Tooele Chemical Agent Disposal Facility: Update on National Research Council Recommendations (1999)

This report reviews the status of the U.S. Army Chemical Stockpile Disposal Program (CSDP) operations at Tooele, Utah, with respect to previous recommendations and observations made by the National Research Council (NRC) Committee on Review and Evaluation of the Army Chemical Stockpile Disposal Program (Stockpile Committee). The committee recognizes actions that have satisfied recommendations, identifies recommendations that require further action, and provides additional recommendations for improving the overall CSDP performance at the Tooele Chemical Agent Disposal Facility (TOCDF), Tooele, Utah, and other sites. In a 1994 NRC report, Recommendations for the Disposal of Chemical Agents and Munitions, the Stockpile Committee established the following general criterion for evaluating CSDP activities: ''The Chemical Stockpile Disposal Program should proceed expeditiously and with technology that will minimize total risk to the public at each site."

The TOCDF is the first operational baseline incineration system for the disposal of chemical agents and munitions in the continental United States. The facility is adjacent to the Deseret Chemical Depot (DCD), where 43 percent of the total chemical agent stockpile was stored before the start of TOCDF operations in August 1996. Since then, more than 20 percent of the chemical agent stored at the DCD has been destroyed. The Johnston Atoll Chemical Agent Disposal System (JACADS), located about 700 miles southwest of Hawaii, was the prototype baseline incineration system and the first to become operational (July 1990). To date, it has destroyed more than 80 percent of the chemical agent and munitions stored on Johnston Atoll (originally about 6 percent of the total stockpile). Baseline facilities are under construction at three additional storage sites in the continental United States (Anniston, Alabama; Umatilla, Oregon; and Pine Bluff, Arkansas).

Chemical agents are stored at four additional sites. Two of these, Aberdeen, Maryland, and Newport, Indiana, contain only bulk quantities of agent (no munitions).

Chemical-based "neutralization" disposal technologies are being implemented at these sites. The remaining two sites, Pueblo, Colorado, and Blue Grass, Kentucky, contain chemical agent in munitions. Alternative disposal technologies, which are presently undergoing evaluation, may be implemented at these sites. The focus of this report is primarily on operational and planned baseline incineration facilities, especially the TOCDF, but broader programmatic matters, such as risk management and public involvement, are also addressed and are applicable to all CSDP sites.

The major aspects of TOCDF operations reviewed in this report are:

• systems performance and plant operations (Chapter 2)

• trial burn tests to establish compliance with Resource Conservation and Recovery Act (RCRA) and Toxic Substances Control Act (TSCA) emission levels (Chapter 2)

• improvements to the monitoring systems for airborne agent (Chapter 2)

• risk management (Chapter 3)

• safety programs and performance (Chapter 4)

• public and community interactions (Chapter 5)

Findings and recommendations are presented in Chapter 6.

The Army and its TOCDF contractor, EG&G Defense Materials, Inc., have satisfied many, but not all, of the Stockpile Committee's recommendations related to system performance and plant operations. The start-up period has been completed, and operations so far indicate that program destruction goals can be met. Because of TSCA permit delays on the deactivation furnace system (DFS), the Environmental Protection Agency

(EPA) mandated a delay in processing M55 rockets, which has significantly slowed the planned reduction of stockpile storage risk. In the interim, operations were continued on bulk items. Following the successful DFS trial burn in November 1998, the Army was processing M55 rockets at half rate under a RCRA permit limitation.¹ When the TSCA permit was issued in mid-1999, the RCRA limitation was lifted. The TOCDF is now authorized to process M55 rockets at the full rate and is proceeding toward meeting its original risk reduction goals as soon as possible.

Some early operational problems were linked to important safety management issues. These problems, and the investigations necessitated by them, have taken time and management resources that might otherwise have been applied to improving operating performance.

Unresolved issues involving the disposal of dunnage and problems with the slag removal system heater are not critical to continuing safe performance, but their prompt resolution (in the interest of minimizing waste and reducing the number of plant shutdowns for heater replacement) remains a high priority. For example, because it is more economical to ship waste brine off site, the Army has not retested the compliance of the brine reduction area (BRA) with particulate emissions standards. If off-site brine disposal becomes infeasible, this could affect TOCDF operations.

The committee has reviewed and evaluated the results of trial burns conducted on the various incinerators comprising the baseline system. Trial burns were conducted in accordance with RCRA and TSCA protocols. The acceptance criteria for the RCRA trial burn of the liquid incinerators, the DFS, and the metal parts furnace have been met. A second test of the DFS polychlorinated biphenyl (PCB) destruction efficiency showed that emissions levels meet TSCA criteria. The TOCDF has been issued a TSCA permit for the DFS, and activities to obtain a national TSCA permit are still ongoing. Certification of the BRA is not required as long as waste brine is being processed off site. An application was submitted in late 1998 for the RCRA permit renewal, allowing six months for regulators to review the application before the present permit expired in June 1999. At the time of publication, the regulators had completed work with the Army on the permit renewal, but its issuance was pending until the conclusion of a public comment period.

False-positive alarms from the current automatic continuous air monitoring system continue to occur and interrupt agent destruction operations. Although the Army appears to be making reasonable progress in addressing the committee's previous recommendations-including upgrading both the automatic continuous air monitoring system and the depot area air monitoring system—the development, testing, and deployment of more reliable agent monitors should be expedited as much as possible. The development and testing of Fourier transform infrared technology for the real-time detection of an agent release is also proceeding, but real-time alarms are still in development.

The risk management program² uses the health risk assessment (HRA) and quantitative risk assessment (QRA) as quantitative tools to evaluate and manage the

facility risks.³ The HRA for the TOCDF, completed by the Utah Department of Environmental Quality before the start of agent operations, showed that risks were well below regulatory thresholds. However, the data from the TOCDF trial burn indicate that a few compounds were measured at higher concentrations than were assumed in the HRA; the detection limits for others were too high to confirm the validity of the assumed HRA emission rates. Furthermore, a review of some of the models used in the HRA revealed that the HRA did not use the air-dispersion and deposition models and risk assessment methods then recommended by the EPA (i.e., all guidance and updates issued by the EPA through December 1994).

Now that the TOCDF trial burns have been completed, the State of Utah or the Army may wish to issue a brief update of the results of the HRA based on actual TOCDF emissions data and the original EPA guidance. Although the risks posed by individual compounds may change in the updated results, the overall estimate of risk is not likely to change significantly. The committee urges that the results of the revised HRA be made widely available.

Although higher emission rates were found during the trial burn, they would not necessarily significantly change the results of the HRA because the HRAs performed to meet regulatory compliance requirements and HRAs directed toward risk management have different focuses. The former use high-biased assumptions designed to provide realistic bounds but may significantly overstate anticipated effects. The latter use more realistic estimates as a basis for risk mitigation. Hence, significant changes in the emission rates of individual compounds, particularly those that do not contribute significantly to overall risk, may not significantly change HRA results.

The implementation of an effective risk management program at the TOCDF will have important implications for the CSDP as a whole. The Stockpile Committee has made several recommendations in previous NRC reports for improving risk management. In both the 1996 report, Review of Systemization at the Tooele Chemical Agent Disposal Facility, and the 1997 report, Risk Assessment and Management at Deseret Chemical Depot and the Tooele Chemical Agent Disposal Facility, the committee observed that certain aspects of risk assessment and risk management at DCD/TOCDF and throughout the CSDP program required further work and refinement. For example, based on experience from the TOCDF, the committee now recommends that Phase 2 QRAs⁴ for chemical disposal facilities under development be performed as soon as feasible. This will allow risk mitigation measures to be implemented through design changes as necessary.

The committee is pleased with the manner in which the Army has responded to safety issues identified in QRAs. However, risk management continues to be an informal, albeit thorough, process. The committee is concerned that an informal process driven by key individuals in the office of the Program Manager for Chemical Demilitarization (PMCD) could break down with a change in personnel or that the risk management process might not be fully transferred to specific sites. Therefore, the committee urges the PMCD to consider the establishment of a formal management program for QRA-identified safety issues, including a tracking mechanism for identifying new issues and monitoring their resolution.

The committee concluded that another critical aspect of risk management is the change management process (CMP). In this process, effects on risk as measured by the HRA and QRA, as well as public input, are used to evaluate proposed system or operational changes. The PMCD claims that public involvement will be part of changes with a significant impact on risk or changes that are of public concern. The committee believes that public involvement is an important element in the timely disposal of the stockpile—including, but not limited to, the CMP.

The committee strongly believes that the Army should rapidly document and formalize the effective risk

management programs being used on the site-specific and programmatic levels. Cross-communication, cooperation, and learning between sites has greatly enhanced the entire program. The Army must continue and strengthen this process to improve safety and environmental performance.

The Stockpile Committee has been monitoring the CSDP safety performance since its evaluations of operational verification testing at JACADS in 1993 and 1994 and has recommended improvements in the overall management of safety, particularly the development of a well qualified, well trained workforce that operates within an established safety culture. Safety at the TOCDF has also become a public issue because of detailed allegations by two former employees that safety programs and performance at the facility were deficient. As a result of these allegations, seven independent assessments of the safety program at the TOCDF have been conducted. All these assessments reached the same conclusion—that agent operations are being conducted safely.

The Stockpile Committee agrees that TOCDF agent operations are being conducted in a manner that protects the public. Nevertheless, instances of failure to wear required protective equipment, poor housekeeping, and some unsafe working conditions observed by the committee during site visits indicate that a total safety culture has yet to take root at the TOCDF. The recent spill of 140 gallons of nerve agent GB within the containment area caused by the improper reassembly of a filter following maintenance suggests that more training and emphasis on following procedures are needed for maintenance and other operations-related activities.

In response to the committee's observations and recommendations, and out of a stated desire to improve safety performance, TOCDF management has implemented several programs and initiatives to develop and maintain a "safety culture" at the site. Despite these efforts, safety performance has not improved significantly since the agent destruction operations began.

The formal and informal communications about safety that are now issued by TOCDF management on a regular basis have reinforced the commitment to safety and created an environment in which safety is valued. These communications should be continued. The committee is satisfied that some progress has been made toward creating a better environment for the development of a safety culture at the TOCDF. Continued attention to balancing the safety of agent operations and traditional industrial safety issues, as well as continued management involvement and commitment, will be necessary.

The Stockpile Committee's recommendations regarding public involvement in the CSDP and emergency management/preparedness in the 1996 Systemization report dealt only with activities at the TOCDF. The recommendations in the 1997 Risk Assessment and Management report were related to risk management in the overall disposal program.

Since 1996, important changes have been made in the PMCD management of the CSDP, especially in the Public Outreach and Information Office (POIO). After a comprehensive self-examination, the POIO redefined its mission and organization and is no longer the primary point of contact for local public involvement activities for specific sites. Much of the responsibility for site-specific public involvement activities has been delegated to on-site contractors. Although it is still too early to assess the impact of the reorganization and realignment of the POIO, the supporting documentation and goals are much improved.

Although reorganization of the POIO and its goals is important, as is the shift to developing strategies to increase public involvement, neither is a satisfactory substitute for an organizational culture that proactively seeks the involvement of stakeholders and the personnel of the local outreach office. Neither the personnel of the local outreach office nor the public had input into the draft CMP prior to the Army's first public presentation of the process. The committee was disappointed by the CSDP's failure to implement its CMP for any proposed change to the facility. The Army needs to engage the public, not only in changes to already established technology at baseline sites—a topic of limited interest as evidenced by poor public turnout—but also in pending decisions on topics of interest to neighboring communities, such as plans for decommissioning a facility.

Despite improvements in outreach at the local level and the reorganization of the POIO, the committee sees little evidence that stakeholder and public views have been incorporated into the decision-making process. The CSDP has clearly expanded its ability

and capacity for public outreach, but it has not yet achieved the meaningful public involvement the committee recommended.

The Chemical Stockpile Emergency Preparedness Program (CSEPP) has also been reorganized. The Army still controls on-site emergency preparedness, but all off-site responsibilities, including budgeting, have been assigned to the Federal Emergency Management Agency. Consequently, off-site preparations are no longer within the scope of the Stockpile Committee's oversight. The committee remains concerned about CSEPP' s relation to the CSDP and the horizontal fragmentation of responsibility at the federal level. Since the TOCDF became operational, local emergency preparedness activities have intensified and have resulted in some excellent preparedness exercises. The emergency communications system in Tooele County is nearly complete, the decontamination equipment is substantially in place, and tone alert radios are being distributed. At least at Tooele, indications are that the activities of on-site and off-site emergency managers are well coordinated.