Evaluation of Alternative Technologies for Disposal of Liquid Wastes from the Explosive Destruction System (2001)

In its previous report (NRC, 2001a), the committee evaluated eight nonincineration technologies for their ability to process liquid wastes produced by the rapid response system (RRS) and the munitions management device (MMD). A technology was selected for evaluation if one or more of the following were true:

• A substantial amount of information about it was available.

• It was under serious consideration and/or evaluation for other demilitarization or waste treatment processes—for example, the Assembled Chemical Weapons Assessment (ACWA) Program—or used widely for commercial waste treatment.

• It was likely to be safe, effective, easily permitted (relative to incineration technologies), and capable of passing the committee’s pollution prevention criteria.

A complete list of the evaluation criteria used by the committee in this report and the previous report (NRC, 2001a) appears at Appendix E. The committee gave the “top priority” criteria more weight than the “important” criteria.

The eight technologies evaluated were the following:

• chemical oxidation using a variety of oxidants, such as hydrogen peroxide, ultraviolet-activated hydrogen peroxide, ozone, and peroxydisulfate

• wet-air oxidation, or a variation using pure oxygen instead of air

• electrochemical oxidation using silver (II) or cerium (IV) ions to oxidize organic compounds

• supercritical water oxidation

• solvated electron technology

• biodegradation

• gas-phase chemical reduction

• plasma arc technology

Process descriptions and evaluations for each of these technologies were given in Chapter 4 of Disposal of Neutralent Wastes (NRC, 2001a).

In its previous report (NRC, 2001a), the committee recommended that the Non-Stockpile Chemical Materiel Product (NSCMP) should pursue a two-track strategy to identify a suitable treatment technology for MMD neutralents. As part of the track-one strategy, the committee recommended that NSCMP should take advantage of available equipment that would require little or no investment—that is, it would piggyback on alternative technologies from the ACWA Program or existing commercial technologies, such as chemical oxidation, wet-air/O₂ oxidation, or existing plasma arc technology. The committee judged that if any of these existing and available technologies can accomplish the task safely, this might be the most rapid and inexpensive course of action.

If, on the other hand, neither the ACWA nor the commercially mature technologies could be used as is—for example, if substantial process or permit documentation would be needed to dispose of nonstockpile neutralents—then as part of the track-two strategy, the committee recommended that the Army should invest research and development resources first in chemical oxidation and wet-air/O₂ oxidation.¹ Only if these technologies cannot be adapted easily did the committee recommend that the Army consider investing re-

sources in the adaptation of ACWA technologies listed below in order of preference:²

1. electrochemical oxidation with Ag(II) or Ce(IV)

2. supercritical water oxidation

3. gas-phase chemical reduction

4. plasma-arc technology.

Biodegradation was judged unsuitable for treatment of MMD and RRS neutralents. The committee recommended that the Army should not invest in further development of biodegradation for nonstockpile wastes.

Since the publication of the committee’s previous report (NRC, 2001a), technology test data have become available from two ongoing programs that bear on the committee’s assessment of the technologies in track two. Although the data are preliminary and many of the tests are incomplete, the committee wished to incorporate the latest results (those available as of July 2001) into its recommendations for the EDS-1 liquid waste treatment technologies in this report. One data source is NSCMP’s Technology Testing Program, in which several ACWA technologies are to be tested on MEA-based neutralents/rinsates or simulated neutralents/ rinsates. The status of this program is summarized in Table 3–1. The other data source is testing from the ACWA Program itself. The ACWA Program has been reviewed by another NRC committee (NRC, 1999a; NRC, 2000; NRC, 2001b).

The EDS-1 uses the same MEA-based reagents for HD, GB, and VX as does the MMD, and it also uses the same aqueous sodium hydroxide reagent to treat CG (see Table 2– 4). The EDS may, however, be called upon to dispose of munitions having fills that the MMD was not designed to process (e.g., lewisite), and for these munitions and fills, other reagents may be needed.

Although the neutralizing reagents used in the MMD and the EDS-1 are essentially identical, there are several operational process differences between these systems that give rise to differences in the compositions of the liquid waste streams. First, the MMD³ was not intended to process

nonstockpile items containing energetics, whereas items processed in the EDS-1 may contain bursters and fuzes. Consequently, the potential exists for some energetics and their reaction products to be found in the EDS-1 neutralents.

Second, in the MMD the chemical agent is accessed by drilling into the item to be processed and draining the agent, while the EDS-1 uses explosives to open the munition and release the agent in its explosion containment vessel. The metal fragments of the munition body are then rocked back and forth with the neutralizing reagent for several hours, which may result in higher concentrations of metals being found in EDS-1 neutralents than in MMD neutralents.⁴ Metals such as lead and mercury may also be present in the mu

nition fuze materials and become dissolved in the EDS neutralents.

Third, after the neutralent is drained from the EDS-1 containment vessel, operating procedures call for further treatment of the vessel contents with water at 100°C, followed by water rinses at about 50°C.⁵ Thus, EDS-1 produces additional aqueous waste streams that must be disposed of.

In evaluating the ability of each of these technologies to process liquid wastes produced by the EDS-1, the committee took account of the above differences in the compositions of the MMD and EDS-1 liquid waste streams and considered the results of ongoing tests of these technologies by the Army in processing simulated and actual liquid wastes. The committee’s evaluation and its reranking of the technologies for application to the EDS-1, where appropriate, are described below.

The use of chemical oxidants, such as hydrogen peroxide, potassium permanganate, Oxone,⁶ and peroxydisulfate, is a promising candidate for treatment of EDS liquid waste streams because of its technical effectiveness for similar wastes, good pollution prevention qualities, robustness, and low cost. Emissions are minimal, and production of chlorodibenzodioxins and chlorodibenzofurans is precluded because of the low temperatures involved.⁷ The biggest potential disadvantage of chemical oxidation is that it may not fully mineralize all of the compounds in the neutralents or that it may not mineralize them rapidly enough to be practical. Additionally, many organics, particularly simple aliphatics and halogenated alkanes, are somewhat recalcitrant to simple chemical oxidation. Long reaction times, large amounts of oxidants, and significant dilution may be required.

Oxidation without UV enhancement is much preferred over treatment with UV enhancement. The problems associated with UV activation are discussed in NRC, 2001 a. These include the need for special equipment, lack of effectiveness when the solution being treated is opaque, and fouling of the optical surface.

The potential presence of residual energetics or their hydrolysis products in the EDS neutralent should not pose any particular difficulty. These materials will probably also be oxidized, although this remains to be verified by testing. If additional metal ions/hydroxides are present in the EDS neutralents, they may consume oxidant or catalyze its decomposition; appropriate testing could answer this question.

The dilute aqueous rinsates produced during EDS operation should be amenable to treatment by chemical oxidation. Typical oxidants such as those mentioned above are commonly employed in aqueous systems and do not react with water.

If arsenic is present, it will probably be oxidized to arsenate ion, which can be stabilized. Organic arsenicals will probably be mineralized, which would be serendipitous since they are difficult to stabilize (Conner, 1990).

The experimental evaluation of persulfate oxidation of EDS neutralents was under way as this report was being written (Table 3–1). The results of this work will be extremely helpful in evaluating the usefulness of chemical oxidation.

Wet air oxidation (WAO) is a hydrothermal process for the oxidative destruction of organic wastes that is carried out in liquid water at temperatures of 150° to 315°C and pressures of 150 to 3,000 psia. The oxidizing agent is dissolved oxygen, which may be derived from air, or pure oxygen may be supplied externally.

WAO operates at temperatures and pressures below the critical point of water (374°C; 3,204 psia). Most inorganic salts that would form during oxidation of EDS neutralents are soluble in subcritical water. Therefore, WAO, unlike supercritical water oxidation (see below), is not prone to plugging by precipitated salts.

WAO can treat any pumpable fluids provided that the chemical oxygen demand (COD) is less than 120,000 mg/L. To achieve this, one technology (William Copa, U.S. Filter Zimpro, personal communication, February 15, 2000) estimates that the MMD neutralent for GB would have to be diluted seven- to ninefold prior to wet air oxidation. The MMD neutralent for VX would have to be diluted 12- to 15-fold. Similar levels of dilution would be required for the EDS neutralents and could be achieved in whole or in part by combining the primary neutralent with the dilute aqueous rinsates. It should be noted that the dilution required to achieve the appropriate chemical oxygen demand ensures that the neutralents are aqueous-based and do not have low flash points.

The presence of small quantities of energetics or energetics reaction products in EDS neutralents may increase the COD above that for MMD neutralents, necessitating additional dilution, but otherwise, process applicability should be similar.

WAO is used routinely in commercial applications to treat sewage sludge containing 10 to 15 percent solids. Therefore, the higher levels of dissolved metal ions or suspended solids

in EDS neutralents compared with MMD neutralents are not likely to compromise the applicability of the process.

WAO does not fully mineralize organics but instead reduces them to short-chain molecules such as acetic acid (the primary component of vinegar). Thus, effluents would need to be treated further by biotreatment, possibly at a POTW. Prior to biotreatment, toxic heavy metals in the EDS neutralent would need to be precipitated, filtered out, and stabilized for disposal in a hazardous waste landfill. Arsenic, if present, would be converted to arsenate ion, a form that is readily stabilized. NSCMP plans for testing of WAO to be completed by the end of 2001.

In the electrochemical oxidation process, a strong oxidizing agent—in this case Ag(II) or Ce(IV)—is generated in concentrated nitric acid in an electrochemical cell. Ideally, the oxidizing agent then reacts with the introduced organic waste material to produce carbon dioxide and inorganic salts. After it has been reduced by the reaction with the waste, the oxidizing agent is regenerated in the electrochemical cell.

The Ag(II) process is being tested extensively by the ACWA program as a primary treatment to destroy neat chemical agents and a variety of energetic materials found in stockpiled assembled chemical weapons. Chemical agents can be destroyed to destruction removal efficiencies of 99.9999 percent; however, the treatment of energetic materials is still very immature (NRC, 1999b; NRC, 2000).

This committee found that the major disadvantage of using Ag(II) is that large quantities of silver salts and chlorides are generated, which could lead to problems with corrosion and precipitation. Recent test data from the ACWA Program indicate that silver salts containing polynitroaromatic compounds (e.g., trinitrobenzene, trinitrobenzoic acid) precipitate on the walls of the vessels and piping during the treatment of liquid wastes containing explosive residues (Winkler, 2001). These solids were difficult to oxidize any further. Another disadvantage is that large quantities of silver and nitric acid (a corrosive) are required for the operation of this technology, which could increase toxic emissions, effluents, and cost. To correct all of these problems, numerous unit operations have been added, making the system extremely complex and immature.

The Ag(II) process has never been tested as a secondary treatment for neutralents. Because neutralents contain a high percentage of water, the concentrated nitric acid solutions will be diluted. This could alter the chemistry and necessitate the removal of water.

Electrochemical oxidation with Ce(IV) avoids some of the deficiencies of the Ag(II) process; for example, unlike Ag, Ce does not form an insoluble salt with chloride ions. Nevertheless, like the Ag(II) process, the Ce(IV) process requires large quantities of corrosive nitric acid and generates large quantities of nitrogen oxides at the cathode, which must be reformed and the waste gases scrubbed.

The committee found that the most serious disadvantage of Ce(IV) is that the technology is not mature enough for immediate use (NRC, 2001 a). The electrochemical cells were designed specially by the vendor and are not commercially available. Furthermore, there is no mechanism for removing salts produced in the reaction from the anolyte solution; the solution must be drained periodically and replaced as the salt concentrations increase. The University of Nevada is using a small Ce(IV) unit to process small quantities of organic waste, including chlorine-containing compounds. Testing of this unit on mustard and sarin neutralent as part of the NSCMP’s Technology Test Program is scheduled to begin in mid-2001.⁸

Supercritical water oxidation (SCWO) is a hydrothermal process for the oxidative destruction of organic wastes. An oxidant and the wastes to be disposed are fed to a reactor in the presence of high concentrations of water heated above the critical temperature and pressure of pure water (374°C, 3,204 psia). These wastes can be fed continuously into the SCWO reactor (continuous SCWO) or, in an alternative design, a small volume of waste is mixed with water and an oxidizer (H₂O₂) in a pressure vessel, heated to reaction temperature above the critical point of water, and then cooled (batch SCWO). The committee evaluated continuous SCWO in its previous report (NRC, 2001a), but did not evaluate batch SCWO, which was still at a very early stage of development.

Although the committee found that continuous SCWO would effectively mineralize agent neutralents, it concluded that issues related to the mechanisms and locations of salt buildup, the chemical composition of the salts produced, and the effectiveness of the flushing of salts are unresolved. Corrosion and plugging of SCWO reactors, erosion of valve seats and nozzles, and pressure containment are other issues to be addressed. As a result, the committee ranked continuous SCWO fairly low as a process for mineralizing MMD neutralents.

Constituents of the EDS-1 neutralent from the GB bomblet operation include small quantities of explosive solids (a maximum of between 0.77 and 1.3 ppm) and liquids possibly containing explosives (<1,000 μg per liter of each constituent). These small quantities of energetics should, if anything, assist the temperature-sustaining exothermic reac-

tion in the SCWO reactor (NRC, 2001 a). Other components of the EDS-1 neutralent, e.g., volatile and semivolatile organics, should be readily destroyed in the SCWO chamber.

One area of concern with regard to continuous SCWO, however, is the presence of metal ions in the EDS-1 neutralent. Ten different metal ions were detected in the analysis of EDS-1 neutralent obtained from destruction of the sarin bomblets at RMA. Although the quantities detected were small, these metals add to the potential for accumulation of salts in the SCWO reactor.

The batch SCWO concept, currently being developed and tested at Sandia National Laboratories, could help address the salt buildup problem, because salts accumulated in the batch SCWO reactor can be flushed out after each batch is processed. The batch SCWO could also help to address the problems of pressure letdown and corrosion (by better control of the pH). However, relatively small quantities of neutralents can be processed per batch. The size of the transportable batch SCWO reactor currently being tested would have to be scaled up to process the volumes of neutralent produced by the EDS.⁹

Although it shows promise as a technology for mineralizing EDS neutralent wastes, batch SCWO is still in the early stages of development. Additional testing at both the laboratory and pilot scales is needed to determine whether corrosion and salt buildup are less problematic with batch SCWO than with continuous SCWO.

It should be noted that the EDS-1 has yet to be tested with munitions containing VX, arsenicals, and other miscellaneous chemical fills, and that the composition of the resulting neutralents is not yet known.¹⁰ SCWO has destroyed hydrolysates from VX neuralization with aqueous sodium hydroxide, but to the committee’s knowledge, it has not yet destroyed MEA-based VX neutralents.

Gas-phase chemical reduction (GPCR) is a thermal treatment technology (850°C, atmospheric pressure) that reduces organic chemicals to water, methane, carbon soot, and other by-products in a hydrogen-rich atmosphere. The by-products include acid gases, phosphorus-containing products from VX and GB neutralents, and arsenic-containing products from lewisite neutralent. These products, as well as the carbon soot, require scrubbing or further treatment, adding to the complexity of the process.

Although GPCR is a well-established thermal treatment technology, it can generate large volumes of effluent gases and is a complex process that requires the management of hot hydrogen gas in the reactor, the scrubbing of effluent gases, and the control of carbon soot buildup. For these reasons, as well as a lack of regulatory experience in the United States, the committee gave GPCR a low rating as a posttreatment process for MMD neutralent.

These same challenges are likely to be faced in the application of GPCR to EDS-1 neutralents. Although small concentrations of residual energetics or energetics decomposition products in the EDS-1 neutralents are not expected to pose a problem, the potential presence of additional metal ions—particularly toxic heavy metals such as mercury and arsenic—is of concern since these must be captured and disposed of.

As indicated in Table 3–1, NSCMP has completed testing of a GPCR system with a variety of actual and simulated neutralent wastes. Initial tests encountered partial plugging in a pipe between the thermal reduction batch processor and the GPCR reactor due to the buildup of a green glassy material and a tarry substance. In tests on DF simulant, hydrofluoric acid was formed, which attacked the metal reactor seals, severely corroding them.¹¹ The suitability of GPCR for treating phosphorus- and arsenic-containing wastes has not been fully demonstrated. Phosphorus-containing wastes have the potential to produce the toxic gas phosphine (PH₃), although a previous NRC report (NRC, 1996) presented theoretical calculations suggesting that phosphine is a less likely product than P(III) oxides. Arsenic-containing wastes have the potential to form the toxic gas arsine (AsH₃), although elemental arsenic may be a more likely product. Russian work on the high-temperature, gas-phase hydrogenolysis of lewisite showed the production of elemental arsenic and As(III) compounds (Petrov, 1998).

Plasma arc technology utilizes electrical discharges to produce a field of intense radiant energy and high-temperature ions and electrons that cause dissociation of chemical compounds in a containment vessel. Operating at electron temperatures as high as 20,000°C, material exposed to the plasma environment is transformed into atoms, ions, and electrons.

In its previous analysis, the committee recognized that plasma arc systems can achieve very high destruction efficiencies, but it noted that they are most efficient when used

to treat low-volume, highly concentrated feed streams (e.g., neat chemical agents) rather than dilute neutralent solutions. Recently a variant of this technology¹² was successful in destroying adamsite (DM), CG, L, H, and HL in tests conducted in Spietz, Switzerland, with a destruction and removal efficiency greater than 99.9999 percent.

Nevertheless, the committee gave plasma arc technology a low rating for treatment of MMD neutralent solutions compared with the other alternative technologies it considered. This rating was based on its reduced efficiency for treatment of dilute neutralent streams, its very high operating temperature (with consequent potential to produce hazardous by-products such as chlorodibenzodioxins and chlorodibenzofurans) and the need for treatment of the off-gases produced.

To provide data on the effectiveness of plasma arc technology on EDS-1-type neutralent, the Army tasked Stone & Webster, Inc., to conduct tests with MGC Plasma AG in Switzerland using that company’s PLASMOX technology to destroy H and GB neutralent simulants (Table 3–1). The preliminary results indicate that the PLASMOX system successfully destroyed a simulated neutralent containing 52 percent water. The initial review of the gaseous effluents indicated that even though low concentrations of dioxins and furans were detected, the system should not have any difficulty in satisfying environmental regulations (Stone & Webster, 2001).

The high temperatures in the reactor vessel are expected to destroy the residual energetics or energetic hydrolysis products in the neutralent. PLASMOX technology has been successful in removing metal ions (including heavy metals such as arsenic) and having them reform into a slag that can be readily disposed of. The data on the fate of the phosphorus during the test are still being evaluated. The unit that was tested was not a commercial unit and needs to have some safety features added to meet U.S. health and safety criteria.

Some reviews of destruction technologies for CWM suggest a greater degree of public acceptance for plasma arc technology as a viable alternative to incineration than the committee had previously perceived. Plasma arc was accepted as an alternative by the ACWA Dialogue group (NRC, 1999b), and the Stone & Webster citizen panel indicated that plasma arc technology might be accepted in some locations and not in others (Stone & Webster, 2001). At least one commercial plasma arc unit has been permitted, according to information provided by the Army (Edward Doyle, Office of the Project Manager, Non-Stockpile Chemical Materiel, communication to the committee, July 10, 2001). These observations, combined with the promising test results discussed above, suggest that PLASMOX may be a technically effective treatment technology for nonstockpile neutralents and a publicly acceptable alternative to incineration in some locations.

The committee concluded in its previous report that biological treatment is a doubtful candidate for treatment of MMD neutralent. Reasons included the relatively low destruction efficiency (typically around 90 percent for most compounds, as opposed to the five or six “nines” usually sought in destroying hazardous materials), the presence of compounds that are known to be difficult to destroy by biotreatment (e.g., chloroform, hexachlorobenzene, and hexachlorobutadiene), the fact that the process yields large volumes of off-gas that must be treated, and that the equipment is bulky and not easily transported. A previous test of biotreatment for GB and VX hydrolysate was unsuccessful (NRC, 2000).

These same disadvantages are expected to pertain to the biotreatment of liquid waste streams of the EDS-1. Although the presence of residual energetics or energetics hydrolysis products in the EDS neutralent should not be a problem (they are biodegradable), the presence of additional metal ions, especially toxic heavy metals such as arsenic, may be detrimental to the microorganisms. The presence of dichlorobenzene, pentachlorophenol, and chlorobenzene (Tables 2–1 to 2–3) in some test samples presents a problem, because all of these compounds are difficult to treat biologically. Overall, it is the committee’s conclusion that biological treatment is of doubtful use for the treatment of EDS neutralent.

As stated in NRC (2001a), solvated electron technology (SET) would not be an appropriate technology for treatment of MMD neutralents because of their high water content. The first step of the SET process is reduction of organic compounds with solutions of metallic sodium in anhydrous liquid ammonia. If used to treat EDS neutralents, the sodium would react violently with the aqueous component of the neutralent, causing a release of hydrogen gas, before reacting with any of the organic components. The large quantities of metallic sodium that would be needed, and the safety problems associated with the hydrogen, preclude the use of SET for EDS neutralents.

Table 3–2 summarizes the earlier evaluations of alternative technologies for processing EDS neutralents (NRC,

2001a). Based on the discussion above and the preliminary results of technology testing, the committee believes that the two-track approach it recommended for selecting treatment technologies for RRS and MMD neutralents remains valid for EDS liquid waste streams. However, the preferential ranking of technologies in the resource investment track (track two above) is modified as follows.

If neither the ACWA nor the commercially mature technologies can be used as is—if, for example, substantial process or permit modifications would be needed to dispose of nonstockpile neutralents—then the committee recommends that NSCMP should invest R&D resources first in further improvements in chemical oxidation and wet-air/O₂ oxidation. Only if these technologies cannot be adapted easily does the committee recommend that the Army consider investing resources in supercritical water oxidation (batch mode).¹³

In summary, the committee retains chemical oxidation and wet air/O₂ oxidation at the top of its preference list on the resource investment track (track two). Based on promising test results of plasma arc technology on simulated EDS neutralents and the possibility of public acceptance, the committee judges that advancements in plasma arc technology confirm its position on track one (existing commercial technology), thus removing it from track two (requiring R&D investment).¹⁴ SCWO (batch mode) is retained in track two, since it promises to address many of the problems associated with continuous SCWO.

Electrochemical oxidation with Ag(II) is dropped from the second track because it seems a poor choice for posttreating the chloride-rich neutralents derived from HD, CG, and L agents, which are likely to be involved in most

EDS applications. In addition, new test data from the ACWA Program indicate that polynitroaromatic compounds (e.g., trinitrobenzoic acid and trinitrobenzene) precipitate during treatment of liquid wastes containing explosive residues, which may be present in EDS neutralents (Winkler, 2001). Electrochemical oxidation with Ce(IV) is also dropped from track two, owing to its relative immaturity and the possibility of corrosion and plugging of the electrochemical cells as a result of higher levels of metal ions in the EDS neutralents.

Supercritical water oxidation (continuous mode) is dropped from track two because the characteristics of the EDS neutralents (high chloride content, higher metal concentrations) are likely to exacerbate ongoing problems with corrosion and plugging. The committee found no reason to change its low ranking of gas-phase chemical reduction (GPCR), because initial testing showed that a variety of undesirable by-products were formed and because the addition of equipment to capture toxic heavy metals such as arsenic and mercury would add significant cost and complexity. Biotreatment and solvated-electron technology (SET), which were judged inappropriate for treatment of MMD liquid wastes, are similarly inappropriate for EDS liquid wastes.

To summarize, the committee recommends that available R&D resources be focused on development of chemical oxidation, wet-air oxidation, and supercritical water oxidation (batch mode). Similarly, it recommends that no further resources be spent on development of electrochemical oxidation, supercritical water oxidation (continuous mode), GPCR, biotreatment, or SET for this purpose.

The biggest data gap relates to the fact that, at this writing, only GPCR had been tested for its ability to destroy actual EDS neutralent. Through its alternative technology testing program, the Army plans to use actual neutralent in some of its tests, but for most tests it will use simulated neutralent. The committee recognizes that there may be a number of good reasons for using simulated neutralent in technology testing, particularly for early testing trials. Actual EDS neutralent may not be readily available, and there may be regulatory issues associated with using actual neutralent during testing (see Chapter 4). However, the committee feels that the credibility of such testing would be significantly enhanced if actual EDS neutralent were to be used.

A complementary approach that could increase confidence in the technology test results would involve the use of tracking compounds.¹⁵ A key factor in the evaluation of any alternative treatment technology is its ability to either destroy the chemical compounds of greatest concern or to change them into a form that allows them to be easily removed or immobilized. Typically, a number of tracking compounds are identified and their fate is examined. Tracking compounds are chosen based on regulatory concerns as well as their resistance to destruction, with the understanding that if the tracking compounds are adequately destroyed or immobilized, the compounds of concern will also be destroyed or immobilized. Some potential tracking compounds—one example might be methylphosphonic acid (MPA), which is stable and difficult to destroy—are identified in Appendix G, along with a discussion of the concept.

Tracking compounds are used to evaluate incinerator performance, and a similar approach would be desired in the regulatory approval process of any alternative treatment technology (see Chapter 4). In addition, an understanding of the thermodynamics and kinetics of the destruction of the tracking compounds is highly desirable, in order to optimize process conditions.

As the Army proceeds with its Technology Testing Program, it would be desirable for it to identify and collect information on the fate of appropriate tracking compounds to better evaluate the performance of the technologies.

Finding: Neutralents from the EDS are similar to those from the MMD owing to similar treatment chemistries. However, there could be three differences:

• The potential presence of residual explosives or explosive-derived organic compounds in the EDS neutralents and rinsates. The MMD and the EDS produce different liquid wastes because of the different ways that munitions are processed. The MMD does not process items containing explosives, while the EDS can handle munitions containing bursters and/or fuzes. The EDS also uses explosives to open the munition and detonate any explosives contained therein.

• Potentially higher concentrations of dissolved or suspended metals (e.g., Hg, Pb, Cu, and Al) in EDS neutralents and rinsates owing to explosive accessing of the munition and/or the presence of fuzes or bursters. The fragmentation of the munition bodies may expose more metal surface to the monoethanolamine (MEA) reagent,¹⁶ which is a good extractant for some of these metal ions. In addition, the detonator materials in fuzes, such as lead azide and mercury fulminate, may introduce these highly toxic metal ions.

• The potential presence of arsenic compounds in the EDS neutralent from a small number of munition fills (the MMD was not intended to treat agents containing arsenic).

Finding: The fills expected to be processed most frequently in the EDS are sulfur mustard (H, HD), phosgene (CG), and—to a lesser extent—sarin (GB). Items filled with other agents—such as lewisite (L), which contains arsenic, or the nerve agent VX—are expected to be encountered much less often, but they do exist in the nonstockpile inventory.

The Army has conducted operational testing of the EDS only for munitions containing H, CG, and GB. Thus, the committee’s analysis focused mainly on the liquid waste streams resulting from EDS treatment of these three types of agent. However, because lewisite munitions are known to exist in the inventory and may be treated in the EDS, the committee also considered in its analysis the effect of high concentrations of arsenic compounds.

Finding: If agents containing arsenic (such as lewisite) are processed in the EDS, additional treatment steps will be needed to remove arsenic from the EDS neutralent or reduce its mobility in treated solids. In these rare cases, however, suitable treatment chemistries are known and have been demonstrated to be effective (Utah, 1998; McAndless et al., 1992).

Finding: The EDS neutralization process and subsequent water rinses produce four liquid waste streams in two categories: (1) organic-rich liquids consisting of the neutralent and a reagent-based rinse and (2) cleaning solutions and final washes containing relatively low concentrations of organics.

Recommendation: The committee recommends that the Army consider separate treatment strategies for organic-rich liquids and these other aqueous liquids, since their chemical properties and regulatory status are different.

Finding: Chemical analyses of EDS neutralents and rinsates obtained from testing of HD, CG, and GB in the EDS may not have accounted for some species, such as energetic compound decomposition products, that may be encountered during operations.

Recommendation: The Army should review the sampling and analytical techniques employed at Porton Down and at RMA to ensure they are sufficiently sensitive and complete to detect any species of agent, energetics, and other components that could be in concentrations high enough to be of concern to human health or the environment.

Finding: The two-track approach¹⁷ recommended for selecting treatment technologies for RRS and MMD neutralents in the committee’s previous report (NRC, 2001a) remains valid for EDS liquid waste streams. However, based on new and preliminary results of NSCMP’s Technology Test Program, as well as results of tests on some of the technologies obtained in the Army’s Assembled Chemical Weapons Assessment (ACWA) Program, the preferential ranking of technologies in the resource investment track has changed, as described in the following recommendation.

Recommendation: The NSCMP should pursue a two-track strategy to identify a suitable treatment technology for EDS liquid waste streams. As part of the first track, the NSCMP should take advantage of available equipment that would require little or no investment, that is, it should piggyback on alternative technologies from the ACWA Program or on existing commercial technologies, such as chemical oxidation, wet-air/O₂ oxidation, or plasma arc¹⁸ technology. The committee judged that if any of these existing and available technologies can accomplish the task safely, it should provide a relatively rapid and inexpensive course of action.

If, on the other hand, none of the existing and available technologies can be used as is—for example, if substantial research and development resources would be needed to adapt them to the destruction of nonstockpile neutralents— the committee recommends that NSCMP, as part of track two, should invest first in chemical oxidation and wet-air/O₂ oxidation. Only if these technologies cannot be adapted easily does the committee recommend that the Army consider investing R&D resources in supercritical water oxidation (batch mode).¹⁹

The committee recommends that no further resources be spent on development of electrochemical oxidation, supercritical water oxidation (continuous mode), gas-phase chemical reduction, biotreatment (by itself), or solvated electron technology for this purpose.

Finding: The Army has an ongoing program to test several alternative technologies for their ability to destroy EDS neutralents. Based on information provided to the committee regarding this test program, the committee has several concerns:

• The tests are often conducted with simulated EDS neutralents mixed from laboratory chemicals rather than with actual EDS neutralents.

• The Army does not appear to have identified key tracking compounds²⁰ that are the most difficult to destroy and whose disposition can serve as indicators for the performance of the treatment technologies.

• The test program does not appear to be designed to provide basic data on the kinetics and thermodynamics of the oxidation of key waste stream components under process conditions.

Recommendation: The test program could be improved if the following steps are taken:

• To the extent feasible, the Army should use a representative range²¹ of actual EDS neutralents obtained from munition destruction in its tests of alternative treatment technologies.

• A limited number of tracking compounds chosen for their ability to gauge process performance and issues of regulatory concern should be identified and analyzed for in the treatment effluent.

• To supplement tests on EDS neutralents, the Army should collect information about the kinetics and thermodynamics of the destruction of these tracking compounds by the preferred destruction technology.

• Physical properties of neutralents such as phase behavior (including suspended solids) and flash point should be determined on neutralent samples obtained from EDS-1 and EDS-2 treatment of actual chemical munitions.