Executive Summary

The National Research Council was asked by the Army to form a special, ad hoc committee to investigate whether incidents involving chemical warfare materiel stored, processed, and destroyed at the two operational Army chemical demilitarization sites provide useful information for the safe operation of future sites.1 To discharge its responsibility, the Committee on Evaluation of Chemical Events at Army Chemical Agent Disposal Facilities examined information on all forms of chemical events and incidents that occurred through the summer of 2001 at the Johnston Atoll Chemical Agent Disposal System (JACADS) 2 site in the Pacific Ocean and at the Tooele Chemical Agent Disposal Facility (TOCDF) in Utah. Information on these events was obtained from sources within the government and from a full range of public sources.

The committee concluded that safe chemical weapons disposal operations are feasible at the new facilities scheduled to begin operating at Anniston, Alabama; Umatilla, Oregon; and Pine Bluff, Arkansas, if their management is diligent in setting and enforcing rigorous operational procedures, in providing comprehensive training, in establishing a strong safety culture encompassing all plant personnel, and in absorbing programmatic lessons learned from the first two operational facilities, JACADS and TOCDF. The committee believes that many of the observations and recommendations made in this report are applicable to all demilitarization facilities, including those that may not use incineration. No evidence derived from previous chemical events causes the committee to doubt that the new incinerator technology plants or the disposal processes they will employ can be operated safely and effectively. The committee joins predecessor committees (NRC, 1994, 1997) of the National Research Council that have found that the risk to the public and to the environment of continued storage overwhelms the potential risk of processing and destruction of stockpiled chemical agent.

Recommendation 1. The destruction of aging chemical munitions should proceed as quickly as possible, consistent with operational activities designed to protect the health and safety of the workforce, the public, and the environment.

THE CHEMICAL DEMILITARIZATION CHALLENGE

How can we safely destroy the current U.S. stockpile of chemical weapons within the time constraints imposed by a dangerous and deteriorating stockpile (U.S. Army, 2001d) and mandated by law? Under congressional mandate (Public Law 99-145), the Army instituted a sustained program to destroy elements of the chemical weapons stockpile in 1985 and extended this program to destroy the entire stockpile when Congress enacted Public Law 102-484 in 1992. The stockpile then included more than 31,000 tons of nerve and blister agents deployed in several million individual munitions and containers. In 1997, the Congress reiterated this commitment by ratifying the Chemical Weapons Convention.3

The U.S. Army, through its Program Manager for Chemical Demilitarization (PMCD), began active destruc-

1  

The statement of task is included in the preface.

2  

Johnston Island, southwest of Hawaii, was the site at which the U.S. Army gathered chemical weapons withdrawn from overseas locations. JACADS, the initial stockpile facility, began destruction activities in 1990 and completed processing in November 2000. Planning for closure operations is currently under way.

3  

Formally known as the Convention on the Prohibition of the Development, Production, Stockpiling and Use of Chemical Weapons and on Their Destruction (P.L. 105-277), the CWC requires the destruction of chemical weapons in the stockpile by 2007 and any non-stockpile weapons in storage at the time of the treaty ratification (1997) within 2, 5, or 10 years of the ratification date, depending on the type of chemical weapon or on the type of chemical with which an item is filled.



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 1
Executive Summary The National Research Council was asked by the Army plants or the disposal processes they will employ can be op- to form a special, ad hoc committee to investigate whether erated safely and effectively. The committee joins predeces- incidents involving chemical warfare materiel stored, pro- sor committees (NRC, 1994, 1997) of the National Research cessed, and destroyed at the two operational Army chemical Council that have found that the risk to the public and to the demilitarization sites provide useful information for the safe environment of continued storage overwhelms the potential operation of future sites.1 To discharge its responsibility, risk of processing and destruction of stockpiled chemical the Committee on Evaluation of Chemical Events at Army agent. Chemical Agent Disposal Facilities examined information on all forms of chemical events and incidents that occurred Recommendation 1. The destruction of aging chemical through the summer of 2001 at the Johnston Atoll Chemical munitions should proceed as quickly as possible, consistent Agent Disposal System (JACADS) 2 site in the Pacific Ocean with operational activities designed to protect the health and and at the Tooele Chemical Agent Disposal Facility safety of the workforce, the public, and the environment. (TOCDF) in Utah. Information on these events was obtained from sources within the government and from a full range of THE CHEMICAL DEMILITARIZATION CHALLENGE public sources. The committee concluded that safe chemical weapons How can we safely destroy the current U.S. stockpile of disposal operations are feasible at the new facilities sched- chemical weapons within the time constraints imposed by a uled to begin operating at Anniston, Alabama; Umatilla, dangerous and deteriorating stockpile (U.S. Army, 2001d) Oregon; and Pine Bluff, Arkansas, if their management is and mandated by law? Under congressional mandate (Public diligent in setting and enforcing rigorous operational proce- Law 99-145), the Army instituted a sustained program to de- dures, in providing comprehensive training, in establishing a stroy elements of the chemical weapons stockpile in 1985 and strong safety culture encompassing all plant personnel, and extended this program to destroy the entire stockpile when in absorbing programmatic lessons learned from the first two Congress enacted Public Law 102-484 in 1992. The stockpile operational facilities, JACADS and TOCDF. The commit- then included more than 31,000 tons of nerve and blister tee believes that many of the observations and recommenda- agents deployed in several million individual munitions and tions made in this report are applicable to all demilitariza- containers. In 1997, the Congress reiterated this commitment by ratifying the Chemical Weapons Convention.3 tion facilities, including those that may not use incineration. No evidence derived from previous chemical events causes The U.S. Army, through its Program Manager for the committee to doubt that the new incinerator technology Chemical Demilitarization (PMCD), began active destruc- 3Formally known as the Convention on the Prohibition of the Develop- 1The statement of task is included in the preface. ment, Production, Stockpiling and Use of Chemical Weapons and on Their 2Johnston Island, southwest of Hawaii, was the site at which the U.S. Destruction (P.L. 105-277), the CWC requires the destruction of chemical Army gathered chemical weapons withdrawn from overseas locations. weapons in the stockpile by 2007 and any non-stockpile weapons in storage JACADS, the initial stockpile facility, began destruction activities in 1990 at the time of the treaty ratification (1997) within 2, 5, or 10 years of the and completed processing in November 2000. Planning for closure opera- ratification date, depending on the type of chemical weapon or on the type tions is currently under way. of chemical with which an item is filled. 1

OCR for page 1
2 EVALUATION OF CHEMICAL EVENTS AT ARMY CHEMICAL AGENT DISPOSAL FACILITIES tion of overseas chemical weapons stockpiles at JACADS in militarization facilities (U.S. Army, 1996a; NRC, 1997; see 1990. In 1996, PMCD commenced destruction of the conti- also Chapter 1 and Appendix E). However, given the inher- nental U.S. chemical weapons stockpile at TOCDF, located ent complexity of the chemical demilitarization task at the at the Deseret Chemical Depot (DCD) in Tooele County, assembled weapons stockpile sites, it is almost certain that Utah. The disposal of the stockpile on Johnston Island was new problems will continue to arise, particularly from aging completed in November 2000, and by September 2001 nearly and deteriorating weapons and the challenges of demilitari- 40 percent of the chemical agent at Tooele, the site of the zation plant closure and decommissioning. There will be largest stockpile, had been destroyed. Between these two future “chemical events,” and serious consequences to both facilities, approximately 23 percent of the original chemical plant personnel and surrounding communities cannot be weapons stockpile had been disposed of by the end of the ruled out. summer of 2001. During the 10 years of JACADS operation and the first WHAT ARE CHEMICAL EVENTS? 5 plus operational years at TOCDF, a number of operational upsets or incidents occurred (U.S. Army, 2001c). Some re- Data and Definition sulted in chemical agent penetrating into normally agent- free areas where workers could be exposed. In others, im- To determine the frequency and nature of chemical proper operating procedures in agent-contaminated areas led events at JACADS and TOCDF, the committee requested to actual or potential worker exposure. Further, in a few of that PMCD provide information on all incidents at the two these events, very small amounts of agent were actually re- sites that the Army considered to be chemical events. PMCD leased outside the building into the ambient atmosphere. provided data on 81 separate incidents (39 from June 1990 JACADS and TOCDF are first- and second-generation through December 2000 at JACADS and 42 from August chemical demilitarization facilities based on the disassem- 1996 through May 2001 at TOCDF; see Appendix B) and bly of chemical munitions and destruction of both the chemi- included independent investigation reports for the most seri- cal agent and the associated energetic munitions, as well as ous events. The committee also solicited and received infor- the decontamination of metal containers in a suite of special- mation on actual or suspected incidents from concerned citi- ized incinerators. In 2002 and 2003, third-generation facili- zens, local and state officials, an organization opposed to ties based on the same disassembly and incineration tech- incineration as a disposal means, and current and former fa- nologies are scheduled to commence operation at three of cility employees (see, for example, Appendix C). Much of the largest remaining stockpiles at Army depots in Anniston, this information was gathered during visits to PMCD, Alabama; Umatilla, Oregon; and Pine Bluff, Arkansas. JACADS, TOCDF, and the recently constructed Anniston This report responds to congressional, Army, and pub- Chemical Agent Disposal Facility. lic concerns by: To gain a perspective on the release of chemical agent to the environment during chemical demilitarization activi- • Providing a context for evaluating the significance ties, the committee obtained data from the U.S. Army Sol- of chemical events, dier and Biological Chemical Command on the rate and se- • Illustrating methods for the analysis of chemical verity of leaks from 1990 through 2000 from the chemical events, weapons stockpiles stored at Johnston Island and Deseret, • Analyzing chemical events at the two initial chemi- Utah (U.S. Army, 2001d). cal demilitarization sites as of September 2001, and The committee determined that current Army criteria • Providing recommendations for minimizing and for classifying events at storage and demilitarization facili- managing potential future chemical events. ties are ambiguous and allow the local depot commander latitude to define as a chemical event accidents or incidents that do not involve release of chemical agent. 4 Other inci- Dismantling and destroying chemical weapons is inher- ently hazardous, but the Program Manager for Chemical dents that clearly involved chemical agent were not defined Demilitarization has incorporated extraordinary safety pre- cautions into both plant design and personnel training (NRC, 1996, 1997, 1999a). The chemical demilitarization incin- eration plants are virtual fortresses built to withstand the consequences of accidents, and, to date, releases of chemical 4For example, Army Regulation 50-6, on chemical surety, provides spe- agent from these facilities have been rare, isolated events cific examples of chemical events which the committee judges to be so broad as to invite widely divergent interpretations by local Army depot involving only small amounts of agent, even under upset commanders, such as example number 7: “Any malfunction or other sig- conditions (NRC, 1996, 1997, 1999a). State-of-the-art quan- nificant activity at a chemical demilitarization plant that could reasonably titative risk assessments have determined that the major haz- be expected to cause concern within the local community or the press, or ard to the surrounding communities arises from potential that in the judgment of the local facility or installation management or lead- releases of agent from stockpile storage areas, not the de- ership could cause embarrassment to the U.S. Army” (U.S. Army, 1995).

OCR for page 1
3 EXECUTIVE SUMMARY as chemical events, because the escaped agent remained erative. Actual chemical events should be used routinely to within the plant’s engineering controls. test the completeness of the QRA, which should be routinely For the purposes of this report, the committee deter- utilized to hypothesize the frequency and consequences of mined that a chemical event is any incident associated with chemical events. The Program Manager for Chemical De- chemical demilitarization operations that results in an actual militarization and the U.S. Army Soldier and Biological or potential release of chemical agent. Chemical Command should use the QRAs to evaluate mea- sures to control future chemical events. The Army should Recommendation 2. The Army should establish a consis- also consider using QRAs to examine scenarios associated tent set of criteria to be used by all chemical-agent-process- with sabotage, terrorism, and war. ing facilities to ensure uniformity in the classification of events, and to facilitate event analysis and comparison. MONITORING CHEMICAL AGENT The committee also reviewed the chemical agent moni- Risk Assessment toring procedures at incinerator-based demilitarization facili- The demilitarization facilities contain relatively little ties. It determined that because the monitoring levels used by chemical agent at any one time, and that agent is under strin- PMCD are very conservative and highly protective of worker gent engineering controls in the demilitarization facility. and public health and safety, there are frequent false positive The published quantitative risk assessment for TOCDF alarms, as well as alarms for actual events that pose no mea- (U.S. Army, 1996a) makes clear that by far the greatest risk surable threat to workers or the public. These conservative to the public arises from accidental or deliberate detonation stack-monitoring thresholds ensure that no significant of stored chemical munitions and the accompanying release amounts of agent can be exhausted into the ambient air with- of large amounts of chemical agent to the environment. Al- out the facility alarming and the agent incineration feed auto- though after the events of September 11, 2001, the Army matically terminating. In-plant air breathed by unmasked delayed publication of its quantitative risk assessments for workers and the output of the scrubbing system for air exit- the third-generation chemical demilitarization facilities, the ing the chemical demilitarization plant are monitored at simi- committee has ascertained that the new risk assessments larly conservative thresholds. confirm the dominance of the risk of continued chemical munitions storage. The committee concluded that, in the Recommendation 5. The Army should maintain conserva- post-September 11, 2001, world, the threat of terrorism and tive chemical demilitarization exhaust stack and in-plant air- sabotage would likely be focused in the storage facilities, borne agent exposure thresholds. If current limits for expo- rather than the demilitarization facilities. sure to stockpiled chemical agents are further reduced, the The committee further finds that quantitative risk as- Army should not further reduce existing monitoring thresh- sessments (QRAs) and health risk assessments (HRAs) are olds unless chemical agent monitors can be made both more critical inputs to the dialogue necessary to ensure adequate sensitive and more specific so that lower thresholds can be public involvement in and understanding of chemical de- instituted without significant increases in false positive alarm militarization activities. Maintaining a prudent balance be- rates or unless health risk assessments demonstrate that lower tween the public’s right to know the risks they face and the thresholds are necessary to protect workers or the public. need to protect sensitive information will be an ongoing challenge for the chemical demilitarization program. With- However, the high rate of false positive alarms seems to out adequate risk information available to the public, it will be causing a “crying wolf” mentality whereby some opera- be difficult to develop or maintain the level of public trust tional personnel tend to discount alarms until they have been necessary for PMCD to accomplish its mission. confirmed by laboratory analyses. PMCD must make it clear that properly responding to alarms is more important than Recommendation 3. The Army should continue its prac- production and, at the same time, show that it is trying to tice of making available to the public the results of its quan- solve the underlying problem by actively developing better titative risk assessments and health risk assessments for each instruments. The committee notes that PMCD’s operating chemical demilitarization site. procedures require that all alarms be treated as real until it has been demonstrated by laboratory analyses that they were The committee also found that the QRAs provide a valu- not triggered by real chemical events. able framework for managing the risk from chemical events, including events arising from sabotage, terrorism, and war, Recommendation 6. To reduce the rate of false positive by placing events in the context of their impact on safety. alarms for both airborne and condensed-materials agent con- tamination, the Program Manager for Chemical Demilitari- Recommendation 4. The quantitative risk assessment zation and the relevant Department of Defense research and (QRA) for each chemical demilitarization site should be it- development agencies, such as the Army Research Office,

OCR for page 1
4 EVALUATION OF CHEMICAL EVENTS AT ARMY CHEMICAL AGENT DISPOSAL FACILITIES the Army Research Laboratory, the Defense Advanced Re- Demilitarization should analyze all chemical-agent-related search Projects Agency, and the Defense Threat Reduction incidents at chemical demilitarization plants for patterns of Agency, should invigorate and coordinate efforts to develop causal factors and should institute program-wide actions to chemical agent monitors with improved sensitivity, speci- address the causes found. ficity, and time response. These efforts should be coordi- nated with, and take advantage of, the increased level of in- The programmatic lessons learned (PLL) database com- terest in and increased resources available for developing piled by PMCD is a large undertaking and should help cap- chemical weapons detectors for homeland defense. ture lessons from past chemical events and help prevent the recurrence of similar events. PMCD is to be commended for creating and maintaining the PLL database. However, infor- CHEMICAL EVENTS ANALYSES mation in the PLL database is relatively hard to use and is In analyzing past chemical events, the committee found not prioritized. The data would be more useful if it were that the basic design of the incineration-based demilitariza- organized in a manner that included a system for prioritizing tion facilities and the processes used to disassemble and de- the data. The data may contain patterns that underlie several stroy chemical weapons and to dispose of residue and waste events and that could be found by “mining” the data for these streams (see Appendix A) are fundamentally sound. The connections. This information would improve the capability committee further found that the investigation of chemical for broad generalization of specific information from an in- events and incidents at demilitarization facilities has been dividual incident. straightforward and honest. However, the committee ob- served that future investigations could benefit from the use Recommendation 8b. Any improvements made in investi- of methodologies such as causal tree analysis (where events gation procedures should become part of a systematically are related to the final outcome) and human factors engi- organized programmatic lessons learned (PLL) database that neering (where data on human performance are related to the makes information easier for the non-expert to find and/or causal tree). Such methodologies would result in uncover- use. This can include prioritization and developing a drop- ing and understanding the complete set of those factors found down “tree” list. Lastly, the Program Manager for Chemical to have contributed to each incident. Demilitarization should ensure that, at the plant level, the data are available to, known by, and useful to operations Recommendation 7. Incident investigation teams should personnel. The proposed contractor for the PLL program use modern methodologies of incident investigation rou- should address these issues. For the program to be useful all tinely at all chemical demilitarization sites to help uncover a stakeholders need to buy into its use and structure. broader set of causal and contributing factors, and to enable greater understanding of the interrelationships between and CHEMICAL EVENT IMPACTS among these factors. Experts in human performance modeling should be included on any incident investigation team. A stand- The committee observed that the computer models used ing incident review board at each site should be established to to model accidental chemical releases in Army and local gov- identify chemical events requiring in-depth investigation and to ernment emergency operation centers (EOCs) are represen- ensure that the lessons learned appropriately influence ongoing tative of the state of the art as of the late 1970s. The Gaussian operations. These boards would meet regularly to review acci- plume dispersion modeling techniques embedded in the dents and incidents, including chemical events, and would be D2PC computer model used to predict agent emission plume fully informed of any findings and recommendations made by extent have more current and accurate implementations. chemical event investigation teams. Adoption of more modern and more accurate emission plume models seems to have been delayed by the failure to inte- In its analysis of JACADS and TOCDF chemical inci- grate better plume models into standard Chemical Stockpile dents and events, the committee observed that repeating pat- Emergency Preparedness Program (CSEPP) emergency re- terns of causal factors occurred across the range of incidents, sponse models. from minor to severe. In particular, deficiencies in standard operating procedures (SOPs), design failures, and under- Recommendation 9a. Stockpile sites that still use the D2PC standable, although inappropriate, assumptions (mind-set) of computer model should, at a minimum, upgrade their emer- operations personnel contributed to almost all of the inci- gency response models to take advantage of the improved dents investigated in depth. Repeating patterns of causal capabilities available in the D2-Puff model. Consideration factors in most incidents did not appear to have been used by should be given to testing and possibly optimizing the D2- management to generalize incident findings beyond the im- Puff model at each site by performing tracer release experi- mediate context of each incident. ments under a variety of meteorological conditions. Recommendation 8a. The Program Manager for Chemical Recommendation 9b. The Chemical Stockpile Emergency

OCR for page 1
5 EXECUTIVE SUMMARY Preparedness Program should undertake a continuing evalu- ness Program should establish a stronger capability and ca- ation of alternative approaches to modeling the release and pacity for the coordination of training, equipment, and plans impact of chemical agents. necessary to respond effectively to an emergency incident, and the commitment to do so in a coordinated and coopera- Recommendation 9c. Accurate agent plume dispersion tive fashion. Additionally, the Army should continue its pro- modeling capability should be coupled with timely commu- gram of outreach—including listening to public concerns and nication of results and appropriate responses to the stockpile responding to them, as well as engaging in more conven- site and surrounding communities. tional public information efforts—to both the public and the relevant government oversight agencies to enhance general The committee also determined that communications understanding of the chemical demilitarization program. during and after incidents and events have not always oc- curred as intended between and among the various stake- A major chemical event can result in several months of holders. The lack of an override function or a hot line dedi- lost chemical munitions processing time. Multiple incident cated to notification that an event has occurred has led to investigations and responses have led to additional delays inadequate communication during chemical events. For ex- in restarting operations when incidents have led to plant ample, the lack of notification and warning between DCD, shutdown. All aspects of such investigations and resump- Tooele County, and other Utah responsible agencies was tion of operations should be accelerated consistent with caused in part by a lack of coordination between the Federal safe operations. Emergency Management Agency’s (FEMA’s) CSEPP and the Army’s Emergency Operations Center, and in part be- Recommendation 11. All stakeholders and involved regula- cause of DCD’s prevailing attitude that its emergency man- tory agencies should agree that a single team will investigate agement responsibilities “end at the fence.” This perspec- chemical events requiring outside review. This investigation tive, if carried to other communities where chemical team should comprise already-appointed representatives from demilitarization facilities are to be operated, can endanger all stakeholder groups and agencies, including members of the ability to provide an effective, coordinated emergency the public who can effectively represent and communicate response to incidents. The memorandum of understanding with local officials and the affected public. Incident findings for information exchange recently agreed to by the DCD and should be documented in a single comprehensive report that Tooele County (see Appendix G) could serve as a model for incorporates the findings, proposed corrective actions, and every community with a chemical weapons stockpile, to en- concerns of the various oversight agencies. sure very close oversight of the disposal plant’s operations. ESTABLISHING A SAFETY CULTURE Recommendation 10a. Chemical demilitarization facilities should develop site-specific chemical event reporting proce- The committee believes that the JACADS and TOCDF dures and an accompanying training program that tests and safety programs and performance have been and continue to improves the implemented procedures and communication be adequate to ensure that chemical demilitarization opera- system. tions are being conducted safely. Even so, there is consider- able opportunity for improvement. Many of the incidents Recommendation 10b. The standing incident review board that have occurred at JACADS and TOCDF could have been recommended by the committee for each site should include significantly mitigated—if not prevented—had a true “safety a qualified member of the public who can effectively repre- culture” been in place and functional at the time. sent and communicate public interests. Recommendation 12a. Much of the needed improvement Recommendation 10c. Each chemical demilitarization site in safety at chemical weapons facilities can come from in- should consider the establishment of a reporting and com- creased attention to factors that contribute to and/or cause munication memorandum of understanding (MOU), of the chemical events. For example, the Program Manager for sort developed between the Deseret Chemical Depot and Chemical Demilitarization and chemical demilitarization fa- Tooele County, which specifies reliable and trusted means cility managers should ensure that standard operating proce- of alerting and informing local officials about chemical dures are in place, up to date, and effective, performing haz- events. These MOUs should be designed to permit ready ard operations analyses on new process steps and design evaluation and updating of the terms of the MOU to take full changes even when such changes are viewed as trivial and advantage of learning across the array of chemical demilita- recognizing that chemical hazards are posed by things other rization sites. than agent (e.g., waste). Recommendation 10d. The Army Emergency Operations Recommendation 12b. Management at the Tooele Chemi- Centers and the Chemical Stockpile Emergency Prepared- cal Agent Disposal Facility (TOCDF) and the new third-gen-

OCR for page 1
6 EVALUATION OF CHEMICAL EVENTS AT ARMY CHEMICAL AGENT DISPOSAL FACILITIES eration facilities should develop or identify and implement Recommendation 13. A generous allotment of time should programs that will result in the establishment of a pervasive, be given to training and retraining chemical demilitarization functioning safety culture as well as improved safety perfor- plant operating personnel to ensure their total familiarity with mance. In doing so, TOCDF and the new chemical demilita- the system and its engineering limitations. All plant person- rization sites should draw on experience in the chemical in- nel should receive some education on the total plant opera- dustry, obtained through industry associations or other tion, not just the area of their own special responsibility. The appropriate venues. The Army should revise the award fee extent of this overall training will be a matter of judgment criteria to encourage each new chemical demilitarization site for plant management, but the training should focus on how operator to demonstrate better safety performance than that an individual’s activities affect the integrated plant and its at the older sites. operational risk. Each facility should develop training pro- grams using the newly designed in-plant simulators to present challenges that require knowledge-based thinking. NEW FACILITY START-UP The training programs should include a process for judging The near-term start of operations at the three third-gen- the effectiveness of the training. Including “design” experts eration chemical demilitarization facilities presents an oppor- in the start-up crew for new plants could be helpful in iden- tunity to get these facilities off on the right foot. Plant start-up tifying latent failures in process and facility design. can be a difficult learning experience for new operating crews. It is probable that conditions will arise in plant operation for The committee’s specific findings are paired with the which no SOP has been written. In these situations operators recommendations noted above and presented together in need an in-depth knowledge of their equipment and its limita- Chapter 6 of this report. tions to handle these unusual conditions and maintain plant security. It is common practice in other industries to include “design” people in the start-up crew for new plants.