Neutralization Technology for Nerve Agent VX
The neutralization of VX nerve agent, like the neutralization of HD described in Chapter 7, can be carried out under mild conditions to give products with greatly reduced toxicity. If alkaline reagents like aqueous NaOH are used, the hydrolysis conditions for HD and VX are similar. The reactions can be carried out in commercially available chemical reactors at temperatures below 100°C and near atmospheric pressure. However, the conditions for hydrolyzing VX with neutral water differ from the conditions for HD. Hydrolysis of VX also results in a different set of reaction products than does hydrolysis of HD, and therefore different subsequent treatment is required prior to disposal (Yang et al., 1995).
The Army has explored several approaches to the hydrolysis of VX, including neutralization in a reactor vessel and an innovative in situ reaction in which the nerve agent is treated with a small amount of water while still in the original storage container (Brubaker et al., 1995). The research has opened some potentially attractive options for the detoxification and disposal of VX (U.S. Army, 1995b). Although the processes for VX neutralization are not as well developed as they are for HD, they have good potential for the safe, timely, and cost-effective disposal of the nerve agent.
In April 1996, the Army selected a process based on alkaline hydrolysis of VX as its preferred candidate for development (U.S. Army, 1996f). This neutralization process is closely analogous to the process for HD except that the reaction conditions are alkaline rather than neutral to acidic. The reasons for choosing this process rather than the superficially attractive in situ process are discussed in the Technology Status section below.
If this technology is selected for pilot demonstration, the TPC states that the neutralization products will be shipped to a commercial TSDF that uses biological oxidation for further treatment prior to final disposal. If further biological treatment at a TSDF is not available, other forms of treatment at commercial TSDFs may be evaluated.
As is true for the neutralization technology for HD in Chapter 7, the Army Alternative Technology Program has been the proponent for the VX neutralization technology for the purposes of the AltTech Panel's study. Therefore, the Alternative Technology Program will be referred to throughout the remainder of this chapter as the TPC.
Although the neutralization of VX with aqueous NaOH resembles the first part of the HD treatment described in Chapter 7, there are significant differences in detail. The current version of this technology is sketched in Figure 8-1. Removal of the nerve agent from ton containers will use the same punch-and-drain process described for mustard agent in Chapter 7, although no solid heels have been found during nonintrusive testing of the VX containers stored at the Newport site.
The drained agent is transferred to a holding tank. From the holding tank, the VX is fed slowly through a recirculation loop with an in-line static mixer to a vigorously stirred reactor containing hot (about 90°C) aqueous sodium hydroxide solution (20.6 wt pct). The total amount of VX added to the reactor is equal to 21 wt pct of the hydrolysate prior to addition of sodium hypochlorite. The mixture is heated for approximately six hours to destroy the VX and a similarly toxic by-product present in trace amounts, labeled EA-2192. After cooling, an equal volume of dilute (5 wt pct) sodium hypochlorite (bleach) solution is added to oxidize a malodorous reaction product and make the hydrolysate more amenable to subsequent biological treatment. After sodium hypochlorite is added, the amount of VX processed is equal to 10 wt pct of the final hydrolysate. The hydrolysate will be analyzed to ensure that the concentrations of both
agent and EA-2192 are below 20 ppb1 before release from the toxics control area.
The hydrolysate is to be shipped for further treatment and final disposal to a commercial waste treatment facility that uses biological treatment to biodegrade organic contaminants. The products could also be treated by incineration (federally owned or commercial), by other existing treatment technologies at
The process for cleaning ton containers resembles the process for HD containers, although no solid residues are anticipated with VX. Drained ton containers are rinsed with hot water to dissolve residual agent. The resulting solution is drained and the rinsing process repeated. The container is then cut open, steam cleaned, and tested for the presence of agent vapor. If no agent vapor is detected by a standard ACAMS monitor, the container is packaged and shipped to Rock Island Arsenal, Illinois, for metal reclamation, as proposed for HD containers (U.S. Army, 1996f).
The process for cleaning VX ton containers was demonstrated using a container from which agent had
The TPC's concept design package (U.S. Army, 1996f) presented the VX neutralization process as described in this chapter. The design included treatment of the hydrolysate with sodium hypochlorite to (1) reduce the odor associated with Thiol, (2) enhance the biodegradability of the hydrolysate, and (3) destroy any residual EA-2192. Because the EA-2192 hydrolyzes, although more slowly than VX, the third reason was no longer important after the duration of the hydrolysis procedure was extended to six hours.
Recently, the TPC has reconsidered this post-hydrolysis treatment option because VX has been detected at concentrations of several parts per million in some upper-layer samples of hypochlorite-treated hydrolysate (Lovrich, 1996). The VX appears to reform during the hypochlorite treatment, since it is not detected (at a detection limit of 860 ppb) in untreated hydrolysate prior to the addition of hypochlorite. The TPC now intends to eliminate the hypochlorite treatment option and instead ship the hydrolysate to an off-site TSDF, either without further treatment or with the addition of isopropanol. Adding isopropanol converts the hydrolysate to a single phase for ease of shipment and enhances its biodegradability by providing additional carbon.
The panel believes that problems associated with the disposal of VX hydrolysate can be resolved in a timely manner. However, the recommendations in Chapter 11 provide alternative disposal options in the event that shipment of hydrolysate to an off-site TSDF is not a viable option. The panel encourages the TPC to make additional efforts to control the hydrolysate odor because it could cause significant concern to the public in the event of a spill during handling or transportation. The TPC can draw on the experience of TSDFs that routinely handle malodorous organosulfur compounds.
been drained. A 3X condition was achieved after 2 hours of spraying with high pressure hot water and steam (U.S. Army, 1996f).
The neutralization processes evaluated for disposal of VX involve hydrolysis of the P-S bond, which is essential to the toxicity of this nerve agent. Figure 8-2 presents the reaction scheme for hydrolysis mediated by sodium hydroxide. The reaction with sodium hydroxide produces the relatively nontoxic ethyl methylphosphonic acid (EMPA), which is present as its sodium salt, and an aminothiol compound (Yang, 1995). The aminothiol, which has a very unpleasant odor but low toxicity, is often referred to as "Thiol," as it will be here (not to be confused with methyl mercaptan, which is also called "thiol"). Much the same reaction occurs during hydrolysis with neutral water, but the resulting EMPA is present as the corresponding acid rather than the sodium salt.
A major advantage of the alkaline hydrolysis process became evident during neutralization studies on impure munitions-grade VX that contained small amounts of a compound containing two P-S bonds. This material (known as "VX-bis"2) reacts with water to form EA-2192 (which is present as the sodium salt under alkaline conditions). EA-2192 is almost as toxic as VX itself and is resistant to further hydrolysis by water alone. The concentration of EA-2192 is low, but it contributes significantly to the toxicity of the hydrolysate. During hydrolysis mediated by sodium hydroxide, EA-2192 is also hydrolyzed by an analogous reaction to form thiol and the sodium salt of methylphosphonic acid (MPA), which has low toxicity. At low temperatures (20°to 25°C), small quantities of EA-2192 may be present for a prolonged period because of slow reaction rates. At higher temperatures (75°C to 90°C), both VX and EA-2192 hydrolyze at acceptable rates to form relatively nontoxic products. The alkaline hydrolysis of VX is exothermic, releasing 32.3 kcal/mole.
The products of VX hydrolysis mediated by sodium hydroxide form two liquid phases. The large, dense, aqueous layer holds nearly all (98 mole-percent) of the phosphorus-containing products (predominantly sodium salts of EMPA and MPA) and about 80 percent of the sulfur-containing products, largely the sodium salt of Thiol. The small upper phase contains the rest of the Thiol and its secondary products, along with a mixture of compounds derived from the stabilizer (diisopropylcarbodiimide) added to the agent during manufacture. When the crude hydrolysate is treated with sodium hypochlorite to destroy the malodorous Thiol, the product continues to have two phases, but the nature and distribution of the sulfur-containing products changes significantly. The Thiol largely disappears from the bottom layer, and its concentration in the upper layer sharply decreases. With a small excess of bleach, the Thiol is largely converted to the disulfide, which becomes the major component of the upper layer. As the amount of bleach is increased, more of the Thiol is converted to the corresponding sulfonic acid, which appears in the lower layer as its sodium salt, along with the sodium salts of MPA and EMPA.
The neutralization technology chosen for destroying VX agent stored at Newport was developed on the basis of previous experience and ongoing research. Although not much has been reported about the alkaline hydrolysis of VX, caustic (sodium hydroxide) has been used to destroy GB (Sarin) nerve agent on a substantial scale. Between 1973 and 1976, the U.S. Army destroyed 4,188 tons of GB at the Rocky Mountain Arsenal by treating it with aqueous sodium hydroxide (NRC, 1993; Flamm et al., 1987). United Nations teams used similar processes to destroy about 70 tons of GB-based agents in Iraq in 1992-1993 (NRC, 1993). The lessons learned from those operations facilitated the development of a VX hydrolysis process through research at Aberdeen by the TPC since 1993.
The VX research was carried out in three stages: (1) laboratory-scale scouting to establish reaction conditions for complete destruction of the agent; (2) process optimization studies in 2- and 12-liter Mettler reactors designed to acquire precise thermodynamic and kinetic data; and (3) bench-scale testing in a 114-liter (30-gallon) stirred tank reactor previously used for HD hydrolysis (described in Chapter 7). Parallel research was conducted on VX hydrolysis under neutral conditions (described below) and on the alkaline hydrolysis method that was ultimately chosen for development. The alkaline hydrolysis was initially tested on a small-scale in the laboratory but was extended to testing in 1-liter glassware in runs that destroyed up to 265 g of agent at a time. These tests established satisfactory processing conditions and provided hydrolysate for developing new analytical procedures, as well as for toxicity testing. A new analytical procedure based on sequential liquid chromatography and mass spectrometry permits detection of both VX and EA-2192 at levels of 10 ppb in aqueous solution (U.S. Army, 1996f). As described below in Agent Detoxification, a 40,000-fold reduction in toxicity is accomplished by alkaline hydrolysis.
The bench-scale studies in Mettler reactors yielded reliable data on heats of reaction and reaction rates. The tests monitored the disappearance of VX and secondary products, such as EA-2192, as well as optimizing the test conditions for the 114-liter reactor. Five tests using the 114-liter reactor destroyed 25 to 30 kg of VX in a typical run, but neutralization of as much as 39.4 kg was demonstrated (Lovrich, 1996). More important, the effects of reaction times and mixing (e.g., stirring rate and the effect of adding a static mixer) were evaluated on a large enough scale to extrapolate to pilot- or production-scale reactors. The basic data generated in these tests will facilitate the design of the reactors to be used at the Newport site. More than 351 kg of VX was destroyed in these bench-scale studies.
The bench-scale studies demonstrated the effectiveness of alkaline hydrolysis and provided valuable operating experience under conditions similar to those proposed for full-scale operations. In addition, the tests yielded large volumes of hydrolysate for biodegradation studies and for testing at a TSDF that was being
considered for the off-site biotreatment of hydrolysate. About 7,000 pounds of hydrolysate from the experiments on alkaline and neutral VX hydrolysis, which were performed at Aberdeen, was delisted under Maryland regulations and shipped off-site for disposal at the TSDF. The treatment at this facility seems to have been satisfactory, but additional studies of the treatability of the alkaline hydrolysate are being carried out at several potential disposal facilities (U.S. Army, 1996c).
In Situ Neutralization
The concept of detoxifying VX in its storage containers (hence referred to as in situ neutralization) is the subject of ongoing research and development by the Army (U.S. Army, 1996f). For in situ processing of VX, a small amount of water (7 to 10 percent) is injected into the storage container. The water reacts over a period of days or weeks and hydrolyzes the VX to a product mixture with substantially lower toxicity. In principle, the viscous liquid product can then be prepared for disposal at a commercial TSDF. On first view, this approach appears attractive for the following reasons:
Handling agent may not be necessary because the neutralization occurs in the agent storage containers.
Substantial reduction of the toxicity of the stockpile could be achieved rapidly because all the ton containers could be treated in rapid succession with little lead time.
The process is conceptually simple, requiring minimal processing equipment and capital costs.
Upon closer evaluation, the potential advantages of in situ neutralization have not been realized because of three obstacles. First, additional handling of agent is in fact required because the ton containers do not have sufficient excess capacity to hold the necessary amount of water for reaction without first removing a significant volume of agent (a few gallons of agent out of about 180 gallons in an average container). Removal of several gallons of agent is probably a simple operation because the valves on the ton containers stored at the Newport site appear to be operable. Even so, the need to transfer agent entails a small (but controllable) risk to plant personnel. Second, although the in situ process does substantially reduce the agent toxicity, further treatment is required to destroy residual EA-2192 completely and reduce toxicity to levels suitable for disposal at a commercial TSDF. Third, the in situ reaction is difficult to control because of poor mixing and inadequate temperature control within the ton containers.
At the time of the panel's evaluation, in situ neutralization of VX had been tested on three ton containers. Several difficulties were encountered during this testing. The low solubility of water in VX resulted in poor reproducibility during preliminary tests. The reaction proceeds slowly for some time after adding the water. As the reaction proceeds, the solubility of the water in VX increases because the initial hydrolysis product is a mutual solvent. In the first full-scale test of the reaction, which was conducted without heat or agitation, the VX and water initially formed two layers in the cylinder. After about three days, the water dissolved, and a rapid reaction ensued. The temperature inside the cylinder rose to a maximum of 98°C, and the VX concentration decreased to 5.4 percent at the end of the first week. Subsequently, the temperature gradually fell; less than 400 ppb of VX remained after four weeks.
Another potential difficulty was found in the second test of in situ neutralization. A change in the procedure for adding water led to an excessively rapid reaction and the formation of solid hydrolysis products that might be difficult to remove from the storage cylinder.
Two approaches to dealing with the immiscibility of VX and water are possible. The simplest is to mix the liquids by rolling the cylinder. The second approach is to add some hydrolysate from a previous VX hydrolysis to the ton container, along with the water needed to react with the VX. The advantage of the second approach is that it produces a smooth, rapid reaction. The major disadvantage is that it requires additional handling of agent and hydrolysate because more agent must be removed from the container to accommodate the volume of the added hydrolysate. The additional handling of the hydrolysate is disadvantageous because the liquid is still toxic, although a thousandfold less toxic than VX, based on intravenous testing in mice. Most of the residual toxicity appears to be due to the presence of EA-2192.
As a consequence of these complications in the in situ process, the TPC elected to concentrate further development work on the alkaline hydrolysis carried out in a conventional reactor. Testing of the in situ process in ton containers continues at the Chemical Agent Munitions Disposal System facility in Utah. Based on the difficulties described above, the AltTech panel does not
recommend additional tests on the in situ process beyond the tests currently in progress.
Operational Requirements and Considerations
Mass and Energy Balances
The TPC provided flow sheets indicating the major equipment, piping, and controls, as well as material and energy flow rates (U.S. Army, 1996f). These flow sheets were derived from operating experience gained in bench-scale testing.
A simplified block flow diagram and corresponding overall mass balance for the VX neutralization process using sodium hydroxide (NaOH) and sodium hypochlorite (NaOCl) are presented in Appendix H. Overall, approximately 8 kg of water, 0.4 kg NaOH, and 0.7 kg of NaOCl will be required per kilogram of VX neutralized. Energy requirements for this process are estimated to be 28,600 kJ/kg of VX neutralized, including 14,000 kJ/kg for steam heating, 9,500 kJ/kg for electricity, and 5,100 kJ/kg for cooling.
Draining, Cleaning, Packaging, and Shipping Ton Containers
Draining, cleaning, and decontaminating the ton containers can be done with the same system described in Chapter 7 for HD containers. The punch-and-drain system is essentially identical to the proven JACAD system (NRC, 1994a). Empty ton containers are cleaned, cut open, decontaminated to a 3X level, and packaged for shipment to Rock Island Arsenal, Illinois, a government metal recycling plant. The principal difference in this operation between the HD and VX facilities is the offgas scrubber. For VX, the caustic scrubber used for HD is replaced with a two-stage scrubber that provides both acid and alkaline scrubbing.
No major difficulties are foreseen with this operation, although some contamination of the equipment and surrounding areas can be expected from the high pressure spray system. Although no heel is expected in VX ton containers, high pressure spray decontamination is used to ensure that no agent remains in microscopic crevices in the container surface. It is difficult to predict the rates of use of water and decontamination solution, the requirements for their subsequent interim storage, and their dilution effect on the neutralization system. The method for treating water and decontamination solution, either separately or with agent batches, has not been defined. The sodium content of the spent caustic solution is not expected to be of concern to either the neutralization or subsequent treatment processes because it replaces some of the sodium hydroxide required by the agent neutralization process. Current monitoring methods developed for the baseline system should be adequate for these operations.
Agent Storage System
The drained agent is pumped to an interim holding and surge tank system, analogous to the proven JACAD system. The capacity could easily be adjusted for local external hazard conditions as needed. Gases vented from these tanks pass directly through a carbon filter bed located in the storage area. Consideration should be given to processing these vent gases through the scrubber prior to carbon filtering and release, along with the vent gas from the neutralization reactor.
The stored agent is fed in batches to one of two independent neutralization trains. Each train consists of a 2.5-m3 (650 gallon) stirred neutralization reactor with both internal and external mixers, overhead offgas condensers, a reactor cooling jacket, and an external heat-exchanger cooling system; a 5-m3 (1,320 gallon) hydrolysate storage tank with mixer; a 152-m3 (40,000 gallon) waste storage tank; and an offgas treatment system.
The neutralization reactor is first partially filled with 11 wt pct caustic and brought up to the operating temperature of 90°C (194°F). The VX is then slowly added in the external recirculation loop just ahead of the static mixer. The mixture is heated for about six hours, as required to destroy the toxic EA-2192 by-product. The temperature is controlled by removing the exothermic heat of reaction through the reactor cooling jacket, the heat exchanger in the external cooling system, and an offgas reflux condenser.
As liquid agent is added to the reactor vessel, the overhead gases are vented through the reflux condenser to condense water vapor and volatile organic
compounds generated during the reaction. The condensate is recycled to the reaction tank. The noncondensable gases pass through a dual scrubbing system. The first scrubber contains acid to absorb organic amines produced by the reaction. The second contains caustic to neutralize the acids formed in the first. A heat exchanger removes the heat of neutralization. The gas then passes through a chiller to reduce the water vapor content, a gas heater to elevate the gas temperature above the dew point, and an activated carbon filter system.
The hot hydrolysate is analyzed for residual agent and, if acceptable, is transferred to the hydrolysate storage tank, where it is combined with water and hypochlorite to oxidize some of the organic products. The primary reasons for adding hypochlorite are to reduce the foul odor of the hydrolysate and to make the hydrolysate more amenable to subsequent biological treatment. Upon completion of the hypochlorite oxidation, the waste is analyzed to ensure complete agent destruction and pumped to an external storage tank, where additional water and hypochlorite are added to prepare the effluent for off-site disposal.
Laboratory testing of the biodegradation of VX hydrolysate has been of limited success to date. The products of hydrolysis do not readily serve as the primary substrate for biological oxidation. Substantial quantities of co-substrate (i.e., other waste with a high-carbon content but low in phosphorus) are required to force the microbial utilization of phosphorus from the methyl phosphonic acid present in the hydrolysate. Because of this need for high-carbon cofeed and because only limited success has been achieved in biodegrading Thiol, the hydrolysate is not a good candidate for treatment by on-site biodegradation prior to final disposal.
The very limited data available as of May 1996 suggest that off-site biodegradation is likely to succeed if the treatment facility receives sufficient quantities of high-carbon waste from other sources to force microbial degradation of the VX hydrolysate products as a source of nutrient phosphorus. Laboratory testing of biodegradation with SBRs has demonstrated significant biodegradation of organophosphonate and organosulfate constituents in VX hydrolysate (U.S. Army, 1996k). When the VX hydrolysate was the only phosphorus source available to the microbial population and isopropanol was provided as an additional carbon source (at about 2,900 mg TOC/liter), greater than 90 percent of the organophosphonate constituents and up to 51 percent of the organosulfate constituents were biodegraded. These results were obtained in laboratory-scale reactors operating in semibatch mode (periodic partial decanting of clear supernatant and removal of settled sludge, followed by refilling) over extended intervals. Hydrolysate biodegradation with a carbon cofeed has not yet been tested at bench-scale.
Because of limitations in the available information, the panel is concerned that off-site treatment to date may have involved primarily dilution of the hydrolysate to an acceptable level rather than complete destruction by biodegradation of the products of concern. However, the preliminary toxicity testing described in the next section suggests that oxidized hydrolysate (VX hydrolysate after being treated with sodium hypochlorite or a similar oxidizing agent) may have sufficiently low toxicity that further degradation of organic constituents is not needed.
If further toxicity testing demonstrates that the hydrolysate poses no threat to human health or the environment, total biodegradation of the organic components during disposal at a TSDF may not have to be demonstrated. Otherwise, the alternative of off-site treatment by biodegradation at a TSDF will require appropriate treatability studies to substantiate that complete biodegradation of the hydrolysate constituents does in fact occur. Such treatability studies would need to be conducted at the TSDFs that are candidates to receive the hydrolysate. The presence of Thiol and methylphosphonic acid derivatives, which are scheduled precursors under the CWC (Chemical Weapons Convention), would subject a TSDF receiving the hydrolysate to destruction verification requirements (including inspection) under the terms of the CWC.
Table 8-1 provides a summary of toxicity testing carried out on hydrolysates from VX neutralization with water and VX neutralization with sodium hydroxide. Intravenous exposure testing was performed on mice. An LD50 is the dose required to kill 50 percent of the test population within 24 hours. The LD50 for VX is included for comparison. Neutralization of VX with sodium hydroxide results in greater than a 40,000-fold reduction in toxicity, compared with a 970-fold reduction achieved by neutralization with water only.
TABLE 8-1 Toxicity of VX and VX Hydrolysates as Measured by 24-Hour Intravenous LD50 in Mice
VX hydrolysate (water only)
VX hydrolysate (NaOH (aq))
LD50 at 24 hours (mg substance per kg body weight)
As discussed earlier, laboratory and bench-scale tests have shown that the primary neutralization process can destroy the chemical agent to less than 20 ppb in the hydrolysate. The Army has considered a residual of less than 20 ppb VX in hydrolysate to be safe enough for release of the hydrolysate from the on-site toxics control area and transport off-site for final treatment. The basis for defining this level as acceptable and its consistency with the 3X release standard for solid materials needs to be demonstrated. The Army also needs to define the standards to be used for transporting and disposing of the hydrolysate, as well as any related restrictions that would limit the pathways for human contact with the hydrolysate. The panel believes that defining these standards will not seriously constrain the off-site disposal options or the disposal schedule because significant quantities of hydrolysate have already been approved for off-site treatment and successfully shipped to an off-site TSDF. These standards and restrictions are therefore not anticipated to impede the successful operation of a VX neutralization facility.
The operations of punching and draining ton containers and then decontaminating and packaging them for off-site shipping constitute a batch process. Although it might be conducted on an eight-hour-per-day basis, the current size limit on the agent storage tank is likely to require a close coupling of the container draining operation with subsequent agent processing. Thus, operations are anticipated to be conducted 24 hours per day, 7 days per week, except for scheduled maintenance periods.
The primary neutralization process is a semibatch process that requires about 8 hours per batch. Thus, although it would probably be more economical to operate this part of the plant around the clock, it could be operated on an 8-hour-per-day basis. Currently, the TPC anticipates operating the neutralization process 24 hours per day, 7 days per week, except for scheduled maintenance periods.
The semibatch mode of operation requires simpler process controls than a continuous operation and reduces the probability of control failure and resultant process upset. In addition, the semibatch mode permits the operators to hold the hydrolysate for confirmation of complete agent destruction.
Emergency Startup and Shutdown
Handling Ton Containers
Draining, washing, cutting, decontaminating, testing, and shipping are sequential steps in a mechanical handling system with no significant surge capability between steps. The steps are started in sequence, and the sequence must be maintained as a mechanical production line. Shutdown simply involves stopping the first step and allowing the in-process containers to continue to completion; normally this is followed by decontamination of the work area. These procedures are common to the front-end systems of all VX destruction technologies and reflect extensive experience at JACADS.
The primary neutralization process can be stopped in an emergency by halting the addition of agent or chemicals. Under these conditions, the reactions will continue until the reactants have been depleted. The maximum temperature rise that could occur from a loss of cooling water and cessation of agent feed is less than 10°C.
Commercial 50 percent sodium hydroxide solution is stored in a 28-m3 (7,400 gallon) tank and then diluted to 33 percent for feed to the neutralization reactor and to 18 percent for feed to the offgas scrubber and for
decontamination fluid. The TPC provided no estimate of the decontamination and gas scrubbing requirements. Based on processing agent at 3,200 kg/day, 3,300 kg/day of 40 wt pct caustic is required for neutralization. Treatment of the resulting hydrolysate before shipping requires 15,900 kg/day of 15 percent sodium hypochlorite solution, which is diluted for use to 5 percent concentration.
Process water is preheated to 90°C for use in the ton container cleanout systems, for which the estimated requirement is 19 liters per minute. Provision is made for using an abrasive in the water-jets, but no estimate of quantity was provided to the panel.
Processing Ton Containers
The ton container cleaning process produces a variety of liquid or slurry waste streams: (1) washdown solutions of contaminated hot water; (2) decontamination solutions, including spent solution and floor washings; and (3) water cutting slurry (contaminated water plus abrasive), which is mostly recirculated. These waste streams will be destroyed by adding them to the neutralization reactors. The residual caustic in them is taken into account as partial replacement for the caustic otherwise added for neutralization. Some of the washings may contain too many solids, which may be separately bagged for off-site shipment along with the ton containers or the hydrolysate. Because the quantities are not known, the exact methods of feeding the waste streams to the reactor can only be approximated and have not been developed.
Small solids, such as valve fittings and removal tools, are also bagged for off-site shipment with the ton containers. Vent gases will be ducted to the scrubber and carbon filter system. Building ventilation air will be treated by an HVAC system with carbon filters.
All liquids are shipped off-site. These liquids are mixed residuals that may not be characterized except for testing to ensure that the VX and EA-2192 concentrations are below acceptable limits. The quantity (including decontamination fluids) is estimated to be 58 metric tons per day at the average agent destruction rate.
All vent gases are treated in the scrubber and carbon filter system. All liquids and slurries proceed to off-site biotreatment.
Secondary wastes include personal protection equipment, rags, small metal parts, etc. These wastes are cleaned with decontamination solution, tested to ensure that they meet the 3X standard, and disposed of as hazardous waste.
Process Stability, Reliability, and Robustness
The neutralization reactions are mildly exothermic with removal of the heat of the reaction by a reactor cooling jacket, an external cooling system, and condensation of the offgases in a reflux condenser cooled by chilled water. Failure of the cooling system would cause a temperature excursion that is estimated to heat a full batch of agent and water from 90°C (1 atm gauge) to 98°C (1.1 atm gauge). The design pressure for the neutralization system is 6.8 atm gauge, but the system runs at 1 atm gauge. Thus, there should be no catastrophic thermal excursions.
The maximum agent present is a full load of 180 kg of VX. An upset of feed stream flow rates could cause minor changes in the reactions and less efficient agent destruction. This condition can be countered by holding up a batch until it has been checked and extending the reaction duration if necessary.
Reliability and Robustness
The system will use standard industrial components for which there is extensive good industrial experience. Materials of construction have not been proposed, but their selection probably does not present serious problems. Although feed rates are important, most of the process phenomena occur relatively slowly so that response time should not be critical.
Operations and Maintenance
All the unit operations included in the overall process have been extensively used in commercial operations with related proven equipment. There has also been significant prior experience with a similar method of agent neutralization, albeit the agent neutralized was GB
(NRC, 1993). An exception may be the analytical techniques, which may be newly developed, depending upon the disposition of process residuals and the related standards. As noted in the Technology Status section, the TPC has gained considerable operational experience with this neutralization process through repeated bench-scale tests using 114-liter reactors. If a contractor is hired to run the pilot plant and full-scale facilities, the contractor should have an established record of experience in operations on a similar scale with hazardous materials.
Maintenance requirements for all systems are similar to the requirements for the baseline technology systems. The TPC and the panel foresee no unusual problems. Maintenance manuals and documented procedures are not yet available for this process.
The equipment lifetimes should all exceed the duration of plant operations, with the exception of the replacement of small pumps and instrumentation. The downtime to replace conventional components is more likely to be governed by agent-related safety precautions than by failure of the equipment.
Experience at pilot-scale for this process does not exist. Thus it is not possible to provide an estimate of operational time versus downtime.
Two considerations are involved in scaling the process to a full-size facility: scale-up from the existing bench-scale work to the pilot plant and scale-up from the pilot plant to the full-scale facility.
Bench Scale to Pilot Plant
The following key unit operations are involved in the complete plant: (1) ton container processing, (2) reagent preparation and storage, (3) agent neutralization, and (4) optional hydrolysate oxidation with hypochlorite. The critical new unit operations are agent neutralization and hydrolysate oxidation. Although off-site treatment and disposal of hydrolysate is also a critical operation, it is considered external to on-site processing. Demonstrating its feasibility will require detailed treatability studies carried out by one or more off-site TSDFs that would be candidates to receive the hydrolysate. Off-site TSDFs appear to be available and are willing to receive the hydrolysate, so availability of a treatment facility is not a constraint.
Agent neutralization tests have been conducted in 114-liter reactors. This size will be extrapolated to the 2.5-m3 reactors for the pilot plant. The principal variables to consider in this extrapolation are likely to be mixing and heat exchange. No catalysis is involved. Heat exchange is not likely to be a problem if the mixing is adequate. Mixing is a concern because the reaction does involve intermediate reactions and requires the proper ratios of agent, caustic, and water to avoid producing undesirable residuals.
The oxidation of VX neutralization products by sodium hypochlorite is similar to current methods used by the Army to dispose of VX wastes. Although the method has not been piloted for this specific purpose and design details are yet to be worked out, neither the TPC nor the panel anticipates a major risk in scaling up this method for treatment of process wastes.
Pilot Plant to Full-Scale Facility
The TPC plans to develop the full-scale facility using the pilot plant neutralization reactors as one of two modules. The full-scale plant will require the addition of another line of equipment. Thus, scale-up of this technology is straightforward, except for scheduling and other matters. Although unit operations other than neutralization may or may not be modular, they are relatively standard unit operations that should not be difficult to scale.
The process safety risk factors for neutralization of VX are the same as for HD. The discussion in Chapter 7 applies equally well to this technology with respect to risk factors inherent in handling agent prior to
neutralization, risk factors inherent in neutralization operations, worker safety issues, and specific characteristics that reduce the inherent risk of the design.
Although a regulatory question concerning an Indiana statute remains to be resolved (see Chapter 9), the TPC has proposed a plausible schedule for destroying the VX stored at Newport using the neutralization process followed by off-site biotreatment. The TPC proposes that pilot-testing be done in one reactor line of a production-scale facility to be built at the Newport site. If the state of Indiana changes its legal requirements to permit this pilot-test, developing and implementing the process for neutralization and off-site biotreatment of VX could be accomplished on the following schedule. The remaining issues about off-site treatment of the hydrolysate by biodegradation should have no effect on the schedule because off-site TSDFs using other treatment methods, such as incineration, could handle the hydrolysate without difficulty.
A pilot plant design is expected to reach the design level required for a RCRA permit in December 1996. If a decision to pilot-test this technology at Newport is made in October 1996, permit applications would be submitted in April 1997. The TPC's plan allows two years for permit acquisition, which seems conservative considering the generally favorable reception of the neutralization technology by the Indiana regulators in preliminary discussions with the panel (see Chapter 9).
During the latter part of the permit acquisition period, a contract for construction of the facility would be let and orders would be placed for equipment items with long lead times for delivery. Construction is scheduled to begin on October 10, 1999, and to be completed about February 27, 2002. Initial systemization of a single production line is scheduled to take nine months. Pilot operations in this reactor train would be carried out until May 28, 2003, during which time systemization of the other reactor trains would start and a low-rate production operation would begin. Full-scale operations beginning in November 2003 would continue for about nine months. The production schedule assumes treatment of five ton containers per day, with the facility operating on a two-shift basis. Plant decommissioning and decontamination is scheduled to begin in late 2004 and is estimated to require one year.