessed zation’s culture by organizing and leading grassroots March 11, 2011. process safety teams while reviewing outcomes and metrics with management.