Review and Evaluation of Alternative Chemical Disposal Technologies (1996)

This chapter discusses the factors that the AltTech Panel considers central to evaluating and comparing the alternative technologies. The factors included here were developed from the panel's review of the Army Criteria Report and the NRC Criteria Report Evaluation , from the concerns and issues raised in public forums conducted by the panel in communities near the two sites, and from the combined expertise and experience of panel members.

The AltTech Panel has essentially adopted three of the four primary factors identified by the Stockpile Committee in the Criteria Report Evaluation: process efficacy, process safety, and schedule (NRC, 1995, pp. 14-19). The fourth factor, cost, will be evaluated independently by the Defense Acquisition Board. In adopting these factors, the AltTech Panel modified the wording of the first two factors (modified portions are shown in italics):

1. Process Efficacy. Does the alternative agent destruction process, when integrated with other necessary destruction system components, effectively and reliably meet agent destruction requirements?

2. Process Safety. Is the alternative technology safe and does it protect public health and the environment? The criterion of "safe" adopted by the Stockpile Committee is minimization of total risk¹ to the public and to the environment (NRC, 1994b).

3. Schedule. What are the impacts of implementation of an alternative technology on the schedule for stockpile destruction?

Each primary factor has several subfactors, which may be interdependent. A negative judgment on a technology for a specific subfactor need not imply a negative overall judgment for the primary factor. The subfactors and their interdependencies are discussed below.

Process efficacy encompasses not only the capability of a technology to destroy the agent of interest but also the status of the technology: its stage of maturation along a spectrum from laboratory-scale to pilot plant development and eventual full-scale operation. Process efficacy also includes whether the process can be controlled, whether it is reliable, and whether it meets applicable regulatory and treaty requirements. The AltTech Panel has defined the following subfactors under process efficacy:

• technology status
• capacity to detoxify agent
• satisfaction of treaty requirements
• satisfaction of environmental and other regulatory requirements
• management of process residuals²
• process stability, reliability, and robustness
• process monitoring
• natural resource requirements (e.g., energy)
• scale-up requirements
• applicability for treating other wastes

By the status of an alternative technology, the panel means the stage to which the technology has progressed toward fully operational practice. In general, chemical-process technologies can be located along a developmental continuum from laboratory-scale, proof-of-concept testing to pilot plant demonstration and ultimately to full-scale operation.

Many considerations are involved in determining whether a technology is ready to move to the next stage or how close it is to being "successfully demonstrated" at a given stage. For instance, at the laboratory-scale, assays and chemical analyses are important in establishing that the desired reactions predominate and that unwanted side-reactions can be controlled. At the pilot-scale, precise mass and energy balances become essential, along with quantitative characterizations of how key process variables affect outcomes. The documentation for a pilot design must be complete enough for a preliminary assessment of risks related to the hazard inventory (e.g., agent concentrations at each process step, reactive materials, pressure) and the adequacy of safety features, such as process interlocks and safe means of releasing excess material or energy. Assigning a status to a technology is, therefore, not a simple classification but rather a running checklist of what has been accomplished to date and what remains to be done.

In assessing the status of a technology, the AltTech Panel had to consider the extent of documentation and evidence provided, as well as the capabilities, resources, and commitment of the TPC. These company-specific characteristics are critical to the successful implementation of any technology, both at the demonstration stage and during disposal operations.

To detoxify a chemical agent such as VX or HD satisfactorily, the reaction that destroys the agent must proceed until the remaining concentration of agent is below a specific limit. The Army specifies this limit in terms of a "destruction removal efficiency" (DRE), defined as the difference between the amount of agent going into the process and the amount remaining, expressed as a percentage of the amount going in. For a process to be acceptable in destroying agent, it must have a DRE of 99.9999 percent or greater. DRE values are often expressed as the number of 9's in the percentage; this DRE is therefore referred to as "six 9's." A DRE of 99.999999 percent is "eight 9's."

In addition to the required DRE for a destruction process, the Army uses the following limits on allowable concentration of agent to determine whether a material must continue to be controlled as (potentially) agent-contaminated, may be released from an agent-control facility for further treatment, or may be released to the environment or to general, "public" use (i.e., any use other than for further treatment to destroy residual agent).

The release of gases to the atmosphere is constrained by a health-based General Population Limit at the site boundary. The limit values for HD and VX are, respectively, 0.1 and 0.003 g per cubic meter of air.

There is no standard established for unconditional release of liquids containing chemical agents. The standard for release of certain specified liquid wastes from incineration facilities to qualified disposal facilities is 200 ppb for HD and 20 ppb for VX. These same limits apply to release of drinking water to soldiers in the field.

The Army has three primary classifications for solids that may be contaminated with chemical agent. The first classification is for solid material that is potentially contaminated and has not been subject to further decontamination or testing. This material cannot be released from agent-control areas under Army supervision. The second classification, called "3X," is for solids that have been decontaminated to the point that the agent concentration in the air above the solid does not exceed the health-based 8-Hour Worker Limit. The limit values for HD and VX are, respectively, 3 and 0.01 g per cubic meter of air. A 3X material may be handled on an unrestricted basis by plant workers but is not releasable to the environment or for general reuse (i.e., not releasable "to the public."). In specific cases in which approval has been granted, a 3X material can be shipped

to approved hazardous waste treatment facilities for landfill disposal. The third classification, called "5X," is for material that has been subjected to thermal treatment of at least 1000°F for 15 minutes to assure essentially complete destruction of all residual agent. A 5X material is releasable to the public.

For this study, the TPCs conducted laboratory tests under Army supervision to determine if the technologies would, in fact, destroy agent. The panel received results of these tests in late June 1996. Although the overall results demonstrated that all the technologies can destroy agent, quantitative data on process residuals were not available to the panel in time for in-depth review. Careful consideration of process residuals will be required for decisions about pilot-testing.

The 1993 Chemical Weapons Convention (CWC) requires destruction of the primary agent and further reaction or destruction so that none of the end products can be readily converted back to the primary agent. (An appendix to the CWC treaty contains a list of compounds that can be readily converted to the agent; these compounds are called "scheduled precursors.") The CWC objective is to remove the military threat from agents, whereas environmental permits are designed to protect human health and the environment. Therefore, the requirements for residual concentrations of agent allowable under treaty negotiations are likely to be less stringent than the requirements under environmental permits for destruction facilities and downstream disposal facilities.

The CWC requires that the destruction system allow for verification that agent has been destroyed. The convention further requires that the destruction of the unitary chemical weapons stockpile be completed within 10 years after the treaty is ratified (ratification was expected in 1996).³

The agent destruction process that is implemented must comply with state and federal regulatory requirements. Key regulatory requirements include specifications for acceptable process residuals and wastemanagement practices. Other regulatory compliance issues include workplace safety and health requirements (e.g., those set by the Occupational Safety and Health Administration) and management of nonprocess wastes, such as decontamination fluids and personal protection equipment.

Disposal of process residuals is a critical aspect of any agent destruction system. The process residuals from alternative technologies differ in physical state, composition, and quantity, but all residuals must ultimately be dealt with. The toxicity of reaction products must be low enough that unwanted process residuals can be managed through aqueous discharge to a conventional wastewater treatment facility, disposed as solid waste in a landfill appropriate for the toxicity of the waste, released as allowable atmospheric emissions, or some combination of these three release routes. In legal terms, the concentrations and toxicities of the materials in aqueous, solid, slurry, or gaseous residual streams must fall below the limits set by the environmental permits needed to operate the agent destruction facility and any downstream waste-management facilities.

One major challenge with some technologies is the management of large quantities of aqueous residuals. On-site management of aqueous residuals requires deciding either to change Army regulations to allow discharge directly to a wastewater treatment facility or to continue to evaporate the water and discharge it as an atmospheric emission, as is done in the baseline system. (The residual material remaining after evaporation is treated as a solid-waste stream.) Some of the hydrogen atoms originating from the chemical agents will ultimately bond with oxygen to form water so that, even with aggressive water recycling, some form of water release will be required. The extent of water recycling will affect cost.

A second major issue is the point at which process residuals can be transferred to off-site, private sector facilities for subsequent management. This question

requires consideration of appropriate waste management options (aqueous discharge, solidification or stabilization, landfill disposal, thermal destruction, etc.) for individual waste streams, the capability of private sector facilities to meet regulatory requirements and to process residual waste streams, the criteria for releasing process residuals to the private sector for treatment or disposal, and the technological capacity of available private sector facilities.

The process residual streams from alternative systems need to be compared in terms of both the composition of the stream and the intended management of it. An appropriate basis for this comparison begins with the mass balances for the overall process and for major chemical elements, such as nitrogen, sulfur, chlorine, phosphorus, and carbon. (Mass balance data that were available to the panel are summarized in Chapters 4 through 8.)

Process stability, reliability, and robustness are key goals. Achieving them depends on many factors, a few of which are described here.

The batches of agent fed to a destruction process will vary in agent purity and in the composition of impurities as a result of variability in the conditions of their production and storage. For example, some containers of HD contain solids, which may make them difficult to feed through a system designed to handle liquid agent. The process must function effectively and reliably in spite of such variations in the process feed, i.e., the process must be sufficiently reliable that it can effectively destroy agent despite a range of variability in the chemical and physical composition of the feed material.

Operating conditions that can result in process instabilities, such as temperature or pressure excursions that can lead to catastrophic failure, must be avoided. Such conditions can include extreme operating conditions (e.g., high pressures, temperatures, or reaction rates) and corrosive reactants, residuals, or process environments.

Control strategies and process flexibility must permit the process to be controlled effectively even in the event of an upset such as a power failure or loss of agitation. The selected process must also provide for the decontamination and management of storage containers and other contaminated metal parts.

Implementing an alternative technology requires techniques to monitor the concentrations of agent and of reaction products in liquid, slurry, or solid process streams. Sampling procedures, response times, and required detection limits must be defined. The monitoring requirements for alternative processes may be quite different from the requirements for the baseline system. A critical issue is whether new monitoring techniques, not commercially available, are required, and if so, what the schedule for developing these techniques would be.

The consumption of resources such as energy and water must be considered in selecting a technology, especially for locations where these resources may be limited. Resource constraints do not appear to be an issue at either Aberdeen or Newport, but a high demand for power or water, for example, may have secondary effects that need to be understood.

Implementation of an alternative technology will require demonstrating the process with near-full-scale equipment prior to full implementation. The equipment required to demonstrate a process may differ for HD and VX. In addition, the scales at which the technologies under consideration have demonstrated the processing of agent are quite different, as is the scale at which these technologies have been used for other applications. Consequently, the engineering development required to scale-up the process will differ for each technology.

Use of an alternative technology that is broadly applicable to treating common industrial wastes (including hazardous waste) is a concern to some in the communities near stockpile sites who fear the facility could be readily converted for treating additional wastes imported from off-site, once stockpile destruction is completed. Thus, selection of a technology that would result in a versatile waste destruction facility may increase fears that the facility will not be decommissioned after

the stockpile is destroyed. A contrary view, also held by some members of the communities, is that versatility could be a virtue at a site such as Aberdeen, which contains numerous hazardous wastes, other than the unitary agent stockpile, that also require disposal.

Process safety encompasses concerns about worker safety, community health risks, and environmental protection. Evaluating process safety therefore includes assessing in-plant safety and health risks, risks to community safety and health, and risks to the environment. For each of these major risk categories, the evaluation should include the consequences of a release of chemical agent and of nonagent, toxic process residuals. Important contributing factors to the overall risk in each category include the risks from storing and handling agent in containers prior to processing, as well as the risk of releases from the destruction process itself.

The discussion below covers, in broad outline, the full range of risk factor evaluation and of risk assessment, preliminary and quantitative, that must be done in the course of developing an alternative technology through pilot-testing and on to construction of a full-scale operational facility.

For this particular study, time constraints and the immaturity and status of design of the candidate technologies precluded making quantitative risk assessments.

However, the panel was able to:

1. make a qualitative evaluation of whether each technology can be operated safely, given the current state of development (assuming adequate attention is paid to the intrinsic safety issues for each technology)
2. identify the intrinsic safety issues for each technology and evaluate the current treatment of these issues by the TPCs
3. provide focus for a future comprehensive, quantitative risk assessment prior to implementation

To avoid confusion, the following discussion refers to the activity of the panel as "evaluating risk factors" and reserves the terminology of "assessing risk" for the future detailed risk assessments. As explained in Chapter 1, the panel insisted that the Army obtain preliminary accidental-release risk assessments for the alternative technologies as input to the decision to be made by the Defense Acquisition Board on pilot-testing one or more alternative technologies. The panel's view of the scope appropriate to these very preliminary and qualitative assessments is discussed below, under Risk Assessments prior to the Pilot-Testing Decision.

In-plant safety and health risks depend on the nature and magnitude of hazards within the processing facility. The panel's preliminary evaluation of an alternative technology included the following components of this risk category: the risk of catastrophic failure and agent release, the risk of exposing workers to agent, the risk of worker exposure to other hazardous chemicals used in or produced during the process, and the risks from hazardous process conditions. These risks are affected by (1) the hazard inventory (agent; stored thermal, mechanical, and chemical energy; and reactive chemicals), (2) process-intrinsic safety (safety features engineered into the process design), and (3) worker controls (e.g., in-plant monitoring for worker exposure, maintenance procedures, and campaign duration).

Although the consequences associated with risks to community safety and health differ from the consequences of risks to the environment, the release factors that cause the risks are generally similar enough to treat both categories together, at least at this stage in evaluating alternative technologies. The release factors include not only those that can cause acute exposure to agent or toxic process residuals but also those that cause latent health effects or gradual environmental damage from long-term, low-level emissions and discharges.

Concerns about both kinds of exposure have led many citizens in the communities near stockpile sites to favor process designs with a "test-prior-to-release" requirement for all process residuals. This testing must be capable of detecting very low-level, continuing concentrations of a hazardous material, as well as one-of-a-kind, brief releases at high concentration.

Factors to consider in evaluating risks to the community and environment (and in detailed risk assessments) include all handling and processing throughout the

projected period of facility operations, the limited scale and finite time of stockpile-destruction operations at each site, and hazards from off-site disposal of residuals. The specific components of risk, many of which require detailed risk assessment to identify and estimate realistically, include:

• risks from agent release and exposure during the destruction process or from storage and handling prior to destruction
• risk of latent health effects from exposure to nonagent releases from the destruction process (realistic information on this risk requires site-specific health effects assessments)
• risks from managing process residuals, whether off-site or on-site, after the destruction process (again, context-specific risk assessment is needed to provide realistic information useful to decision-making)
• risks to the community or environment associated with the total environmental burden (burden as quantified by total residual process streams that are released to the environment), including the potential impact on natural resources (agriculture, bodies of water, etc.) from aqueous discharges, atmospheric emissions, or solid-waste management

The first and third bullets in this list may require special consideration in future detailed risk assessments. One such consideration includes consequences for emergency preparedness or emergency response-for example, the extent of the area that would be affected by an accidental release of agent or of toxic nonagent materials.

Risks to the community and environment from agent storage have been cited as a reason for prompt destruction of the stockpile (NRC, 1994b). These storage risks have been the focus of ongoing debate in communities near several stockpile sites. The storage risks that vary with the agent destruction system, whether that system uses an alternative technology or is the baseline system, depend primarily on the duration of storage and therefore on the overall schedule for each option. Actions that can reduce storage risk at individual sites, other than shortening the storage time, are for the most part independent of the technology for stockpile destruction. The Army is currently assessing the storage risks at all stockpile sites in the continental United States and may consider reconfiguring individual stockpiles based on the results of the evaluation.

Before any technology is implemented at a stockpile site, two site-specific risk assessments will be required: a comprehensive quantitative risk assessment in which the likelihood of events leading to the unintended release of agent or toxic materials and the consequences of such a release are analyzed, followed by a health and environmental risk assessment in which the potential consequences of accidental or continuing low-level exposure of the community or the environment are assessed. These assessments cannot be properly performed until after pilot-testing of a technology and detailed engineering planning of the full-scale facility. However, the AltTech panel believes that a preliminary, comparative assessment of risks associated with the alternative technologies is necessary for a decision to recommend a technology for pilot demonstration. If the pilot demonstration is successful and the alternative technology is selected for full-scale implementation, the two more-rigorous, site-specific risk assessments must be completed before a full-scale facility is built and agent destruction operations begin.

As noted above, the panel encouraged the Army to support a preliminary accidental release risk assessment before the pilot-testing decision. A preliminary assessment for each alternative technology should be prepared as input to the decision on whether to pilot-test one or more of them. This assessment should include the kinds of accidental release scenarios that can reasonably be envisioned during the operation of the technology, a measure of the probability of various accidental release scenarios and their likely magnitude (the probability measure could be qualitative), a measure of the impact of potential accidental release scenarios on worker health and safety, and a preliminary assessment of the impact of a release on public health and the environment.

To compare the effect of alternative technologies on the implementation schedule for stockpile destruction, the panel needed estimated schedules for each alternative technology at each potential site. These technology-specific schedules had to include time ranges for technology development, pilot-scale evaluation, and full-scale implementation and operation. The panel requested schedules from the TPCs and the Army indicating major milestones—and the assumptions made in

estimating them—for (1) laboratory and bench-scale development, if applicable; (2) pilot plant design, construction, and, operation, with subsequent analysis of pilot plant data; and (3) design of the full-scale plant, acquisition of equipment, and the construction, startup, operation, and decommissioning of the full-scale facility.

Public opposition, regulatory review, and permitting requirements can cause significant delays in the implementation schedule, but informed public acceptance and support can help to overcome regulatory or statutory hurdles. The actual time required to implement a system and eliminate the stockpile will not only affect compliance with the CWC but will also significantly affect the overall risk at each site, because storage risk depends on the duration of storage.

The panel met with members of the communities near the Newport and Aberdeen sites, with representatives of the Indiana and Maryland CACs, and with state regulators to solicit information and learn how these groups see issues affecting the implementation of each alternative technology. In particular, regulators were asked to provide information on technology-specific permitting requirements. CACs and local communities were asked to discuss their specific concerns about the technologies selected for evaluation and their views on criteria that should be used in the evaluation.

The factors and subfactors described in this chapter provided the framework for the panel's assessments and evaluations. For example, the framework of factors was used as the outline for the information to be gathered and presented in the detailed individual assessments of the alternative technologies (Chapters 4 through 8). The framework was also used to generate the detailed questionnaires that were sent to the TPCs and regulators (Appendix J). The framework was also the basis for the public forums and for the panel's discussions with CACs (Chapter 9). Following the information-gathering stage, the panel refined the framework of factors and subfactors to derive specific evaluation criteria for comparing alternative technologies (Chapter 10).