Skip directly to: content | left navigation | search

Attachment #3

    Report for Congress: Review of the U.S. Army Proposal for Off-Site Treatment and Disposal of Caustic VX Hydrolysate From the Newport Chemical Agent Disposal Facility

    Review of the Toxicology and Health Hazard Considerations for Safe Management of Newport (Indiana) Caustic VX Hydrolysate
    By

    Agency for Toxic Substances and Disease Registry in collaboration with the Centers for Disease Control and Prevention Atlanta, Georgia
    November 3, 2004


    Review of the Transportation and Risk Management Provisions for Caustic VX Hydrolysate
    By

    Centers for Disease Control and Prevention in collaboration with the Department of Transportation
    November 3, 2004


    INTRODUCTION
    CDC prepared this report to analyze DuPont’s Transportation Safety Assessment and Risk Management Plan Safety, dated March 3, 2004. CDC considered this component of the response from two perspectives, described as follows:

    First, CDC determined whether the transportation plan is consistent with Department of Transportation (DOT) requirements for shipping hazardous materials from the point of generation—Newport, Indiana—to the point of final treatment and disposal—Deepwater, New Jersey. This determination differs from typical CDC reviews because of the different hazard characteristics and larger volumes involved; therefore, CDC requested and received assistance from DOT in conducting this part of the review.

    Second, CDC determined whether the safeguards, emergency planning, and other risk management considerations that will be applied to this proposed project are comparable to transportation of other potentially hazardous substances, such as recovered chemical weapons material (RCWM). Some of the criteria considered by CDC included route selection considerations, shipping containment provisions, emergency planning, and notification activities. CDC is conducting this analysis directly. Considerable overlap exists in the safety considerations required by DOT and the safety provisions considered by CDC in reviews of RCWM transportation plans.

    BACKGROUND

    The Newport Chemical Agent Disposal Facility proposes to treat agent VX with sodium hydroxide to produce caustic VX hydrolysate (CVXH) with no agent detected ≤20 parts per billion (ppb). This clearance criteria is equivalent to the Army’s drinking water standard for nerve agents for field use by soldiers, and CDC considers it appropriately conservative for use as a clearance criteria for shipment of waste.

    The CVXH can be characterized as being predominantly caustic and aqueous with a smaller organic fraction, the extent of which depends on the VX loading rate used in the batch process. Batch VX loadings of 8%, 16%, and 33% have been examined for the Newport facility. The current plan calls for plant startup using an 8% loading of VX stabilized with diisopropylcarbodiimide (DIC), and this is the only VX loading rate fully evaluated in this review. Please note that in this report, the more technically accurate term CVXH generally is used in place of Newport caustic hydrolysate or NCH.

    Other major by-products of interest in the caustic VX hydrolysate are ethyl methyl phosphonic acid (EMPA), methyl phosphonic acid (MPA), thiolamine, and EA 2192. EMPA and MPA are of interest because of their potential for persistence in the environment, and thiolamine is of interest because of its strong and disagreeable characteristic odor. As a general matter, EA 2192 exhibits nerve agent properties similar to VX. However, EA 2192 will be limited to ≤1 part per million (ppm) for a cleared batch of CVXH, a concentration deemed by CDC to be suitable for the risk management practices contained in this proposal. Toxicity considerations of these by-products are discussed in the full CDC report.

    DISCUSSION

    CDC considers four broad functional areas applicable to the proposed Newport CVXH transportation plan.

    • Packaging and Containment—The DuPont transportation plan discusses several options for the containment, including dedicated tank trucks and transportable tote containers (“ISO tanks”). The materials of construction and strength of the container design were considered, as were placement of valves, remote operability characteristics designed to minimize personnel potential exposure to tank contents, and vulnerability of the valves to bump hazards. DOT, in correspondence to CDC, noted that the plan “proposes to use equipment and procedures that go beyond what the regulations require for materials with the specific hazard and risk involved.”1

    • Personnel Qualifications—The transportation plan proposes use of two hazardous materials shippers that have “excellent safety records” as evidenced by “very low DOT recordable accident rates” and “very favorable DOT safety ratings.” Each of the two shippers reportedly maintains high qualification standards by employing experienced personnel who have passed rigorous background checks. Extensive training, including hazardous materials spill response, will be required of the drivers for this project. A team of two prequalified drivers will be used for each trip.

    • Route Planning—DuPont analyzed potential risk associated with four identified highway routes and one combined rail and highway route for transporting the CVXH from Newport, Indiana, to Deepwater, New Jersey. Factors considered included number, length, and duration of each trip; accident potential based on historic truck accident rates for each route; general population exposure potential for each route; potential environmental impact from accidental CVXH release for each route; and emergency response capability for each route. A commercially available risk analysis algorithm was used to quantitatively estimate total potential impact potential for each route option analyzed.

    • Emergency Preparedness—DuPont describes its Integrated Emergency Response Plan (IERP) used to support ongoing transportation incidents. A detailed specific emergency response plan would be developed for this proposed CVXH shipping plan and shared with appropriate state and local responders along the selected transportation route. DuPont also has IERP teams in place in Belle, West Virginia, and Deepwater, New Jersey, to serve as regional service centers to support incident responses if needed. In accordance with the IERP, these teams consult with and advise on-scene DuPont personnel and local emergency response personnel. As needed, additional on-scene advisors or response resources may be deployed.

    DuPont’s transportation analysis is predicated on the assumption that the CVXH poses a corrosivity hazard with no attendant nerve agent properties. Most transportation plans reviewed by CDC involve limited amounts of chemical warfare agents moved in one or a very limited number of moves. This plan differs both in the volumes of and predominant characterization of the material to be moved.

    CDC asked DOT personnel to review the DuPont transportation plan for overall consistency with DOT requirements for hauling hazardous materials. DOT determined the plan generally met or exceeded DOT requirements. However, DOT recommended that the shipping designation for the CVXH be reconsidered to reflect that it is a corrosive liquid, basic, inorganic, not otherwise specified, rather than the organic corrosive designation described in the plan. DOT’s review reflected DuPont’s characterization of the CVXH.

    In evaluating RCWM transportation plans, CDC also typically reviews agent air monitoring. Air monitoring for chemical agent before and after a move of RCWM is usually an integral part of a plan to detect any breech in containment so corrective action can be taken. For the CVXH, the Army and DuPont have stated that VX agent is required to be destroyed to ≤20 ppb to qualify for shipment.2 Because this clearance level would produce minimal safety hazard when compared with the corrosive nature of the CVXH, agent air monitoring for VX would not be useful and consequently was not included in the DuPont proposal.

    Batch processing studies indicate that, if VX survives, it would partition into the organic fraction of the caustic VX hydrolysate. The Army has stated that, at an 8%–16% VX loading, the organic fraction should be limited to approximately <1%–3% of VX hydrolysate. In the absence of mixing or agitation, the organic fraction separates, and layers on top of the aqueous component of the CVXH. At an 8% VX (DIC-stabilized) batch loading, the organic layer remains nearly indistinguishable from the much larger inorganic, aqueous fraction. The CVXH will be reprocessed if VX is detected above the MDL. However, the current sampling and analytical method used for process batch clearance does not attempt to evaluate potential VX in the organic layer of CVXH but instead evaluates the organic and aqueous components as a mixture.

    Examination of the impact of potential agent VX survival in the organic fraction of the CVXH requires estimation of an upper-bound level for the VX concentration within this fraction. On the basis of existing batch studies, CDC believes a reasonable upper-bound estimate is approximately 1–10 ppm of residual VX. This assumes a maximum of ≤20 ppb VX for the CVXH mixture and a VX loading of 8%. CDC noted, however, that one study showed a VX residual of approximately 2100 ppm in the organic layer (at a VX feed rate of 33%) of VX/sodium hydroxide (NaOH) batch hydrolysate,3 despite analysis showing that the hydrolysate mixture had ≤20 ppb VX. CDC contacted the lead author on this study to ask whether follow-up work was conducted to resolve and clarify this finding. Although recommended, the study was not repeated. The author believed, however, that this VX finding in the organic layer resulted from a sample mishandling in the laboratory and is not consistent with his laboratory’s other studies of VX/NaOH hydrolysate.

    A maximum credible event could involve a 5000-gallon tank truck or tote in an in-transit accident that ruptures the containment. If the above study result is the outlier it appears to be, then human exposure to VX at an estimated maximum of 1–10 ppm could occur with direct, unprotected contact with the organic fraction of the spilled material. The nerve agent effects of this level of VX and possible concurrent EA 2192 at the 1-ppm level are difficult to assess. However, to reach this maximum exposure to VX, the organic fraction (estimated at <0.5% by volume of the total contents for the 8% loading level CVXH) would need to remain undiluted from any mixing from the spill, which CDC believes is highly unlikely. Mixture and dilution of the organic fraction with the much larger aqueous fraction, to the extent that the corrosivity of the spilled material would present the most significant hazard, would be more likely.

    Inhalation exposure to VX vapor in a spill is believed to be negligible given its low initial assumed concentration in the CVXH and the relatively low volatility of VX. Because of the corrosivity of the bulk of the CVXH, emergency responders are required to take appropriate precautions to avoid contact with the spilled material; consequently, prevention of exposure to low residual VX, even if the organic fraction remains intact, should not require extraordinary measures. As with any release of hazardous liquid materials, untrained observers and the public should be kept away from the active
    response zone.4

    To be thorough, CDC sought to evaluate the likelihood and potential impact of a shipment of off-specification CVXH that could contain residual VX above the clearance level (≥20 ppb VX). At CDC’s request, the Army’s contractor evaluated the probability of human or system error resulting in shipping of off-specification CVXH.5

    The review of off-specification scenarios identified a potential cross-contamination link (a three-way valve that controls flow of both hydrolysate and agent) that could result in agent VX reaching the CVXH holding tank after batch reactor sampling. This potential link, without mitigation, reportedly would result in a calculated annual event frequency of shipping off-specification CVXH of approximately 1 per 20,000. Processing estimates for NECDF range from a low of less than 200 shipments per year up to a maximum of about 900 shipments per year if the entire stockpile is processed in one year. For cross-contamination to risk health or safety of transportation personnel would require coincidence of the event with a shipping accident large enough to release the VX hydrolysate and to splash the drivers or other people who might be in the area of the accident. The DuPont transportation review estimates the maximum likelihood of an accident involving a release of CVXH at 1 in 13,000. This estimate is based on actual observed transportation accident statistics in the United States. Combining the probabilities of two independent events—an off-specification shipment of CVXH involved in an accident severe enough to release its contents—yields an event likelihood of well under 1 in 1,000,000, which risk management specialists consider insignificant. Add to this the probability of a responder or other person being splashed during the event, and the total risk would be further reduced. Nonetheless, Dupont should consider deferring CVXH shipment during severe weather, such as heavy prolonged rains, icing, and snowstorms, to reduce accident risk.

    CDC believes the potential agent-related risk to human health and safety from a transportation accident involving off-specification CVXH is negligible. Nonetheless, the Material Safety Data Sheet (MSDS) for CVXH should recommend as a precaution that medical response personnel evaluate anyone having direct skin contact with released CVXH for possible nerve agent effects so appropriate medical intervention can be taken if needed. However, nerve agent effects are extremely unlikely, and the corrosiveness of caustic VX hydrolysate is likely to be the major concern.

    Finally, the highly odorous nature of normal-process CVXH should be noted. Although the cause of the odors would not be expected to result in adverse health impacts directly, knowledge that the spilled material originated from a facility processing agent VX could result in considerable confusion and possible panic during the event. This characteristic of CVXH should be described clearly to avoid potential misunderstandings. The MSDS for CVXH should alert responders to its disagreeable odor characteristics to help inform both responders and the public and to minimize possible confusion or concern over exposure to airborne VX.

    CONCLUSIONS
    This transportation analysis was based on information about CVXH produced with VX at the 8% loading level and stabilized with DIC. The remainder of the stockpiled VX, which is stabilized with DCC or with a mixture of DIC and DCC, is not addressed in this review because of inadequate characterization of the organic layer.

    The DuPont plan appropriately addresses CDC’s key risk management considerations, as well as DOT’s requirements for transporting hazardous materials. The predominant potential hazard during transportation of CVXH is its corrosivity. Precautions used to manage this hazard in a spill are adequate to protect response personnel from the low-level residual agent VX or residual EA 2192 at levels estimated for maximum credible event analysis.

    MATERIALS REVIEWED

    1. E-mail correspondence from Reeves (DOT) to Decker (CDC), May 19, 2004, re: Transportation Plan for Chemical Weapons Waste

    2. Presentation to CDC by Parsons and U.S. Army Chemical Materials Agency, May 24, 2004, re: Response to CDC questions regarding proposed operations at the Newport Chemical Agent Disposal Facility.

    3. Manthei J, Way R, Gaviola B, Burnett D, Bona D, Durst H, Thompson S. Toxicological Evaluation of VX Decontamination Wastestreams According to DOT Test Procedures, February 1999.

    4. Manthei J, Way R, Gaviola B, Bona D, Burnett D. “Alternative Technology Program: Intravenous Toxicological Evaluation of Four VX Wastestreams in Mice.” U.S. Army ERDEC, ECBC-TR-173, August 2001.

    5. “Quantitative Subsystem Hazard Analysis of Potential for Off Site Transfer of Hydrolysate Containing Above the 20 ppb Method Detection Limit”, Mary Kay O’Connor Process Safety Center, Texas A&M University System (TAMUS), August 2004.

    6. DuPont Technical Assessment on U.S. Army Newport (Indiana) Project,
      Executive Summary, E.I. du Pont de Nemours and Company, March 2004.

    7. Burke C. Transportation Safety Assessment and Risk Management Plan – Shipments of Newport (Indiana) Caustic Hydrolysate (NCH) Newport IN to Deepwater NJ, DuPont Safety, Health and Environment Excellence Center, March 2004.

    8. Zimmerman G, Ensminger J, Saulsbury J. Transportation Analysis for the Off-Site Shipment of Liquid Process Effluent from the Newport Chemical Agent Disposal Facility at the Newport Chemical Depot, Indiana, Oak Ridge National Laboratory for the U.S. Army Chemical Materials Agency, December 2003.

    1 E-mail correspondence from Reeves (DOT) to Decker (CDC), May 19, 2004, re: Transportation Plan for Chemical Weapons Waste
    2 Presentation to CDC by Parsons and U.S. Army Chemical Materials Agency, May 24, 2004, re: Response to CDC questions regarding proposed operations at the Newport Chemical Agent Disposal Facility.
    3 Manthei JH, Way RA, Gaviola BI, et al. Toxicological Evaluation of VX Decontamination Wastestreams According to Department of Transportation (DOT) Test Procedures, U.S. Army ERDEC, 1999 February.
    4 The risk concerns of residual VX discussed herein also would apply to the low level residual EA 2192 that could reside in the hydrolysate.
    5 “Quantitative Subsystem Hazard Analysis of Potential for Off Site Transfer of Hydrolysate Containing Above the 20 ppb Method Detection Limit”, Mary Kay O’Connor Process Safety Center, Texas A&M
    University System (TAMUS), August 2004.

    Previous | Table of Contents | Next