A Modified Baseline Incineration Process for Mustard Projectiles at Pueblo Chemical Depot (2001)

The criteria to be used by the Defense Acquisition Executive to evaluate the modified baseline design and other technologies being considered for Pueblo are spelled out in the NOI for Pueblo (PMCD, 2000):

The committee’s evaluation of the modified baseline process in Chapter 3 is based on these criteria. Chapter 4 covers safety and risk factors and assesses the Army’s interactions with stakeholders concerning the Pueblo site.

The chief components of a safety and risk evaluation are a quantitative risk assessment (QRA), a health risk assessment (HRA), and hazard evaluations (U.S. Army, 1997). The QRA process is conducted in two phases. A Phase 1 QRA is completed prior to facility construction as a component of the EIS. A Phase 1 QRA provides only a point estimate (an assessment without consideration of uncertainty) of public risk from accidental releases of agent. A Phase 2 QRA, which is based on the constructed plant facilities and operations, assesses both public and worker risk from accidental releases and plant upsets. The HRA, which is typically prepared by the Army as part of the RCRA permitting process, considers routine emissions and off-normal operations and assesses all possible human health risks at the site deriving from releases of agent, other emissions, or operating procedures. The HRA and QRAs are complementary in that the former considers normal and off-normal operations and the latter consider accident and upset conditions. Hazard evaluations are detailed assessments of specific operating hazards (including mitigation procedures derived from the QRAs) and are used for managing plant operations safely.

The Army contends that the modified baseline process will be simpler and safer to operate than the baseline system because, among other things, the process will be confined to a one-story building; the LIC furnace and its associated agent drain system and PFS/PAS will not be there; control instrumentation and heating, ventilation, and air conditioning systems should accordingly be less complex; and the DFS, if required, would be located in a separate structure. For the modified baseline process, there will be a requirement for a special accessing procedure (punching), freezing equipment and associated conveyers and holding space, and disposal of spent decontamination solution in the MPF afterburner. Until additional analyses associated with the Phase 1 quantitative risk assessment and the health risk assessment are completed, there are insufficient data by which to quantify the safety of the modified baseline process relative to the baseline system.

A site-specific Phase 1 QRA for a baseline incineration system at Pueblo was prepared and published in 1998 (SAIC, 1998). The causes of potential accidents considered included failures of equipment, human error, and external phenomena such as earthquakes and airplane crashes. Intentional acts, such as sabotage, were not included, nor were nonagent health risks (which will be covered in the HRA). The Phase 1 QRA concluded that the probability of one or more public fatalities from operation of the baseline system is very much lower than the risk of storing the stockpile for 20 years. However, the probability of fatalities at Pueblo under either scenario was estimated to be very much lower than at the other baseline sites (Table 4–1). As Table 4–1 shows, it is much less risky to destroy the stockpiled agents at these sites than to continue storing them. The table also shows that a baseline facility at Pueblo has the

lowest risk among the sites listed because the PCD inventory is entirely mustard munitions, whereas other sites have significant stores of nerve agents. Mustard agents are much less volatile and less toxic than nerve agents, so lethal doses are less likely to be widely dispersed in the event of an explosive disruption. However, because mustard agent is carcinogenic, the Phase 1 QRA included the estimated risk of a cancer fatality from exposure. The risk was found to be extremely low—less than a one in a million chance for one cancer fatality during continued storage or less than a one in a billion chance during processing. The highest risks in the Phase 1 QRA for Pueblo are from external events such as an earthquake or an airplane crash.

A number of studies have been initiated to establish a basis for comparing the risks from a baseline system facility at Pueblo with proposed alternatives (including the modified baseline process) (U.S. Army, 2001b). These studies include risk assessments on (1) the transportation of materials into and out of the PCD; (2) the risks of MAVs versus ONCs; (3) a comparison of a ground-floor storage and unpack area for a modified baseline process with the two-story, ONC-dependent configuration used in the baseline system; and (4) an estimate of the number of leaking projectiles. Additional studies that are planned but have not yet been initiated include (1) the risk analysis for the processing of frozen, undrained munitions in the MPF; (2) the Phase 2 QRA; (3) the HRA; and (4) supporting analyses of safety and hazardous operations.

All these assessments will provide information for comparing the risks associated with the modified baseline process with those associated with other alternatives. The Stockpile Committee previously recommended that the primary criterion for the selection of technology should be the “minimization of cumulative adverse consequences from all relevant risks” (NRC, 1994). Cumulative risks include risks to the public and workers, as well as the economic and schedule risk measures provided in the NOI for Pueblo (PMCD, 2000).

Baseline system operations at Pueblo are estimated to present an extremely small risk to the public—in fact, a much smaller risk than at any other continental U.S. stockpile site. Because the estimated risk with the baseline system is so small, the Army must carefully consider the challenges of attempting to develop a modified process. A modified baseline process, whatever its final configuration, is expected to present a negligible risk to workers and the public, at least in theory. An uncertainty, however, is the behavior of frozen agent-containing munitions introduced to the MPF, which has never been tested. The possibility of plugging of the accessed agent cavity, which could lead to an expulsion of either liquid or solid agent and a consequent spike in the MPF temperature, must be tested and evaluated.

Results of the Phase 1 QRA suggest that the risk to the public from operation and storage at Pueblo is very small. Therefore, a framework for comparison for Pueblo might include an agreement that risks associated with alternative designs of the modified baseline process be considered equivalent if they are confidently below a specified (and small) reference value. This kind of framework would remove the potential for eliminating some viable, low-risk technologies from consideration because of an insignificant difference. Similar risk comparisons must be made for all risks, not just the public risk calculated in the Phase 1 Pueblo QRA. Until a Phase 2 QRA is completed, risk attributes of the baseline system configuration for Pueblo cannot be determined. Based on the values in Table 4–1, the framework described above would probably be applicable only to the Pueblo site.

The discussion and development of the framework for risk comparisons must involve community stakeholders. The Pueblo city and county government and many members of the community are anxious for demilitarization activities to be completed as soon as possible in a way that ensures the safety of all concerned. Although some redevelopment on the PCD site has already begun, full commercial development of the site cannot begin until demilitarization of the PCD stockpile is completed.

Finding 4–1. A Phase 1 QRA for using a baseline incineration system at Pueblo has been completed, and the point estimation of the impact on public health indicates that the risk to the public due to accidental releases would be extremely low. The HRA for Pueblo has not yet been completed. Several additional analyses are being conducted to support operational and design decisions for a modified baseline process. It is not clear how the Army is going to use the collective risk information it has or is seeking and in what framework this information will be used.

Recommendation 4–1. Before the HRA is completed, the Army should work closely with all stakeholders to decide

how the risk of a modified baseline process will be compared with the risk of the baseline incineration system. Such a framework should ensure fair comparison of both configurations as a basis for deciding on the acceptability of the modified baseline process.

Finding 4–2. One specific criterion for evaluating the modified baseline design, as documented in the NEPA notice of intent for Pueblo, is that the facility be “as safe as” baseline incineration. Because of the very low risk at Pueblo, it is not clear how this and other criteria can be implemented.

Recommendation 4–2. The Army should expedite obtaining necessary risk information so interested parties can com-pare the baseline incineration system and the modified baseline process. An important step for ensuring that the necessary risk information is obtained in a timely manner is the establishment of a comprehensive risk management framework. Such a framework would clearly identify the risk measures of interest and reflect the criteria specified in Public Law 105–261 and the NEPA NOI concerning a disposal facility for the Pueblo stockpile.

The modified baseline process is considerably simpler than the baseline system. The optimal modified configuration involves using only one furnace (the MPF) at any time, whereas as many as four different furnaces can be used at one time in the baseline system.¹ If only one furnace is operated, a much less complex control system will be required than for the baseline system. Furthermore (as discussed in Chapter 3), the use of an agent/munition freezing operation may greatly reduce maintenance requirements, resulting in fewer entries by personnel in DPE suits and, thus, fewer opportunities for hazardous exposure. At this time, the committee cannot evaluate the potential risk of introducing frozen munitions to the MPF, because, to the committee’s knowledge, this operation has never been tested. Two other features of the modified baseline process could also increase risk: (1) more agent would be in the MPF at any one time than there was during disposal operations at JACADS and (2) the carbon filters used in the PAS may require replacement over the course of disposal operations, contributing an added source of worker risk. Carbon filters are also being installed at the Anniston, Umatilla, and Pine Bluff facilities, but they were not used at JACADS or TOCDF.

On a qualitative basis, it appears that worker risk in the modified baseline process will be as low as or lower than risk in the baseline system. Quantitative estimates of risk will be provided in the HRA and various hazard evaluations. The committee has stated several times that safety must be the foremost objective in any demilitarization operation involving chemical agent, including operations at PUCDF (NRC, 1996, 1997). Safety should be weighted higher than production rate or cost control.

Training requirements for process workers will essentially be the same for the modified baseline process as for the baseline system. General Physics, the systems contractor for training PUCDF workers, operates the Chemical Demilitarization Training Facility in Edgewood, Maryland, for all CSDP and contractor workers (U.S. Army, 2000c). Initial training takes place in Edgewood, and a simulator for PUCDF operators that will be installed will also be available for ongoing training. Separate training is provided for ammunition handlers and maintenance personnel. A self-paced curriculum is being developed to provide refresher training to control room operators.

Finding 4–3. The modified baseline process as currently configured may be simpler than the baseline incineration system. It may possibly be safer if new operations, such as the processing of frozen munitions, are found not to increase risk.

Recommendation 4–3. Safety should be given the highest priority during the construction, systemization, operation, and closure of the Pueblo Chemical Agent Disposal Facility, regardless of the technology configuration.

Part of the committee’s statement of task for this report (cited in Chapter 1) is to receive briefings from stakeholders in addition to PMCD project personnel and NEPA document preparers. These briefings were expected to augment the information received from PMCD sources and to enable the committee to put the latter information into perspective. The Stockpile Committee has consistently advised the Army on the importance of providing appropriate opportunities for meaningful public involvement in decision making concerning the CSDP, most recently in A Review of the Army’s Public Affairs Efforts in Support of the Chemical Stockpile Disposal Program (NRC, 2000).

Stakeholders that provided input for this study were the PCD, the Chemical Stockpile Emergency Preparedness Program (CSEPP), the Colorado Department of Public Health and Environment (DPHE), the Working Integrated Project Team, the Colorado Citizens Advisory Commission (CAC), and, most importantly, people who live in Pueblo.

PCD is commanded by an active-duty Army lieutenant colonel, who is assisted by a staff of about 150 civilian employees, one of the most active and effective members of

which is the public affairs officer. PCD’s strategic priorities at present are to (Megnia, 2000):

• store the chemical weapons safely and securely

• sustain base operations

• remediate several environmental hazards arising from operations at the PCD since it was established in 1942

• destroy the weapons and residues

• transfer the property and facilities to the reuse authority and close the base

The committee believes that good progress on all of these is being made. There have been no major storage disruptions at the base and only two leakage incidents, one in 1997 and one late in 2000. In both instances, PCD took prompt, effective action to contain them, and no detectable agent was released to the atmosphere outside the affected igloo.

Judging from site visits by committee members, the base appeared to be adequately maintained. Two major remediation projects are proceeding. One is to purge volatile organic compounds (mostly trichloroethylene) from the water table to prevent an underground plume from spreading and contaminating the water supply of Avondale, a nearby community. The other is to clean up explosive residues (including unexploded ordnance) that underlie some areas of the base. Very recently, TNT decomposition products have been identified in the Avondale water supply, and a remediation order from the Colorado DPHE is expected soon.

The PCD commander believes he has good working relations with PMCD, regulators, county government, CSEPP, and the local community. He is willing to work cooperatively on whatever plans evolve.

Construction beginning in 2001 has been approved for infrastructure (steam, electric power, natural gas utility service, and roads) that will be necessary for any of the disposal technologies being considered, with a budget of about $10 million. Plans have been completed for the evacuation of base personnel in the event of an agent release or other disaster, although one road on the base must still be connected to a nearby public road to increase current evacuation routes. A fairly detailed plan for transferring facilities to the Pueblo Development Authority has already been put into effect. Some facilities have already been turned over to the PDA and are on lease to several tenants. The PDA was concerned about the proposed location of the electric power substation and has proposed building and owning the substation to ensure its preferred location. PMCD and the U.S. Army Corps of Engineers have indicated that the substation should be closer to the PUCDF to minimize line losses and to comply with PMCD policies for electric power transmission to chemical agent disposal facilities. The disagreement caused some delay in implementation of the substation project. It was also noted that construction of the personnel support building will be delayed beyond 2001 because of the need to substantially increase the size of the facility.

The Chemical Stockpile Emergency Preparedness Program (CSEPP) for the Pueblo site is housed in a modern control center in Pueblo County facilities that are shared with other county emergency preparedness units. All emergency preparedness units, including CSEPP, report to the county emergency management coordinator (EMC). The heads of both these groups told committee representatives that, except for completing two projects, CSEPP is ready to deal with any emergency at PUCDF. Evacuation routes from the Transportation Research Center immediately north of PCD and from the industrial park at the Pueblo airport still require increased capacity. Also, the installation of 850 indoor tone-alert radios in residences and businesses in the potentially affected area must be completed. Full-scale emergency drills are held periodically, and relations with PCD are good. CSEPP is funded by the Federal Emergency Management Agency with funds provided by DoD.

The Colorado DPHE federal facilities program manager and another staff member met with committee representatives in September 2000. The office of the federal facilities program manager regulates various cleanup efforts at PCD and is the chief environmental regulator involved in the PUCDF project. Interaction with the PUCDF project occurs through participation in an environmental working integrated project team (WIPT); other members include the Pueblo County Commissioner’s Office, EPA, PCD, the Army Soldier Biological Chemical Command, the U.S. Army Corps of Engineers, PMCD, the Program Manager for Assembled Chemical Weapons Assessment (PMACWA), and their contractors.

The goal of the WIPT is to identify and solve problems at an early stage, thus expediting the regulatory activities of the PUCDF project (e.g., approving the infrastructure construction projects) (WIPT, 2000a, 2000b, 2000c). The DPHE manager believes the WIPT process is working fairly well. Initial WIPT meetings in Denver and in Edgewood, Maryland, were not open to the public, but the WIPT meeting in Pueblo on December 6, 2000, was open to the public. Some concerned citizens were present, although public attendance was much smaller than meeting sponsors had expected. Also, the relationship between the DHPE and PMACWA staff appeared to be well established.

A second WIPT, the “Acquisition WIPT,” was set up to provide a single government interface with the contractors and the public throughout the acquisition phase (the design and construction of PUCDF). Members will include PCD, PMCD, PMACWA, the U.S. Army Corps of Engineers, Operations Support Command, and contracting personnel (WIPT, 2000a, 2000b, 2000c; U.S. Army, 2000b).

The CAC “provides a vital link between the Pueblo community and the Army by providing a forum for exchanging information about chemical weapons disposal. It exists to represent community interests” (U.S. Army, 2000e). The nine CAC members are appointed by and serve terms at the governor’s discretion. They include local business people and citizens, a labor leader, an environmentalist, and a representative of Colorado DPHE; a county commissioner presides. The CAC is funded by DoD but is “independent of Army influence” (U.S. Army, 2000e).

At meetings in September and December 2000 attended by members of the Stockpile Committee and PMCD and PMACWA representatives and their contractors attempted to explain the technologies under consideration in the EIS process for the Pueblo site. The technologies and the EIS process are very complex, and some CAC members and members of the audience appeared to be somewhat confused. Personnel associated with the PUCDF project had some difficulty describing clearly the features of the designs and justifications for them, as well as putting levels of risk in perspective. The following concerns were expressed by CAC members about the modified baseline process:

• the multiple handling of munitions if enhanced reconfiguration were employed (this issue was resolved by the PMCD decision to abandon the enhanced reconfiguration for the modified baseline process)

• the ability of the modified baseline process to prevent the release of mercury in gas emissions

• the shipping of energetics for off-site treatment

• the reasons for accelerating the NEPA EIS decision process

Additional meetings are planned to ensure that CAC members understand the technology options. The CAC did not take a position on a choice of technology, but in October 1999, had voted 4–3 (1 abstention, with eight of nine members present) in favor of continuing to pursue the ACWA alternatives. At that time, the county commissioners and city council members indicated a preference for incineration (U.S. Army, 2000b). At the September 2000 CAC meeting, attended by several Stockpile Committee members, five citizens in the audience expressed strong feelings in favor of ACWA alternatives and against incineration. They also expressed dissatisfaction that they were excluded from the WIPT process. Whether others in the community or the CAC agreed or not was not evident, and no one spoke in favor of the incineration option.

A 1999 survey of 1,068 randomly chosen Pueblo County residents revealed that a majority favored incineration over the alternatives (Williams et al., 1999). This conclusion was reconfirmed in a follow-up survey after TOCDF experienced a small release of agent in May 2000.

Based on the committee’s limited contacts, no firm conclusions could be drawn about stakeholder relations. It was evident from comments made at the CAC meetings, however, that the CAC members respect the PCD commander and the depot public affairs officer and were satisfied with PCD operations. It was also evident that some members of the CAC and the public may not understand the merits of the technology alternatives and the risks they may entail.

The CSEPP appears to have emergency management activities under control and to enjoy the trust and respect of Pueblo County officials. CSEPP personnel recognize that more needs to be done to develop evacuation routes and procedures and to complete the distribution of early warning radios.

The WIPT process appears to be reasonably effective and is attempting to work through regulatory issues as they emerge. The process has been viewed with suspicion by some stakeholders, but opening WIPT meetings to the public may allay such suspicions.

Although the PCD commander and staff, along with PMCD, generally enjoy good relations with the Pueblo community, there are apparent communication difficulties that have resulted in lingering confusion about the future. Some misunderstandings seem to have arisen about what would comprise a pilot-scale facility and when it might be installed and tested. The NOI alternatives listed in the NOI published in April 2000 as part of the NEPA process are:

1. A baseline incineration facility,

2. A full-scale facility to pilot test the single-story incineration process,

3. A full-scale facility to pilot test the alternative technology successfully demonstrated by the Assembled Chemical Weapons Assessment Program—neutralization followed by supercritical water oxidation,

4. A full-scale facility to pilot test the alternative technology successfully demonstrated by the Assembled Chemical Weapons Assessment Program—neutralization followed by biodegradation,

5. No action, an alternative which will continue the storage of the mustard agent and munitions at Pueblo Chemical Depot.

Some community leaders apparently believe that installation of a small-scale pilot test of one or more of the ACWA alternatives was promised. However, the wording of the NOI implies that, for at least three options, the choice was a “full-scale facility to pilot test.”

Finally, concerns have also been expressed by some community members about the safe disposal of condensed heavy metal vapor from incineration-based processes and options for energetics disposal.

Finding 4–4. Some community leaders perceive that the office of the Program Manager for Chemical Demilitariza

tion has not been adequately responsive and forthcoming regarding information requests from concerned citizens and CAC members. Presentations to the public describing the modified baseline process have not been as clear or informative as possible.

Recommendation 4–4. The Army should make a greater effort to educate the public about possible disposal processes, as well as the relative risk of continued stockpile storage versus disposal. Army officials responsible for the Chemical Stockpile Disposal Program, in close coordination with the commander of the Pueblo Chemical Depot and other local Army representatives, should strive to maintain and improve open communications with the public, the Citizens Advisory Commission, and interest groups. A Review of the Army’s Public Affairs Efforts in Support of the Chemical Stockpile Disposal Program may be used for guidance. The Army should be particularly responsive to questions and requests for information from local officials and other stakeholders, taking every opportunity to discuss the Chemical Stockpile Disposal Program with the public. All presentations for the public and other stakeholders must be targeted, clear, and of a high professional caliber.