Review of Systemization of the Tooele Chemical Agent Disposal Facility (1996)

The Army Chemical Stockpile Disposal Program and earlier chemical demilitarization activities have had a long history. National Research Council (NRC) committees have been involved since 1969. The Army's development of its baseline incineration system for the destruction of the nation's stockpile of lethal unitary chemical agents and munitions has its roots in research and development activities that took place at least as early as the late 1970s and early 1980s.¹ More recently, in 1992, Congress, in Public Law 102-484 (appendix A), directed the Army to dispose of the entire unitary chemical warfare agent and munitions stockpile by the end of 2004. Appendix B contains a brief history of the Chemical Stockpile Disposal Program (CSDP), of NRC reporting on that program, and a description of the principal elements and components of the baseline system.

The two extant examples of the baseline incineration system are the Johnston Atoll Chemical Agent Disposal System (JACADS), located on Johnston Island in the Pacific Ocean more than 700 miles southwest of Hawaii, and the Tooele Chemical Agent Disposal Facility (TOCDF) located at the Tooele Army Depot in Utah about 80 miles west of Salt Lake City. JACADS was established as the first full-scale version of the baseline system and commenced Operational Verification Testing in July 1990. Construction of the TOCDF, the first baseline system disposal facility located in the continental United States, was begun in 1989, and systemization was begun in August 1993.

At the request of the Secretary of the Army, the NRC Committee on Review and Evaluation of the Army Chemical Stockpile Disposal Program (Stockpile Committee) was established in 1987 to provide the Army with technical advice and counsel on the Chemical Stockpile Disposal Program. In evaluating the adequacy of testing and operations at JACADS, the committee produced the following reports in 1993 and 1994 (whose recommendations are summarized in appendix C):

• Evaluation of the Johnston Atoll Chemical Agent Disposal System Operational Verification Testing: Part I and Part II;
• Review of Monitoring Activities Within the Army Chemical Stockpile Disposal Program;
• Letter report to the Assistant Secretary of the Army to recommend specific actions to further enhance the CSDP risk management process; and
• Recommendations for the Disposal of Chemical Agents and Munitions.

Based on the recommendations contained in the above reports, this report assesses the Army's changes and improvements to the safety and operations of the baseline incineration system and the associated monitoring system as implemented during systemization of the TOCDF. In addition, this report reviews the conduct of the Army site-specific risk assessment for the TOCDF. The recommendations in this report indicate actions the Army should take both prior to the start of and during the first year of agent operations at the TOCDF.

At this writing, the start of agent operations at the TOCDF is planned for the first quarter of 1996. Although the Army has the overall responsibility for the decision whether to start agent operations (subject to meeting all regulatory requirements), the Stockpile Committee was requested to provide this report as additional input to that decision. Specifically, the report has been prepared in response to the Army's informal

request to the NRC for a review and assessment of the systemization of the TOCDF.

In developing the findings and recommendations in this report, the Stockpile Committee found it necessary to evaluate the safety and environmental performance of the TOCDF systemization. Safety performance was assessed by direct committee observation and the evaluation of third party information (e.g., published reports prepared by the Army, Army contractors, and state agencies as well as discussions with plant personnel, citizens, and regulators) obtained by the committee. In addition, the committee assessed the Army's Pre-Operational Survey, which was to be used as a basis for making the decision to start agent operations at the TOCDF. Environmental performance was assessed primarily by evaluating the results of surrogate trial burns demonstrating that each component of the baseline incineration system destroys selected compounds to a destruction removal efficiency of 99.9999 percent (6-nines).

The Stockpile Committee continues to monitor the Army's public affairs and community relations program for the Chemical Stockpile Disposal Program. The committee offers observations on the Army's efforts to encourage local community involvement and to work closely with the Utah Citizens Advisory Commission and responsible emergency management agencies on the Chemical Stockpile Emergency Preparedness Program (CSEPP) in the state of Utah.

Finally, the committee offers observations on the Army's response to the committee's risk-related recommendations. The committee describes the Army's plans for risk assessment and risk management at the TOCDF as well as the current status of those activities.

The Stockpile Committee has completed this review of the present status of the TOCDF design modifications and systemization. This review is based on the committee's knowledge of the baseline system, on information provided by the Army and others, and on four site visits to the TOCDF, located at the Tooele Army Depot, Utah. The visits took place in November 1991, March 1993, May 1994 (shortly after the start of systemization), and March 1995 (towards the end of systemization). In addition, four subgroups of the committee made separate visits to the site between February and June 1995. The findings and recommendations presented here are based on information obtained prior to September 30, 1995, before the completion of all requirements that must be met prior to the start of agent operations. The recommendations are based on the committee's knowledge of the current status of the TOCDF and on the committee's understanding of the actions that must be completed before the Army authorizes the start of agent operations. As of this writing, some final activities still require completion by the Army and Army contractors.

Except for reducing the confirmation time for false positive alarms and establishing a system to track laboratory errors, the recommended monitoring system improvements have been made. The reduction of confirmation time is a difficult problem, but efforts are under way to solve it. The system for laboratory operations error analysis was scheduled for implementation by December 1995.

The brine reduction area and the dunnage furnace have been modified and have passed systemization tests at the TOCDF; they will be tested and fully certified shortly after the start of agent operations. The TOCDF liquid incinerator is expected to be capable of meeting nitrogen oxides emission requirements without an additional abatement system.

A hot-slag removal system has been installed on each liquid incinerator. These have not been tested under hot-slag conditions because the Army has decided that there is no adequate surrogate test material for the highly variable slag to allow for complete testing of the system. The Army will test the system as slag accumulates from agent-destruction operations. If the slag removal system does not allow for hot-slag removal, the current manual removal procedure will be resumed.

MITRE Corporation has performed a detailed study (MITRE, 1994) and concluded that changes made to the munitions handling system minimize the chances of a recurrence of the misfeeding problems that occurred at JACADS. The current system is also being analyzed in detail in the risk assessment study being performed by Science Applications International Corporation (SAIC), providing an independent review of the safety of the munitions handling systems.

The Army has developed an improved agent-extraction verification system to handle gelled

mustard and to avoid the need for manual intervention in routine operations.

The TOCDF has organized a program management unit, staffed by the Army, to focus on safety, quality assurance, and environmental oversight. A similar organization exists within the Edgerton, Germerhausen and Grier, Inc. (EG&G) program management organization (the prime contractor for the TOCDF). Environmental oversight includes both permitting activities and environmental compliance. SAIC has been retained by the Army to provide additional staff through an environmental compliance field office. Close working relationships have been established with the Utah Department of Environmental Quality to maintain responsive interactions to facilitate preparation of the Resource Conservation and Recovery Act (RCRA) permit as the TOCDF makes final modifications in response to the systemization experience and other factors. These organizations are charged to maintain compliance with all environmental relations throughout the lifetime of the facility.

The management structures in place at the TOCDF and limited evidence gathered by Stockpile Committee members during site visits suggest that high-quality safety management systems for agent operations are in place, although the committee believes that more attention must be paid to nonagent safety issues to promote a total safety culture at the facility. All action items identified as incomplete in the Pre-Operational Survey must be resolved by the Army prior to the start of agent operations. The Army has initiated a Programmatic Lessons Learned program, created a Field Lessons Learned Review Team, commenced Subject Area Reviews, and established a Risk Management Plan which is being developed by SAIC as an outgrowth of their TOCDF risk assessment work. The Army has developed a process to identify ''precursor events'' at operating facilities (e.g., equipment failures and human errors) and suggest improvements to maintain a continuing emphasis on a philosophy of totally safe operations.

The TOCDF risk assessment has been performed by SAIC under the observation of an outside Expert Panel. One Stockpile Committee member monitored most of the Expert Panel meetings. Because of time constraints, the entire risk assessment will not be completed before the start of agent operations at the TOCDF. However, risk assessment pertinent to each succeeding campaign (i.e., destruction of a particular munition or container type and a particular agent) will be completed and lessons learned will be applied to the facility before the start of each campaign. The first analysis for campaigns 1 and 2 (disposal of GB and VX M55 rockets with co-processing of bulk items) was completed in April 1995. A few resulting modifications to the TOCDF and its operations are being made and will be completed before the start of agent operations. The modified final analysis for the first two campaigns was completed on June 26, 1995. The complete TOCDF risk assessment is expected to be published in the first quarter of 1996. The committee believes that the Army's approach to the TOCDF risk assessment for the first two campaigns meets the spirit of the committee's recommendation.

Capability for positive identification of chemical agent species has been added to the laboratory with the new gas chromatograph/mass spectrometer. This capability is not presently available with field monitors.

The need for continuous monitoring capability for all agents has been met by placing Automatic Continuous Air Monitoring System (ACAMS) monitors calibrated for the various agents in the unpack areas of the TOCDF. A multiagent ACAMS is under development. Continuous monitoring is not yet in place in storage areas, but the Army has told the Stockpile Committee that plans are under way to implement continuous monitoring there.

The need to reduce the time for confirmation of false positives has not been met. ACAMS alarms still require the laboratory analysis of samples from the Depot Area Air Monitoring System (DAAMS) to confirm a false positive. A single unconfirmed alarm requires shutdown of agent operations but does not by itself initiate the response appropriate for a major agent release. False positive signals result in plant disruptions and increase the potential for human error and equipment degradation. The Army expects that the multiagent ACAMS will have a lower false positive rate. (A dual detector ACAMS is expected to be ready for field tests in December 1995).

Procedures for periodic testing of field sensors to eliminate false negatives if a significant release should occur are based on a more comprehensive schedule for testing field sensors. Some monitors have also been installed that can reliably detect higher levels of agent should a significant release occur.

Faster response monitors, set to higher detection levels (immediately dangerous to life and health), have been installed in the unpack area of the facility. A new type of detector (based on ion mobility) with about a 30-second response time is presently being developed.

The TOCDF will undergo a series of required RCRA and Toxic Substances Control Act state-supervised trial burns for the first 24 months of operations; a minimum of 16 trial burns (each consisting of three or four runs) are scheduled. The results of the trial burns will be used to revise the Health Risk Assessment and to formulate a monitoring approach for products of incomplete combustion, particulates, heavy metals, halogenated dioxins and furans, and all volatile and semivolatile organics.

The TOCDF laboratory uses a new bar code system for identifying samples and an improved automated record system. The Stockpile Committee recognizes major improvements in quality control over the procedures at JACADS but reiterates the need for an additional system to track and evaluate lab errors. The additional system would also provide input for continuing improvements to laboratory operations. Such a system is scheduled to be in place by December 1995.

Laboratory personnel are now given a variety of tasks. In addition to analyzing for agents, they now perform nonagent analysis of flue gas samples using mass spectrometry and gas chromatography. Laboratory operators are not presently cross-trained in all of these techniques. However, in the next few months the Army plans to start cross-training in the operation of various instruments, handling hazardous waste, testing new monitors, and performing statistical analyses of data. Cross-training will enhance personnel attentiveness and enhance individual job performance.

The Army has established and implemented a system for providing double blind challenges to the laboratory in a way that will provide frequent and valid checks on the quality and reliability of normal laboratory operations.

SAIC completed the first phase of an accident quantitative risk assessment for the TOCDF, and a summary of the results was provided to the committee on June 26, 1995. This first phase covers the portions of the facility that will be involved in the first two campaigns at the facility (disposal of GB and VX M55 rockets with co-processing of bulk items). The full accident quantitative risk assessment for storage risks and remaining campaigns is expected to be completed in the first quarter of 1996. The techniques used and scenarios considered in the quantitative risk assessment are in accordance with the Stockpile Committee's recommendations. Latent health effects from nonagent exposures are not included in SAIC's scope of work. The State of Utah is analyzing latent health effects independently, following Environmental Protection Agency guidelines for Health Risk Assessments.

SAIC has recognized capability for performing risk assessments but is involved in the Chemical Stockpile Disposal Program because it has been a long-term support contractor for the Army. The Army reports that SAIC was chosen to avoid a long procurement process to select an independent contractor, which would have introduced delays in the program. The Army retained another long-term support contractor, MITRE Corporation, to organize an Expert Panel to oversee the SAIC risk assessment. Five experts were chosen for the panel. The Stockpile Committee has reviewed their qualifications, and they are all well qualified. One of the members is a combustion expert from Brigham Young University in Salt Lake City, who provides some degree of local perspective for the panel. At the request of the Army, the Stockpile Committee was invited to have a member who is an expert in risk assessment monitor the meetings of the Expert Panel. The Stockpile Committee believes that the Army has complied with the intent of the recommendation that an independent risk assessment be made.

Although SAIC briefed the Utah Citizens Advisory Commission in July 1994 on its methodology and plans for conducting the risk assessment, they have not established a dialogue that is perceived by the committee to be interactive with either that group or with other interested local parties. Additional information on local concerns about the quantitative risk assessment was obtained from newspaper articles. The committee believes interactive communication between the risk assessment contractor and local community groups about future risk analyses is essential.

Several risks were identified as a result of the detailed risk assessment, and appropriate changes in equipment and procedures were made to mitigate them.

The Recommendations report reiterated the recommendations from the OVT, Monitoring, and site-specific risk assessment reports. Three of the recommendations pertaining to the TOCDF were related to public involvement, carbon filtration, and program staffing.

The committee finds that the Army's efforts in Utah to obtain community input into the risk assessments were substantial, but not especially productive. The committee believes that such involvement prior to the beginning of a risk assessment, as well as during its implementation, is essential to improving risk communication and gaining public acceptance of the results.

In March 1995, the committee found that the long-delayed approval of personal protective equipment for emergency management responders, associated training, and funding for core response personnel continued to be roadblocks to implementation of the Chemical Stockpile Emergency Preparedness Program (CSEPP) by both the Utah Division of Comprehensive Emergency Management and the Tooele County Department of Emergency Management. At this writing, the personal protective equipment has been approved, but unresolved issues remain pertaining to training in the use of the equipment.

The risks presented by the stockpile require that emergency response plans be completed and exercised to ensure preparedness and successful response should there be an actual release. The Utah Division of Comprehensive Emergency Management has indicated that the absence of national planning standards has meant insufficient guidelines for reentry, emergency medical services, and recovery phase operations, and that this in turn has led to a lack of fully effective training and exercises. Moreover, some counties opted not to participate in exercises, raising concerns about their level of preparedness for an emergency. The lack of national planning standards cannot be permitted to interfere with planning for public safety and, if necessary, the Army must step in to rectify the situation.

The committee finds that the local CSEPP emergency planning efforts are incomplete as is evident by the fact that the Tooele County Emergency Operations Plan appendices are still in draft form.

The committee also finds that the Communication Plan for Tooele County and the planned implementation of the communications system linking the important operations centers in the emergency planning zone are not yet complete. In addition, as of this writing, public notification tone alert radios are not yet in place.

The Army has begun to implement a large and comprehensive public information program in Tooele. This program is impressive because few activities and resources had been devoted to public information and outreach until recently. Yet the list of activities, either planned or under way, still suggests that not enough attention is being paid to soliciting citizen input for programmatic decisions. Additionally, public concerns about emergency management issues are likely to increase when the TOCDF starts agent operations. Efforts should be made by the Army to include the public in discussions of CSEPP issues and to integrate better the public outreach program with elements of the CSEPP program.

The Army is currently evaluating whether the installation of carbon filter equipment for incineration exhaust streams at continental U.S. disposal sites would be warranted. If so, the Army has chosen the TOCDF as the site for a demonstration unit, which would be located on one of the TOCDF incinerator exhaust gas systems. Space for the addition of systems to all incinerator exhaust gas streams has been kept available at the TOCDF in the event that carbon filters are warranted for the TOCDF site-specific conditions. Studies are planned to evaluate whether the installation of carbon filtration systems would be warranted at any of the other continental U.S. sites after the TOCDF evaluation program is completed.

The committee has observed the addition of qualified personnel, both in the office of the Program Manager for Chemical Demilitarization (PMCD) and in contractor organizations at the TOCDF. The present level of staffing at the TOCDF appears appropriate for safe and environmentally compliant operation of the facility. Additional PMCD personnel may be needed as operations expand.

The retirement of the Tooele Chemical Agent Disposal Facility EG&G general manager so close to the

start of agent operations raised committee concerns. A new general manager has since been appointed, and he worked with his predecessor during the months of July and August 1995 to facilitate the transition. The retiring general manager will remain on call as a consultant through the start of agent operations.

Based on the Stockpile Committee's evaluation of the status of the TOCDF with respect to recommendations made in previous reports, the committee is generally satisfied with the progress made and recommends the following actions pertaining to safety and performance be taken at the TOCDF:

Recommendation 1. The development and implementation of the overall safety program at the TOCDF must be given high priority.

Recommendation 2. Safety and environmental performance goals should be given at least equal weight with production goals in establishing award fee criteria.

Recommendation 3. Applicable portions of the accident quantitative risk assessments must be completed and all safety-related concerns resolved before the start of specific agent-destruction campaigns.

Recommendation 4. A substantial effort should be made by the Army to enhance interactive communications with the host community and the Utah State Citizens Advisory Commission on issues of mutual concern (e.g., various elements of the Chemical Stockpile Emergency Preparedness Program (CSEPP), decontamination and decommissioning, future use of the facility, and risk reduction).

Recommendation 5. The Army should increase efforts to work with the Utah Division of Comprehensive Emergency Management to ensure that first-responders have been adequately trained to use the personal protective equipment approved by the Occupational Safety and Health Administration. Tooele County must ensure their capability for responding to an emergency, especially because this condition relates to state requirements for the start of agent operations.

Recommendation 6. The Army, and where appropriate the Federal Emergency Management Agency (FEMA), should ensure that local and state Chemical Stockpile Emergency Preparedness Program plans for responding to potential chemical events are complete and well exercised as soon as possible.

Recommendation 7. The Army/FEMA should provide the necessary resources for completing the communications system planned by the Tooele County Department of Emergency Management.

Recommendation 8. All mandatory requirements of the Army's Pre-Operational Survey must be satisfied.

Recommendation 9. The liquid incinerator and deactivation furnace system must have demonstrated a destruction removal efficiency of 99.9999 percent (6-nines) during surrogate trial burns.

Recommendation 10. High quality, adequately staffed safety management systems must be completely implemented (including procedures for testing critical equipment; all necessary operating, maintenance, and emergency procedures; management of change procedures; training and cross-training programs; programmatic lessons learned activities; subject area reviews; and other safety oversight activities).

Recommendation 11. The liquid incinerator and the deactivation furnace system must pass all required Resource Conservation and

Recovery Act trial burns; the deactivation furnace system must also pass required Toxic Substances Control Act trial burns.

Recommendation 12. Testing and certification of the brine reduction area and the dunnage incinerator should be completed at the TOCDF, or a satisfactory disposal alternative must be implemented.

Recommendation 13. Performance of the slag removal system for the liquid incinerators should be demonstrated when sufficient slag has accumulated.

Recommendation 14. The Risk Management Plan must be fully implemented.

Recommendation 15. A comprehensive, integrated, and clear TOCDF risk assessment study, including a full description of all significant acute and latent agent and nonagent risks associated with disposal operations, as well as with the continued maintenance of the Tooele chemical stockpile, should be completed. A full explanation of the uncertainties associated with the various estimates should be included.

Recommendation 16. A system for documenting and tracking unexpected upsets, errors, failures, and other sources of problems that have led to "near misses" during operation of the facility should be developed as soon as possible. A program for integrating this information into a plan for continual safety improvements at the TOCDF should be implemented.

Recommendation 17. An active program for continual improvement of monitoring instrumentation, including techniques for more rapid recognition of significant levels of agent release, should be pursued.

Recommendation 18. Evaluations of the stack-gas carbon filter bed system should be continued.