Review of Proposed Process Changes for Expedited Disposal of the Aberdeen Stockpile of Bulk Mustard Agent (2001)

Advisers to the Nation on Science, Engineering, and Medicine

National Academy of Sciences

National Academy of Engineering

Institute of Medicine

National Research Council

The modified process now under consideration consists of the following:

1. Nondestructive draining of the ton containers. Accelerated disposal will be accomplished by using a process in which ton containers are drained in a glove box type of chemical agent transfer system (CHATS), a system used previously at Aberdeen during research studies, but not for high-throughput processing. The system consists basically of a horizontal support on which the ton container is placed and sealed from the room atmosphere by two expandable seal rings. One plug (inside the sealed area) is then removed and a suction hose is fitted onto the open hole to remove the agent. A second plug is removed and a vent hose is attached. Again, this design is not new; it has been used by the Army for several years. The difference for the expedited disposal process will be that a relatively large number of containers will have to be processed on a relatively compressed time schedule. Because of this change, the Army will have to pay careful attention to the human factors aspect of the CHATS design. The process must be able to accommodate frozen plugs.

   Drained agent is subsequently destroyed by on-site neutralization (hydrolysis). Draining of the containers is followed by their triple rinsing and return to the storage area. This rinsing step uses water only and will not necessarily remove all agent and sediment that may be left in a container. If there is any indication that agent still exists in the ton containers after rinsing, caustic will be added before they are placed in storage for final cleaning and disposal.

2. Destruction of the chemical agent drained from the ton containers, as well as any agent present in the container rinse water, by on-site neutralization in existing tank reactors. This portion of the process is essentially the same as the design under construction—i.e., the reaction is the same and the agent storage and neutralization tanks are the same as those planned for the original process.

3. Disposal of the hydrolysate through off-site posttreatment at a commercial facility. The Army, as a part of the original design, was considering off-site treatment. Preliminary talks have been held with two types of TSDFs, one a nonincineration treatment facility and one an incineration facility. The composition of the hydrolysate is known to the disposal facilities, which have tentatively indicated they can accept it.

4. After triple rinsing, disposal of the ton containers through a ton container cleanout (TCC) process and shipment of the container halves to Rock Island, Illinois. The TCC process is essentially the same as that previously intended for Aberdeen.

This modified process differs from the facility design currently under construction at Aberdeen in several ways (NRC, 2000), including the following:

1. The agent will be removed from the ton containers in CHATS glove boxes that are located in the existing (new) warehouse building at Aberdeen. This building will have an air flow system with pressure levels different from those contemplated in the original design. The building is not as robust (hardened) as the building in the original design, and the airflow is not as firmly segregated between levels. For example, the ton containers in the CHATS are separated from the rest of the room space of the building by two bladder-type seals. As in the original design, the air from the building is discharged through activated carbon. Also, all spill control requirements for the building are to be met. Mustard is the only agent that will be processed, and because it has a relatively low vapor pressure, the change in building design described above is not critical.

2. Any necessary thawing will be done prior to delivery to the facility. However, limited information is available on how this will be accomplished.

3. The ton containers are delivered to the disposal facility in essentially the same manner as originally proposed, but are drained to one or more neutralization tanks using a vacuum system rather than a punch-and-drain method.

4. The triple-rinse step is done through the same opening used in the draining step. This step may not remove all agent or sediment from the containers, in which case caustic will be added to the containers when required. Complete decontamination of containers is to be done in the ton container cleanout (TCC) process, which is not a new step.

5. The controls provided are similar to those used in a batch pilot plant and include a data acquisition package.

6. The hydrolysate will be sent off-site to a commercial posttreatment facility rather than treated on-site by a biological process.

Reduction of the time needed to destroy the mustard agent stored at Aberdeen will also require several activities for expediting the environmental permitting process. The Army’s intent is to meet all of the control and monitoring requirements that currently apply but to drastically shorten the time needed to obtain the necessary permits. As presented by the Army, this approach will require a National Environmental Policy Act (NEPA) exemption; a Resource Conservation and Recovery Act (RCRA) emergency permit, along with an expedited RCRA delisting of mustard agent HD hydrolysate;¹ and an expedited Clean Air Act (CAA) administrative permit modification (Flamm, 2001). These issues have not yet been discussed with the appropriate regulatory authorities. The Army proposes to initiate these and all outreach activities after receiving formal approval to proceed with this approach.

The Army is currently pilot testing various key components of the modified process. Studies are under way to determine agent-loading parameters, to confirm drain station performance, and to demonstrate the ton container triple-rinse operation. These pilot studies are expected to be completed by January 2002. The loading of the mustard agent in the neutralization step has not been finalized. The original design calls for 3.5–8.6 percent agent loading, but if the hydrolysate is to be incinerated off-site, a higher organic concentration will be preferable. There appears to be an upper limit of about 17 percent agent loading (based on laboratory data), at which point phase separation occurs. The hydrolysate loading part of the ongoing pilot program will be completed by December 2001. The concept of using a triple-rinse operation is derived from RCRA regulations for cleaning drums that contain hazardous wastes. The ability to clean the ton containers using the proposed triple rinse has not been confirmed. There is a real possibility of solids and/or gelled agent remaining in the containers after draining. There is also the possibility that any sludge, whether remaining in the containers or removed during the rinse step, may contain metals. The pilot program is aimed at addressing these concerns.

The use of 18 MINICAMS, including four spares and two mobile units, is proposed for monitoring agent concentrations in the air. The 12 MINICAMS will be located in the warehouse (2), neutralization cubicle (1), drain stations (3), airlocks (3), filter (2), and ton container staging area (1). The Army should consider a means of providing more complete coverage. As was the case with the original design, no perimeter monitoring is contemplated. The control limits for agent concentrations are the same as those currently in place (see Federal Register, Volume 53, Number 50, March 15, 1988, pages 8504–8508).

To date, no hazards analyses, in particular none with a focus on operations, have been completed for the modified configuration proposed. To accelerate the design steps, the Army is proposing to use a process whereby the hazards analyses can be done concurrently with the final design rather than in steps, as is the more common practice.

As indicated above, meeting the accelerated schedule requires several actions to expedite the permitting process. The major change will be minimal or no opportunity for public input (hearings) concerning the new proposed process for expedited agent disposal and various components of the process, including monitoring. The Army intends to meet the intent of the regulations and to maintain an equivalent degree of control on preventing release of agent. Preliminary modeling indicates that the one percent lethality distance² will be contained within the fence line of the disposal facility site as currently proposed, and that the concentrations of agent at points of human exposure will be below levels acceptable according to the 1988 regulations noted above (i.e., the time-weighted average exposure limit for workers, general population limit, and so on).

The approach proposed reflects a considerable departure from the current philosophy of providing the maximum degree of protection to the operators and to the public. The building where agent operations occur is not hardened, and the ventilation system has fewer levels of separation between contaminated air and work areas. The Army is proposing that workers not use demilitarization protective ensemble (DPE) suits (Level A protection), but use commercially available protective gear at Levels B, C, and D, depending on to the operation. According to the Army, this source of an apparent increase in risk (which may be small, but was not quantified) will be offset by the more rapid destruction of the agent stockpile and, presumably, a reduction in total risk over time.

The Army has proposed a very ambitious schedule for disposal of mustard agent at Aberdeen, with a significant number of critical milestones and decision-points. So that the schedule can be met and the plant operated successfully with as few as possible events resulting in as little as possible release of agent, it is critical that a single team be created that can function in a seamless manner. In similar situations in industry, the pilot plant staff, design engineers, operations personnel, health and safety experts, and permitting experts all have clear means of communicating frequently in the planning and execution of the entire program. This typically is done face to face to ensure the needed level of trust. Also, during actual operation of the facility, more stringent supervision, especially during start-up, may be required. Federal contracting regulations applicable to the Aberdeen site are not conducive to this type of operating environment because the kinds of expertise mentioned above reside in a variety of organizations within the government and with private contractors.

No details concerning outreach activities for the proposed modifications were provided by the Army, which said only that an outreach program will be started upon approval to use the modified approach. This approach, which may provide less protection to the public, will require a well-thought-out public discussion.

Finding 1. The ad hoc panel supports the Army’s proposed concept for expediting the disposal of mustard agent at Aberdeen. The chemical reactions utilized in the modified alternative technology process have been essentially proven from existing data and there appears to be little likelihood that the mustard agent will not be neutralized. The degree of agent loading has to be carefully determined to avoid phase separation. Whether the triple-rinse operation can remove most of the agent and/or sediment from the containers is currently uncertain.

Finding 2. Outside of engineering controls, the degree of protection available to prevent release of agent is not as high in the proposed modified process as it is in the original design, because the building (a warehouse in this case) is not hardened and the air-handling system is less robust. That is, the plant air ventilation is a once-through cascade system without the redundant blowers and filters normally found at chemical agent disposal facilities. The air monitoring appears approximately equivalent to that in the original design for the Aberdeen facility, but the ad hoc panel believes that additional monitors will provide a higher level of safety since the building is less robust and has fewer levels of containment by which to control contaminated air.

Finding 3. The ability to shorten the schedule for agent disposal appears to be reasonable if the data collection under way confirms the validity of the underlying assumptions; nonetheless, the regulators must agree to the expedited permitting. Not enough information was available to the ad hoc panel to allow assessing either of these items. Meeting the accelerated schedule will require tight (seamless) control of all elements of the project, from pilot studies through construction and operation.

Finding 4. The analyses required to confirm agent destruction in the hydrolysate at the higher loadings now being contemplated have not been finalized. To dispose of hydrolysate off-site, treatment, storage, and disposal facility personnel, as well as the appropriate regulators, will have to agree on the analytical protocols, including sampling frequency, sampling procedures, and detection limits for all media.

Finding 5. Hazards analyses, which will identify the areas of greatest risk to workers or the public, have not been completed. Their completion and utilization in developing final equipment designs and operating procedures are critical to preventing and/or minimizing the occurrence of any events whereby agent could be released.

Finding 6. It appears that the chemical agent transfer system (CHATS) for agent removal will require greater worker involvement than will the automated handling system for agent removal from ton containers developed for currently operating chemical agent disposal facilities. This change will require close analysis and control so that worker safety and process integrity are not compromised. In developing a safety program and worker protection procedures, the Army should carefully consider the human factors involved in the procedures.

Recommendation 1. The Army should complete as quickly as possible its experimental program to evaluate the proposed modifications for expedited disposal of the Aberdeen stockpile, giving particular attention to the draining (via the chemical agent transfer system) and cleaning of the ton containers.

Recommendation 2. The Army should provide more monitoring stations as one means of assuring workers and the public that the proposed new approach to agent disposal at Aberdeen is as safe as the original approach to disposal at Aberdeen.

Recommendation 3. The Army should create a tight, well-functioning team of design, safety, permitting, and operations personnel who are responsible for the accelerated disposal process from the pilot stages through the completion of disposal operations.

Recommendation 4. Required hydrolysate analysis protocols should be finalized as soon as possible.

Recommendation 5. The Army should conduct hazard analyses, such as focused quantitative risk assessment (QRA) evaluations, as soon as possible and update these analyses as required, using new information as it is received. The focused QRA evaluations should take advantage of existing risk evaluations and focus on potential operational risks. These analyses should be made part of an ongoing risk management process that will span both the design and operational phases of the facility.

Respectfully yours,

Peter B.Lederman, Ph.D.,

Lead

Ad Hoc Panel on Review of Proposed Process Changes for Expedited Disposal of the Aberdeen Stockpile of Bulk Mustard Agent

Flamm, Kevin. 2001. Briefing to the ad hoc panel of the Stockpile Committee on November 26, 2001, by Kevin Flamm, Project Manager for Alternative Technologies and Approaches.

National Research Council (NRC). 1996. Review and Evaluation of Alternative Chemical Disposal Technologies. Panel on Review and Evaluation of Alternative Chemical Disposal Technologies. Washington, D.C.: National Academy Press.

NRC. 2000. Integrated Design of Alternative Technologies for Bulk-Only Chemical Agent Disposal Facilities. Committee on Review and Evaluation of the Army Chemical Stockpile Disposal Program. Washington, D.C.: National Academy Press.