Disposal of Neutralent Wastes: Review and Evaluation of the Army Non-Stockpile Chemical Materiel Disposal Program (2001)

The NSCMP is preparing to test two prototype transportable systems, the RRS and the MMD, to destroy a range of nonstockpile chemical agents and militarized industrial chemicals. The RRS is designed to treat recovered CAIS, which were used to train soldiers in the detection and identification of chemical agents and decontamination procedures. The MMD is designed to treat nonexplosively configured chemical munitions (i.e., munitions containing chemical agents but no fuzes, propellants, or burster charges).

Because CAIS and recovered munitions contain different materials, the RRS and MMD must use different reagents to destroy the chemical agents. Although both systems appear to be effective for destroying the target agents (to ppm levels or below), they also produce liquid waste streams containing complex mixtures of reaction by-products, excess reagents, and organic solvents. These liquid waste streams are referred to in this study as “neutralents.” ¹

Because neutralent waste streams from the RRS and MMD are expected to be classified as hazardous wastes under RCRA, 2 the Army had planned to ship them to a permitted hazardous waste incinerator for final disposal. However, because the incineration of chemical agents has aroused considerable opposition among public interest groups, and because this opposition may be extended to the incineration of neutralents, the Army is also investigating alternative (nonincineration) technologies for disposing of neutralents.

This chapter describes the composition, quantity, and toxicology of the neutralent waste streams expected to be generated by the RRS and MMD. In keeping with the Statement of Task for this portion of the study, the committee accepted the treatment processes and neutralent compositions as given. The committee did not consider upstream changes in the treatment chemistry or process conditions that might produce neutralents with different characteristics.

Approximately 110,000 CAIS were produced in various configurations from about 1928 to 1969. These sets contain (1) neat chemical agents and/or (2) agents dissolved in chloroform in glass vials or glass bottles and/or (3) agents adsorbed on charcoal in glass bottles. The chemical agents include blister agents, sulfur mustards (HD and H), nitrogen mustard (HN-1 and HN-3), and lewisite. ³

The treatment chemistry of the RRS is based on the oxidation of chemical agents with 1,3-dichloro-5,5-dimethylhydantoin (DCDMH) dissolved in a mixture of chloroform/t-butyl alcohol/water. This reagent was selected because it does not react with chloroform, which is present in some CAIS items, it maintains a reasonable reaction volume, and it does not generate significant amounts of heat or gaseous products during the reaction.

Depending on the CAIS item and the agent it contains, one of four reaction processes is selected. The processes (blue, red, charcoal, and charcoal-L) have different proportions

of DCDMH and different solvents (see Table 2-1 ). The blue process is used to treat neat sulfur mustards (H or HD). The red process is used to treat sulfur mustard, nitrogen mustard (HN-1, HN-3), and lewisite that are dissolved in chloroform. The charcoal process is used to treat sulfur mustards and nitrogen mustards adsorbed on charcoal, and the charcoal-L process is used to treat lewisite adsorbed on charcoal.

The chemical reactions of the RRS processes are complex, and a large number of products are present in the neutralent waste streams. However, all four processes effectively destroy the chemical agents and produce waste streams that could be shipped to hazardous waste incinerators. The compositions of the neutralent waste streams are shown in Table 2-2 . The dominant constituents consist of the reaction solvents and excess DCDMH. A large number of reaction by-products (e.g., sulfones, sulfoxides, etc.) are present in low concentrations.

Arsenic found in lewisite is converted in the red and charcoal-L processes into chlorovinylarsonic acid (CVA) in quantities of up to 3 percent by weight in the neutralent waste stream ( Table 2-2 ). The fate of CVA depends on the post-treatment processes. In a SCWO reactor and in the GPCR caustic scrubber brine, the CVA is expected to be converted to sodium arsenate salts (e.g., Na₃AsO₄ and Na₄As₂O₇), which can then be treated with ferric chloride to produce ferric arsenic salts for disposal in a hazardous waste landfill. This treatment scheme, which was developed in Canada, is the basis of the treatment of bulk lewisite at the Chemical Agent Munitions Disposal System facility in Utah.

If 40 to 45 CAIS ampoules or 12 to 15 CAIS bottles are treated per day, the estimated volume of neutralent generated would be less than 15 gallons per day (U.S. Army, 1999a). The 1,189 CAIS items located at Deseret Chemical Depot are expected to generate a total of about 468 gallons of liquid neutralent (Gieseking, 1999). 4

The treatment chemistry of the MMD is based on (1) the hydrolysis of HD and GB with monoethanolamine (MEA) and water or (2) the hydrolysis of VX with MEA-aqueous sodium hydroxide solution. MEA was chosen as the reagent based on previous experience with it in Russian chemical demilitarization programs. The advantages of MEA include good solvent properties for agents, miscibility with water, noncorrosivity to stainless steel under operating conditions, and low flammability. MEA cannot be used in the RRS because it reacts violently with chloroform, the solvent present in many CAIS items.

Depending on the type of agent in the munition, one of three reagents is selected for the MMD process (see Table 2-3 ): a mixture of MEA and water; MEA and aqueous sodium hydroxide; or just aqueous sodium hydroxide. Phosgene is reacted with aqueous sodium hydroxide to form simple inorganic salts. A minimum ten-fold volume excess of reagent solution is used to ensure destruction of the agent and to control viscosity.

The compositions of neutralent waste streams from the MMD are complex, as shown in Table 2-4 , Table 2-5 , Table2-6 to Table2-7 . Because many of the munition bodies introduced into the MMD have already degraded during their long burial, metals and debris may be mixed with the neutralent during the cutting and rinsing processes. If a large amount of sodium hydroxide is added during processing, the pH of the neutralent may exceed 14 (hazardous waste).

In current testing of the MMD, one CWM munition or container can be processed per day. In future tests or in normal operation, the rate may be two items per day (U.S. Army, 1999a). The initial testing of the MMD at Dugway Proving Ground in Utah is expected to generate approximately 6,412 gallons of neutralent, an average of 57 gallons of liquid waste (including liquid wastes from rinsing the system after processing) for every gallon of agent or industrial chemical processed (Gieseking, 1999).

A number of reports produced or sponsored by the Army describe dermal and inhalation toxicity studies of the oxidant/solvent systems (O/SS) used in the RRS and MMD,

as well as a waste stream from these systems (e.g., U.S. Army Research Office, 1994; DOT, 1997; and U.S. Army, 1999a). The effects of exposure to individual components of the O/SS are shown in Table 2-8. In general, the toxicity (i.e., exposure response data) of both RRS and MMD neutralents is comparable to the toxicity of components of the O/SS (U.S. Army, 1999a).

The inhalation toxicity of neutralents from the RRS red process treatment of HD, HN, and lewisite was tested in rats by 14-day exposures. The neutralent contained 53 percent chloroform, 30 percent t-butyl alcohol, trace amounts of DCDMH, and less than 1 ppm HN or HD, or 37 ppm lewisite. The toxicity of the waste stream was compared with that of an aerosol containing 58.3 percent chloroform, 39.1 percent tert-butanol, and 2.6 percent water (the vehicle control). Concentrations of 24,000 ppm of the vehicle control or neutralent killed all of the test animals. Lower doses caused excessive salivation, ocular and nasal discharge, lack of coordination, listlessness, difficult breathing, and corneal opacity. The inhalation effects of the neutralent on test animals were consistent with those of the t-butanol and chloroform components of the O/SS (Morgan et al., 1997).

The dermal toxicity of RRS neutralents from all four processes and O/SS was tested by exposing the skin of rabbits to the solutions under an occluded patch for 24 hours. All solutions caused redness and swelling, but with the exception of the charcoal process, the effects of the neutralent were less severe than those caused by the O/SS. The dermal effects of the charcoal process neutralent were comparable to those of the O/SS because of the moderately toxic HD degradation products produced when DCDMH reacts with HD in the absence of water (DOT, 1997).

The vesicant (blister formation) properties of the neutralents from RRS processes were tested by dermal application to hairless guinea pigs. The only neutralent that caused vesication was from the blue process (treatment of neat HD). Because the concentration of HD in neutralent (less than 50 ppm) is too low to cause vesication, the blistering was attributed to the presence of HD oxidation products (Olajos et al., 1997). ⁵

Unlike the reaction between DCDMH and HD in the RRS, reaction between MEA and HD in the MMD produces relatively few toxic breakdown products. This is reflected by the lower toxicity of the HD neutralents from the MMD process.

The dermal toxicity of O/SS and simulated neutralents from the treatment of HD, GB, and VX were compared by exposing the skin of rabbits to the solutions under an occluded patch for four hours. The effects were recorded 24 hours after exposure. Severe redness and swelling were observed in all cases, but for the most part, skin injuries from the neutralents and O/SS alone were comparable. No systemic toxicity resulted from 24-hour dermal exposures to either the O/SS or neutralent solutions (Olajos et al., 1996).

The vesicant properties of the O/SS and the HD/MEA/ water neutralent were studied by dermal application to hairless guinea pigs. Blisters did not result from dermal exposure to either the O/SS or the neutralent solution (Battelle, 1997).

Because neutralents contain compounds that are classified as hazardous, they will be regulated under RCRA Subtitle C. They will also be regulated under the CWC based on the chemical agent(s) they contain and under the U.S. Department of Transportation (DOT) regulations because both hazardous materials and chemical agents have special requirements for transport in the United States.

RCRA Subtitle C is the “cradle-to-grave” approach of managing hazardous waste, including generation, storage, shipment, treatment, and disposal. Under RCRA, neutralents produced by the RRS and the MMD may be classified either as listed or characteristic hazardous wastes. If the neutralent waste stream contains phosgene, it will either be classified as a listed hazardous waste, or, if it is corrosive (pH >10), as

a characteristic hazardous waste (40 CFR 261; NRC, 1999a). ⁶ The neutralent waste stream could be regulated under the federal or state requirements (or both) of RCRA. In some cases, state requirements are more stringent than federal requirements.

The storage of GB and VX neutralents from the MMD is subject to the constraints of the CWC. Some of the breakdown products (e.g., amiton [S-[2-(diethylamino)ethylphosphorotioic acid O,O-diethyl ester]), are listed as Schedule 2 precursors to the manufacture of chemical agents (i.e., chemicals that could be used to remanufacture chemical agent). This means that, theoretically, the precursor chemicals in the neutralent could be reprocessed from the neutralent and used to remanufacture chemical agents. To prevent the manufacture of chemical weapons, the CWC requires that Schedule 2 precursors derived from existing

agents be destroyed in the same time frame as the chemical agents.

The reader should note that Table 2-4 , Tabel 2-5 , Table 2-6 to Table 2-7 describing the composition on neutralent wastes derived from bench tests of the RRS and MMD neutralents do not list any Schedule 2 precursor compounds. Yet the Army suggests that a major argument against storage is that neutralents may contain Schedule 2 compounds and therefore must be destroyed per the CWC schedule. The committee does not find this to be inconsistent. Whenever chemical agents are treated, Schedule 2 breakdown products could be produced. The bench test data in the tables indicates that this did not occur in these tests. However, until more is known about compounds produced by the reactions occurring in the RRS and MMD and a significant body of data has been established for large-scale operation, it is best to take a conservative approach and assume that some Schedule 2 breakdown products will be present in sufficient quantity to preclude storage.

Significant concerns have been raised about the transport of CWM. Indeed, concern about the movement of chemical agents has been a driving force behind the development of

transportable treatment systems. As far back as 1969 (P.L. 91-121 ), Congress placed severe, almost insurmountable, restrictions on the transport of CWM, including a requirement for advance notification and coordination of shipments with the U.S. Department of Health and Human Services and Congress (except in cases of emergency). In 1995, Congress placed restrictions on moving nonstockpile CWM out of any state except to the closest permitted CWM storage facility, and then only under very strict conditions. Public concerns about transporting CWM have effectively fore-closed even this option except in extraordinary situations.

The neutralents generated by the RRS will primarily include hazardous waste and hazardous materials, which make them subject to RCRA and DOT requirements. If these are the only constituents, the neutralent could be transported as a routine hazardous waste or hazardous material under existing laws and regulations as long as it was properly packaged, marked, manifested, and shipped as required by those regulations.

However, the neutralent from the RRS and MMD may contain trace amounts of residual chemical agents (e.g., GB [= 25 ppb], sulfur mustard [= 50 ppm], VX [= 1 ppm]). At these levels, the toxicity studies cited above indicate that the concentration of residual agent would be too low to affect the overall toxicity of the waste streams. It is not known how much, if any, chemical agent would be present at concentrations lower than the detection limit. Thus, based on the information provided by the Army, transporting neutralent wastes from the RRS and the MMD should not be subject to any restrictions beyond the applicable federal RCRA, DOT, and state regulatory requirements. However, the public perception of “residual chemical agents” from the MMD waste streams may arouse concerns.

Although these residues will be in extremely small amounts, the public could consider the overall neutralent waste mixture as “tainted” with chemical agent and, therefore, of special concern. The Army should address this potential problem proactively. This could be done in several ways. In a previous report, for example, the committee recommended that a comparative risk assessment be performed of the disposal of CWM in CAIS in an incinerator and the disposal of typical hazardous waste (NRC, 1999a). The Army could assess the comparative risk of transporting and disposing of neutralent and transporting and disposing of typical hazardous waste. Providing the public with this type of information would increase the transparency and credibility of the process.