Evaluation of the Army's Draft Assessment Criteria to Aid in the Selection of Alternative Technologies for Chemical Demilitarization (1995)

The Army will rely heavily on information developed by the research, development, test and evaluation (RDT&E) program to decide whether to construct a pilot plant to demonstrate a neutralization-based process. The RDT&E results will be evaluated in the context of the assessment criteria that are the chief subject of this report. The adequacy of the technical information resulting from the RDT&E program is expected to be the topic of a second NRC report on alternative technologies, which is scheduled to be published in mid-1996. This chapter outlines the types of technical information about any neutralization-based process that will be needed to make a sound decision about whether to proceed to a pilot-scale demonstration.

Along with its earlier recommendation to research alternative technologies (NRC, 1994b), the NRC advised that several processes should be considered and a narrowing-down procedure should be implemented to avoid carrying all alternative processes to full development. The Army has included this concept in its RDT&E program (U.S. Army, 1994a).

There are major differences between the Aberdeen and Newport sites, including differences in the agent stored, differences in agent storage methods, and differences in local site characteristics. Thus, complete site-specific information should be developed for each site, independent of whether the same process is used for both sites or a different process is used at each site. Similar data must also be available for the baseline system for comparison purposes.

These performance characteristics, which are prescribed in the draft treaty, must be met by any alternative process for a particular agent. The Chemical Weapons Convention specifies use of “a process by which chemicals are converted in an essentially irreversible way to a form unsuitable for the production of chemical weapons.” The Chemical Weapons Convention language has been clarified through the development of schedules of unacceptable precursors. Language indicating a need to cleave the carbon–phosphorus bond in VX hydrolysis products such as methylphosphonic acid is likely not to be incorporated into the implementation protocols (Evans, 1995).

The character of the agent to be destroyed and the containers to be decontaminated must be determined before the successful performance of any alternative process can be assured. Examples include mustard “gelling” and the existence of field-grade mustard and VX, rather than pure agent materials, in the stockpile.

Information about all waste streams from all processes must be considered because waste-stream components will vary with both agent type and process. It is also possible that the allowable limits will be different for the two different sites owing to local regulations. Information should be provided on the limits for both characteristics and quantities for all solids, liquids, and gases, including percentage of total, if required.

The information for decision should include a listing and discussion of the operational, transportation, and discharge limits and related precautionary rules and regulations for all flammable or explosive materials used in all steps of the process, including final disposal.

The Army should determine if there are any limits imposed by instability of reactions, such as reactions that operate in a temperature regime where additional exothermicity could cause a runaway reaction-rate problem. Additional process instabilities may result from extreme pressures, phase changes, or corrosivity of process reactants or products. If such reactions are anticipated, it may limit the amount of agent or other hazardous components that can be processed at one time.

The chemical and petroleum industries have always faced the problems of bringing new processes on line. These same problems must be faced by the Army with respect to technology development for agent neutralization and biotreatment alternatives. The transition from bench-scale laboratory work, through pilot-plant demonstrations, to an eventual operating plant is difficult and expensive. Industry has found it efficient to integrate the engineering and research efforts early in the process. The engineering design will indicate the highcost items and the problems and data requirements that researchers are likely to overlook. Continuous communication between researchers and design engineers can lead to optimization of process conditions and, for example, avoidance of conditions requiring expensive materials of construction.

Development of the Army's neutralization-biotreatment technology would benefit from industry 's approach. The committee believes that a continuing collaboration between process designers and researchers is desirable; in fact, collaboration is essential to address both the committee's and the Army's assessment criteria. The design information that is needed to decide whether to proceed to a pilot plant is discussed in the subsections that follow.

The purpose of the conceptual design package is to demonstrate the feasibility of using an alternative set of process unit operations to conduct all activities that are required to complete the disposal program and to provide a basis for its comparison with the baseline system. At a minimum, the design package should include the elements discussed in the following sections.

The conceptual design must address all steps, including the decontamination of all storage containers, equipment, carbon filters, and any other agent-contaminated wastes. In addition the conceptual design must address the activities involved in removing wastes or byproducts and must include all subsequent steps if wastes or byproducts are taken off site. The conceptual design must also indicate all material flows to each site, as feeds, and from each site, as waste discharges.

The level of detail required to allow both meaningful technical review and the development of meaningful cost estimates requires a complete conceptual design, including full technical data for all process steps, equipment lists, and designs. This information is needed to estimate the cost of purchasing and installing individual equipment components using standard cost-estimating methods and construction-cost factors. Additional design data to demonstrate the proposed reactions, reaction rates, and equipment sizing should also be included in the conceptual design. Overall, the conceptual design should contain at least the following types of detailed information:

1. Process description. The information package should include a description of the total process that details how the data have been used to develop the equipment requirements and how the various agent destruction, decontamination, and waste-processing steps are conducted. The description should also provide an adequate basis for establishing that, after pilot testing, the process will have a high probability of successfully performing agent-destruction and waste-disposal functions.

2. Process data. Chemical and physical properties of all process materials should be provided to the extent that these data are needed to design each unit operation in the overall process.

3. Flow sheets. Flow sheets should be provided that show all equipment, piping, and general control methods, including:

• material and energy balances that indicate all material flow rates and heat or energy requirements or heat-generation and heat-removal rates for each step of the process;

• process monitoring and control activities that include all process monitoring instrumentation and describe the method(s) used to control the process; and

• descriptions and characterization of all process waste streams.

4. Storage facilities. Storage for all feed materials and all wastes prior to their final disposition, including off-site shipment, should be described.

5. Utility requirements. Utility requirements, including process requirements for both fuel and electricity, should be included. Backup requirements to allow for emergency shutdown of the process and related pollution-control systems should also be provided.

6. Feed materials. Requirements should include both quantities and qualities of all chemicals.

7. Equipment lists. All major pieces of equipment for the destruction process, secondary treatment systems, and pollution-control systems should be listed.

8. Equipment designs. Design sketches, sizing calculations, and materials of construction for all major pieces of process equipment should be included.

9. Plant layout. The layout and working space for the major pieces of equipment, plot plans for the current storage facilities, and planned methods for transporting agent containers from the storage area to the destruction facilities should be provided.

In addition to conceptual design of a full-scale chemical demilitarization plant based on the alternative technology under consideration, preparation for the Army Systems Acquisition Review Council and Defense Acquisition Board decisions should include planning the development and demonstration activities that will be needed if the decision is to proceed with construction of a pilot plant. These activities include:

Plant scale-up requirements. The principles to be used to scale-up from the current laboratory-data level to the pilot-plant scale and from the pilot-plant scale to the full-production level should be described. Multiple modules may be acceptable for some components, but not necessarily for all processing steps.

Additional laboratory tests. Laboratory tests that may be appropriate during the pilot plant development period should be outlined.

Pilot-plant test program. The steps recommended for conducting pilot-plant testing and demonstration should be described.

Good engineering practice also requires preparation of several other types of plans associated with building, operating, and closing a process facility. The development of these plans often generates unforeseen questions that may require additional research and development activity. Some types of plans commonly required are:

• construction plan;

• operations plan;

• subsequent waste-disposal plans, if applicable following operation of the facilities;

• decommissioning plan;

• schedule (time requirements) for each of the above steps and summarized for all steps;

• preliminary hazards analyses; and

• costs for further development and continued storage and for facility construction, operation, and decommissioning.

It is common practice to include factors in designs, time schedules, and cost estimates to provide for uncertainties and unplanned events. The magnitude of these factors is normally greater for less-developed systems because there are greater uncertainties. These uncertainty factors are in addition to those used to achieve reasonable variables in plant operation. These additional factors should be identified, recognizing that they are dependent on judgment.

Three types of risk assessments—accident, environmental, and health—will be needed before an alternative technology is implemented at any site. These assessments may need to be done in two phases. In the short time before the decision whether to demonstrate an alternative technology, the risks of the alternative should be compared with those of the baseline system, at least in preliminary fashion. If the demonstration succeeds and a decision is made to implement the alternative technology at full scale, a second phase of the risk assessments, i.e., rigorous quantitative assessments of risks of accidents, environmental impacts, and health impacts on both on-site workers and the public should be carried out before a “production facility” is built.

Risk analyses generally address the basic questions:

• What can go wrong?

• What is the likelihood of something going wrong?

• What are the consequences of something going wrong?

• What are the potential human health and environmental impacts of routine operations? (Even if nothing goes wrong, what are the impacts of routine emissions and discharges?)

A full quantitative risk assessment for a complex process like chemical agent destruction is not possible until a complete process design has been developed. However, preliminary responses to the questions above at a level needed to assess the feasibility of an alternative technology should be doable in the time available, especially when comparing features of the alternative to those of the well-developed baseline system. As an example, one might compare the likelihood and consequences of the failure of the reaction chambers for the two systems. Such an analysis would contribute to improving the design of the alternative system as well as assessing its safety. For each failure scenario judged to be critical, the three types of assessment studies listed below should be conducted.

A site-specific update of the prior accident risk assessment for the baseline system and a preliminary, accident-risk comparison for the selected alternative technology should be prepared.

A preliminary, site-specific environmental-impact comparison based on projected process emissions and local geographic considerations (e.g., population distribution, ecosystems) should be prepared.

A preliminary, site-specific health-impact comparison from both agent and nonagent emissions should be developed for the alternative technology under consideration and the baseline system with which it is to be compared. The comparison should include potential health impacts of all discharges during routine operations.

Uncertainties in specific aspects of the stockpile disposal program limit the information available for the decision process. Some of the uncertainties arise because the RDT&E schedules do not allow adequate time to collect all of the data that would be desirable prior to deciding whether to conduct the pilot demonstration. Delays in the initiation of bench-scale testing may limit the amount of data available when the information for decision is finalized. While decisions based on incomplete information are common in process development, the level of uncertainty should be recognized.

Important areas of uncertainty are (a) unforeseen process development difficulties, and (b) implementation delays for the baseline system or the alternative technology. In addition, specific information may not be available at key decision points because of programmatic schedules (e.g., time for development of site-specific environmental and health risk assessments) or research and development schedule constraints (e.g., specific process data or permit requirements). To manage these areas of uncertainty several decision options must be addressed:

Option 1. Decide not to proceed to pilot evaluation of an alternative process because information gathered to date is clearly negative.

Option 2. Decide not to proceed because specific questions cannot be answered with respect to safety,

cost, or schedule issues. Uncertainty and information gaps are sufficiently great as to be considered negative information.

Option 3. Proceed to pilot demonstration of an alternative process on the assumption that initial information indicates that the process will meet technology requirements and continue to gather information.

Option 4. Delay the decision and gather more information.

Option 4 may result in additional program delays and increased disposal costs resulting from extended stockpile storage times. This option may increase the likelihood that a future decision on pilot demonstrations is negative because of the detrimental impact of schedule delays on overall programmatic risk and cost.