2
Changes at the Tooele Chemical Agent Disposal Facility in Response to National Research Council Recommendations

Review of NRC Recommendations Regarding Operational Verification Testing at JACADS

In 1994, the National Research Council (NRC) Committee on Review and Evaluation of the Army Chemical Stockpile Disposal Program (Stockpile Committee) published its report Evaluation of the Johnston Atoll Chemical Agent Disposal System Operational Verification Testing: Part II (NRC, 1994a). This report contained a general recommendation to proceed with systemization of the Tooele Chemical Agent Disposal Facility (TOCDF) and, during systemization, to conduct needed tests and make improvements (this recommendation is coded OVT2-2 in the present report; see appendix C). The specific tests and improvements listed in this recommendation are reviewed in the following sections of this chapter.

Brine Reduction Area

Satisfactory operation of the brine reduction area was not demonstrated during the initial JACADS Operational Verification Testing (OVT). As a result, the Stockpile Committee recommended that this activity be addressed satisfactorily prior to the start-up of the TOCDF:

Complete the brine reduction area (to include its pollution abatement system) performance tests, or develop a satisfactory brine disposal alternative. (OVT2-2B1)

The Army subsequently made a number of equipment modifications and completed the JACADS test. The Army report (U.S. Army, 1995a) shows that the system could be operated under steady-state conditions with only a small amount of salt accumulation in the ducts and filter bags and with stack emissions of particulates, Resource Conservation and Recovery Act (RCRA) regulated heavy metals, and hydrogen chloride all within Environmental Protection Agency (EPA) allowable limits.

Flow measurements and a total mass inventory were taken for a 30-hour run, and a mass and energy balance was established for the 8-hour steady-state run. Overall, the test demonstrated 100.1 ± 5 percent mass balance based on the total solids analysis for the 8-hour test. The mass accountability for the total 30-hour operation, including system start-up, shutdown for an overnight test break, restart, and shutdown, was 98.7 ± 6 percent, based on total solids analysis. The individual mass balances for the evaporator and dryer during the 8-hour test period were 103.9 and 103.7 percent, respectively.

The mass balance for RCRA-regulated heavy metals showed 115 percent recovery of chromium and 192 percent recovery of lead. The apparently higher recovery of lead probably reflects difficulties in accurately measuring the low concentration of lead in the feed brine because there was no other source of lead in the system. The percentage of chromium recovered was within reasonable limits of uncertainty for trace metals. Chromium stack emissions averaged 69.5 percent of the EPA limit, and the maximum among three samples was 142 percent. Arsenic and cadmium concentrations in the brine and salts were below the uncertainty limit of measurement.

Average particulate matter stack emissions were less than 1.7 percent of the EPA limit of 30 mg/dscm

1  

Appendix C includes extract listings of the recommendations from the 1993 and 1994 reports prepared by the Committee on Review and Evaluation of the Army Chemical Stockpile Disposal Program (Stockpile Committee). The findings from the 1994 Recommendations report are also included. To assist the reader, an alpha-numeric code reference has been added to each finding and recommendation shown. These code references are applied to each use in the text.



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 11
--> 2 Changes at the Tooele Chemical Agent Disposal Facility in Response to National Research Council Recommendations Review of NRC Recommendations Regarding Operational Verification Testing at JACADS In 1994, the National Research Council (NRC) Committee on Review and Evaluation of the Army Chemical Stockpile Disposal Program (Stockpile Committee) published its report Evaluation of the Johnston Atoll Chemical Agent Disposal System Operational Verification Testing: Part II (NRC, 1994a). This report contained a general recommendation to proceed with systemization of the Tooele Chemical Agent Disposal Facility (TOCDF) and, during systemization, to conduct needed tests and make improvements (this recommendation is coded OVT2-2 in the present report; see appendix C). The specific tests and improvements listed in this recommendation are reviewed in the following sections of this chapter. Brine Reduction Area Satisfactory operation of the brine reduction area was not demonstrated during the initial JACADS Operational Verification Testing (OVT). As a result, the Stockpile Committee recommended that this activity be addressed satisfactorily prior to the start-up of the TOCDF: Complete the brine reduction area (to include its pollution abatement system) performance tests, or develop a satisfactory brine disposal alternative. (OVT2-2B1) The Army subsequently made a number of equipment modifications and completed the JACADS test. The Army report (U.S. Army, 1995a) shows that the system could be operated under steady-state conditions with only a small amount of salt accumulation in the ducts and filter bags and with stack emissions of particulates, Resource Conservation and Recovery Act (RCRA) regulated heavy metals, and hydrogen chloride all within Environmental Protection Agency (EPA) allowable limits. Flow measurements and a total mass inventory were taken for a 30-hour run, and a mass and energy balance was established for the 8-hour steady-state run. Overall, the test demonstrated 100.1 ± 5 percent mass balance based on the total solids analysis for the 8-hour test. The mass accountability for the total 30-hour operation, including system start-up, shutdown for an overnight test break, restart, and shutdown, was 98.7 ± 6 percent, based on total solids analysis. The individual mass balances for the evaporator and dryer during the 8-hour test period were 103.9 and 103.7 percent, respectively. The mass balance for RCRA-regulated heavy metals showed 115 percent recovery of chromium and 192 percent recovery of lead. The apparently higher recovery of lead probably reflects difficulties in accurately measuring the low concentration of lead in the feed brine because there was no other source of lead in the system. The percentage of chromium recovered was within reasonable limits of uncertainty for trace metals. Chromium stack emissions averaged 69.5 percent of the EPA limit, and the maximum among three samples was 142 percent. Arsenic and cadmium concentrations in the brine and salts were below the uncertainty limit of measurement. Average particulate matter stack emissions were less than 1.7 percent of the EPA limit of 30 mg/dscm 1   Appendix C includes extract listings of the recommendations from the 1993 and 1994 reports prepared by the Committee on Review and Evaluation of the Army Chemical Stockpile Disposal Program (Stockpile Committee). The findings from the 1994 Recommendations report are also included. To assist the reader, an alpha-numeric code reference has been added to each finding and recommendation shown. These code references are applied to each use in the text.

OCR for page 11
--> (corrected to 7 percent O2); stack emissions of hydrogen chloride were less than 1.8 percent of the EPA limit of 0.026 grams per second (0.206 lbs/hr). During the past year, the Army also made a number of changes and conducted some acceptance tests at the TOCDF. Satisfactory completion of all the tests is required by the RCRA permit. Changes in the TOCDF Brine Reduction Area Facility The changes made for the TOCDF brine reduction area facility are described in the Required Report for the Operational Verification Tests (U.S. Army, 1993). The principal changes are outlined below. Salt Particulate. JACADS experience showed that some salt particulate was escaping the dryers, causing overloading of the baghouses. The TOCDF has increased the processing rate by installing four pollution abatement system baghouses in the brine reduction area (versus two at JACADS) to handle the increased gas flow and particulate loading from the brine evaporators and dryers. Condensate. At JACADS, there was condensate formation in the off gas ducting even though this gas was heated with an in-line burner to prevent condensation. Nevertheless, condensation occurred when the evaporator exhaust (which is almost 100 percent water vapor) was mixed with the dryer exhaust. A larger burner was installed at the TOCDF, and the evaporator exhaust was rerouted to enter downstream of the burner to keep gas from condensing. Duct Work Leaks. Leaks, apparently caused by the accumulation of corrosive condensate, developed in JACADS duct work; the leaks were repaired by sealing and installing drain tubes. The TOCDF system was redesigned using a larger heater, a different configuration for introducing the evaporator exhaust, and sloping ducts. Although no condensation is expected, the design ensures that any condensate will drain back to the evaporator or to a knockout box. Holes in Bag Filters. The bag filters at JACADS developed holes and had to be replaced. If the problem was caused by poor quality control of the bag fabrication or improper installation, the problem might recur at the TOCDF. For now, the TOCDF bags have been redesigned to be shorter (longer bags require a stronger pulse of air for cleaning than shorter bags). Overfilling the Tanks. The high-level alarms in the brine storage tanks at JACADS malfunctioned, indicating that the tanks were full when they were not. The false alarm required the incinerators to be shut down to prevent the tanks from overflowing. The JACADS instrumentation was subsequently repaired, and improved instrumentation has been installed at the TOCDF. In addition, there are four 40,000-gallon storage tanks at the TOCDF (versus the two 26,000-gallon tanks at JACADS), which will allow greater flexibility before a tank-full signal would require incinerator shutdown. Poor Drum Dryer Availability. The JACADS drum dryers proved to be unreliable, owing primarily to failures of the conveyor bearing and an inefficient conveyor wiper. The TOCDF design incorporates corrosion resistant conveyor bearings and a stainless steel wiper with a polyvinyl blade. It also includes inspection ports so that operators can check the wiper for salt buildup. Acceptance Test Results Certain portions of the brine reduction system were tested as part of the TOCDF systemization. Individual components (boilers, the pollution abatement system, surge tanks, evaporators, and drum dryers) were tested, and a systems material balance test was performed. Final stack emission tests have not been conducted because only simulated brine feed was available for testing. The results of testing should be similar to the successful JACADS runs (reported above) with modifications for the difference in processing rates. The results of the Tooele material balance performance tests are reported in Tooele Chemical Agent Disposal Facility, Phase 3 Systemization Demonstration Report, Brine Reduction Area Lines 1&2 (EG&G, 1994a). This report shows a discrepancy between the amount of salt collected and the theoretical amount of salt produced in the feed stream, i.e., the amount of dry salt collected represented only 80 percent of the predicted amount. However, a post-run inspection of the baghouses and condensate knockout boxes showed that few salts were present. This suggests that the discrepancy was primarily due to uncertainties in the salt content of the brine feed stream and the deposition of

OCR for page 11
--> salt in the duct work. Operational procedures require monthly inspections and cleaning the duct work. As a result of the above changes, the committee believes the test results are consistent with the successful demonstration of the brine reduction area at JACADS. Dunnage Furnace Satisfactory operation of the dunnage furnace and its related pollution abatement system was not demonstrated during the JACADS OVT. Therefore, the Stockpile Committee recommended that satisfactory operation be demonstrated prior to the start of agent operations at the TOCDF: Demonstrate the dunnage furnace performance with various levels of chlorinated waste; if needed, either modify the pollution abatement system design (e.g., add acid gas scrubbing) or limit feed materials to those that can be handled by the existing design; alternatively, satisfactory land disposal options must be identified. (0VT2-2C) The Army subsequently made a number of equipment modifications at JACADS and conducted the required RCRA trial burn tests during December 5–8, 1994. The following results of these tests were reported by the Army (U.S. Army, 1995b): The test feed stream consisted of cardboard boxes containing 260 pounds of wood dunnage along with polyethylene carboys containing 10.3 pounds of GB nerve agent absorbed in cellulose. No agent was detected (detection limit is 0.00006 mg/m3) in the stack at any time during the four test runs. The particulate emission level averaged 6.7 mg/dscm, with a maximum concentration of 16.0 mg/dscm. This maximum concentration is less than 9 percent of the allowable rate of 180 mg/dscm and occurred during baghouse air flow pulsing (for bag cleaning) during which time the effectiveness of the ''filter cake'' on the bags was reduced from normal. The hydrogen chloride emission rate varied from 0.59 to 0.72 lb/hr, i.e., only 15–18 percent of the allowable 4 lb/hr limit. The unit was operated with a rolling average carbon monoxide concentration of 46.1 ppm in the exhaust gas downstream of the afterburner, well below the allowable 100 ppm level (one-hour rolling, average on a dry basis corrected to 7 percent O2), and with five-minute average concentration peaks of only 62.2 ppm, well below the 200-ppm five-minute peak limit. Separate composite leachate samples were made from furnace ash and baghouse ash. Toxic Characteristic Leaching Procedure (TCLP) analyses were performed for the eight TCLP metals as well as for additional metals. These analyses indicated that sufficient quantities of seven of the eight TCLP metals were present to warrant treatment of baghouse ash as a hazardous waste. Analysis of the furnace ash demonstrated sufficient quantities of four of the eight TCLP metals to require treatment as a hazardous waste. Dioxins and furans were both present in baghouse and furnace ash samples at the nanogram per kilogram level. There are no regulatory limits on dioxins/furans concentrations in ash. Bis (2-ethylhexyl) phthalate was the only semivolatile compound found in the ash samples; it was detected (1,640 ng/kg) in one sample of furnace ash. No semivolatile compounds were found in the baghouse ash. No volatile compounds were detected in either ash sample.2 The results of the following measurements were in compliance with EPA standards: Exhaust-stack emissions including: O2 (oxygen); CO2 (carbon dioxide); volatile and semivolatile products of incomplete combustion; polychlorinated dibenzo-p-dioxins and dibenzofurans (PCDD/F); trace metals; particulate breakthrough; total hydrocarbons; NOx (nitrogen oxides); SOx (sulfur oxides); HF (hydrogen fluoride); Furnace scrubber liquid process samples for: agent GB; volatile and semivolatile products of incomplete combustion; polychlorinated dibenzo-p-dioxins and dibenzofurans; pH (acidity or alkalinity); trace metals; and reactivity. 2   Waste materials will be tested routinely. Toxic liquids can be disposed of at a hazardous waste facility through deep well injection and toxic solids in landfills.

OCR for page 11
--> Modifications to the Dunnage Furnace The principal changes to the TOCDF dunnage furnace pollution abatement system (U.S. Army, 1993), are described below. Temperature Control. The temperature of the JACADS dunnage furnace was to be controlled by limiting inlet air and causing oxygen-deficient combustion. However, air leaking in through the furnace-door seals allowed the dunnage load to burn at an uncontrolled rate. The JACADS unit was modified to operate with an excess of air by controlling the fuel content of the feed batch rather than the air intake. The TOCDF unit was redesigned to operate in a similar way. Induced Draft Fan Failure. Operating difficulties were experienced with the JACADS dunnage furnace after the induced draft fan failed with dunnage fuel still in the primary combustion chamber; the failure of the fan also locked out the furnace burner fuel supply. At both JACADS and the TODCF, backup emergency induced draft fans have been added, as well as cooling water spray nozzles located in the top of the primary combustion chamber. The spray cools the dunnage fuel, generating steam that helps purge the chamber. The induced draft fan and the water spray system are designed to keep negative pressure in the system and to keep the combustible gas concentration below the lower explosive limit during cooling of the dunnage fuel. Lift Shaft Fire. On one occasion the JACADS dunnage furnace ram was unable to push into the furnace a box of dunnage that had caught on the edge of the furnace door. Radiant heat from the furnace ignited the box in the entry shaft. The TOCDF dunnage furnace feed system has been redesigned with a higher ram pushing force and a shorter cycle time. In addition, a new heat shield reduces preheating of the box before the furnace door is completely open. Lift Shaft Hydraulic Fluid. Furnace heat caused the ethylene glycol/water hydraulic fluid used at JACADS to thicken gradually. This eventually caused the lift to bind. Both JACADS and the TOCDF will now use FYRQUEL 220, a high-temperature hydraulic oil. Also, the lift will be cycled several times a day during furnace operations to keep the lift seals and shaft lubricated properly. Ram Failure. The hydraulic ram on the JACADS dunnage furnace failed to retract on one occasion, damaging bearings and instrumentation. The TOCDF dunnage furnace ram system has been completely redesigned. The original JACADS hydraulic unit has been replaced with an electrical powered unit, and a heat shield has been added to protect the ram during door cycling. Fuel Oil Strainers. JACADS fuel oil strainers frequently became plugged up, requiring unscheduled maintenance. The furnaces at the TOCDF will burn natural gas rather than fuel oil, thus eliminating this problem. Baghouse Pressure Drop. The JACADS RCRA permit requires that the baghouses operate with a pressure drop of 1–20 inches of water column. Higher pressure might cause bag plugging, and lower pressure might indicate that the bags are torn or leaking. There were problems at JACADS maintaining a minimum pressure drop because the combustion rate and exhaust gas flow rate were lower than expected. However, periodic visual inspection revealed no problems with the bags, and the permit was modified to allow for a lower pressure drop at reduced operating rates. The TOCDF flow rates and bag sizes will differ from those at JACADS. However, operational control over the dunnage feed rates will continue to be required in order to maintain a proper pressure drop. Pressure Excursion. A maintenance operator inadvertently shut off the emergency quench water spray at the JACADS dunnage furnace during an incineration operation, also locking out the furnace fuel feed burners. Continued burning of the dunnage load in the furnace (operating in an oxygen-deficient atmosphere) then produced combustible gases that migrated into the lift closure where they ignited, causing overpressurization that damaged the shaft and surrounding wall panels. Both JACADS and the TOCDF have since added a water (steam) purge to the primary combustion chamber to cool dunnage fuel and to keep the combustible gas concentration below the explosive limit in case the burners lock out while the dunnage load is still burning. Dunnage Furnace Nonagent Test Results Two major nonagent tests were conducted at the TOCDF (EG&G, 1994b, 1994c). These tests were not intended to be performance acceptance tests, which will be conducted after agent processing begins.

OCR for page 11
--> A thermal capacity test of the modified TOCDF furnace was conducted to demonstrate the following capabilities: (1) the capacity to handle dunnage disposal at the required rate; (2) operability of the furnace and related pollution abatement system in automatic mode; and (3) the operability of various components, including the dunnage furnace pollution abatement system. Over a continuous eight-hour period the tests demonstrated a thermal capacity to dispose of 431 lb/hr of mixed dunnage as compared to the original design capacity of either 1,000 lb/hr of wood dunnage or up to 24 mine drums (560 pounds of metal) per hour. The Army maintains that the dunnage furnace is an optional part of the overall disposal operation and is primarily used to minimize waste disposal costs. There is also a program to minimize the feed stream to the dunnage furnace. Nevertheless, improving the operating efficiency of the TOCDF dunnage furnace is important because it is currently designated as the principal system for reducing facility hazardous waste. A water spray purge test was performed to determine if the emergency fan, combined with water spray purge cooling in the event of a primary induced draft fan power failure could maintain a negative pressure in the primary reaction chamber. Each of the three test runs was successful, indicating that water spray cool-down could be used, along with the emergency induced draft fan in the event of a failure of the main induced draft fan and a lockout of furnace burner fuel supply. During tests with 200-pound and 500-pound dunnage loads, furnace pressure remained negative, no combustible gas concentrations were detected, and carbon monoxide concentrations remained below RCRA limits (<100 ppm) at all times (EG&G, 1994c). In the final analysis, the level of exhaust gas emissions from the modified dunnage furnace at the TOCDF is likely to be similar to or lower than the level of gas emissions at JACADS. Overall Assessment of Dunnage Furnace After reviewing the changes and tests described above, the committee believes that the dunnage furnace will perform acceptably at the TOCDF. Nitrogen Oxide Emissions During JACADS OVT operations, nitrogen oxide (NOx) emissions during VX destruction were acceptable, but the Stockpile Committee recommended that NOx emissions be analyzed with respect to more stringent requirements that might apply to U.S. Sites: Review the probable levels of NOx production from VX destruction and the allowable emission levels at the other continental U.S. sites requiring VX destruction; if appropriate, develop needed NOx abatement systems. (OVT2-2D) There are currently two general regulatory standards to be met for nitrogen oxides: (1) the prevention of significant deterioration of the air quality which sets a limit of 250 tons per year of NOx (based on a 12-month rolling average) at any single facility; and (2) the requirement that the NOx concentration not exceed 1 µg/m3 at the facility boundary, again based on a 12-month average. If calculations show that either of these limits may be exceeded, then the facility must demonstrate by modeling that the excess NOx will not damage the local air quality. If the modeling demonstrates a deterioration of air quality, NOx controls must be added to the air pollution abatement system. The Ralph M. Parsons Company (Parsons, 1994) has estimated NOx emission levels for the TOCDF using an equilibrium model to calculate mass and energy balances on the various furnace systems comprising the baseline system. Because NOx formation is kinetically controlled, the equilibrium model may predict much higher thermal NOx emissions than will actually be produced. (Note: This was confirmed during testing of the TOCDF.) The thermal capacity tests at the TOCDF during systemization provided NOx test data that can be used to develop an empirical relationship (equation) between NOx emissions and furnace temperature for the liquid incinerator, the deactivation furnace system, and the metal parts furnace. This equation makes possible the prediction of NOx emissions, as a function of temperature only, for furnaces from any system. The resulting equation was used to adjust NOx concentrations predicted by the Parsons equilibrium model and energy balances. Because no munitions are being burned in the tests, and the fuel is natural gas, the only nitrogen source is the combustion air. When nitrogen-containing constituents are combusted in baseline system furnaces, the NOx levels will be higher. Currently, there are no data for "fuel NOx" because the TOCDF has not started processing munitions. The worst case scenarios for annual NOx emissions were calculated for each of the eight continental U.S.

OCR for page 11
--> sites (see appendix B for stockpile locations). First, the total contribution of NOx emissions was calculated for each incinerator, burner, and boiler using only fossil fuel. Second, the total contribution of agent (fuel) NOx emissions was calculated using design processing rates for each incinerator and a worst case processing schedule. (Note: Calculations are based on operations of 6,000 hr/yr and furnace idle time of 2,760 hr/yr.) The principal source of agent fuel NOx is in the VX feed-stock. The Stockpile Committee believes the incineration of VX should not significantly increase the Parsons estimates for the following reasons: There are relatively small quantities of VX in the stockpile. Annual NOx emissions are regulated by 12-month rolling averages; short VX campaigns will be averaged with non-VX campaigns during any 12-month period, thus reducing the impact of processing VX. The estimates for generating NOx from VX incineration, although not as conservative as the estimates for non-nitrogen-bearing munitions, still assume a significant conversion efficiency of agent fuel nitrogen to NOx. The Parsons mass and energy balances indicate that none of the sites will exceed the limit of 250 tons of NOx emissions per year. Calculation of the point dispersion impact (perimeter concentration levels) are site-specific and contingent on local meteorological conditions as well as the distance to the facility boundary. Point dispersion calculations have not been completed for all sites owing to the unavailability of site-specific meteorological data, but the Stockpile Committee believes that NOx emissions will not cause a major impact at any location because of the relatively small amount of NOx expected to be generated. The main stack at each facility will include an NOx meter and a total exhaust gas flow meter. The actual NOx emissions both during furnace idle (not processing) and during agent operations (including trial burns) will be recorded and tracked against the estimates. Liquid Incinerator Slag Removal The combustion of agent in the liquid incinerator converts most of the carbon and hydrogen to volatile gases. However, inorganic elements are changed to acids, oxides, or salts, which condense on the walls of the slightly cooler secondary combustion chamber to form molten slag that slowly flows down the walls and collects in a pool at the bottom of the chamber. (The TOCDF slag production rate will be on the order of 100 pounds per hour, depending on the agent being destroyed.) Because the accumulated slag must be removed periodically, the JACADS liquid incinerator was designed with this in mind, using principles that were then being developed and tested at the Chemical Agent Munitions Disposal System (CAMDS) facility. These involved the use of a sloped furnace chamber bottom with a central tap hole from which the liquid slag could be drained. However, subsequent CAMDS tests (after construction of JACADS was begun but prior to JACADS start-up) showed that the system would not operate effectively because the slag solidified. Consequently, the liquid drain system was never used or tested at JACADS. Periodic shutdown of the system to allow operators to enter the secondary chamber and remove the hardened slag manually has contributed significantly to downtime of the JACADS liquid incinerator. Based on the JACADS experience, shutdown and manual removal procedures would limit availability of the TOCDF liquid incinerator to only 30 percent of peak design throughput. The Stockpile Committee recognized this as a significant problem: Develop and demonstrate the proposed hot-slag removal system for the liquid incinerator system. (OVT2-2E) To minimize this problem, a program was initiated to develop a new, heated, liquid slag removal system for installation at the TOCDF and other plants in the continental U.S. The complete system is referred to as the slag management system. This development program initially involved a review of existing commercial methods and equipment for removing slag in related industrial applications. A specific design was then developed from these various alternatives. Because this design was not completed in time for demonstration at JACADS or implementation at the TOCDF prior to the start of systemization tests, the committee recommended that the demonstration be completed prior to the start of agent operations at the TOCDF. Once it became clear that the JACADS design would not work, the TOCDF furnaces were redesigned using a flat bottom instead of a sloped bottom. The bottom was

OCR for page 11
--> elevated so that subsequently designed equipment could be installed under it. The new slag management system design involves cutting a central hole in the existing flat furnace bottom, attaching a cylindrical extension of the furnace under the central hole, and providing a side tap system from this extension to drain molten slag at intervals during operations. The side tap system involves a vertical slide valve to stop slag flow, a remotely operated external drill to drill through the frozen slag into the inner molten zone, and electrical resistance heaters to keep the drilled channel molten during periodic drainage operations. The heaters are turned off after the drainage operations are completed, allowing the outer layer of slag to solidify because the solid is less corrosive. At peak operating rates, slag drainage will be necessary about once a week. Slag draining operations will be conducted with the furnace hot and operating, but agent feed will be stopped to avoid any possibility of agent escape with the slag. The newly installed slag management system at the TOCDF will be mechanically tested prior to the start of agent destruction operations. However, because of the variability of the slag, the Army decided that there was no adequate surrogate for complete testing of the system, which will be tested only when slag accumulates from agent destruction operations. If the slag removal system does not allow for hot slag removal, the current manual removal procedure will be resumed. The committee will complete assessment of the slag removal system after testing with agent at the TOCDF. Furnace Feed System The Stockpile Committee recommended that better methods for tracking various types of munitions be implemented to avoid furnace feed errors of the kind encountered during OVT testing at JACADS: Eliminate furnace feed errors by improved monitoring and control of the deactivation furnace and metal parts furnace feed systems and by improved methods for tracking the various types of munitions. (OVT2-2F) The Army responded to this recommendation by retaining the MITRE Corporation to assess munition tracking problems identified at JACADS and to provide appropriate recommendations for resolving munitions tracking problems that have been identified. The MITRE Corporation report (MITRE, 1994) presents a number of recommendations for modifications at the TOCDF. In addition to the problems observed from JACADS, MITRE identified 13 other process limitations that might compromise the effectiveness of tracking munitions. MITRE concluded that the enhancements implemented for the TOCDF rocket processing line and the projectile processing line should be adequate to minimize the likelihood of repeating munitions tracking failures during future chemical agent disposal facility operations. MITRE's analysis concentrated on each processing campaign (rockets, projectiles, bulk items) separately and on the critical steps in each campaign. JACADS OVT malfunctions that interfered with the successful completion of the critical step were analyzed, and recommendations for correcting the deficiency were proposed. Additional improvements were also recommended to the Army for consideration. The following changes for the rocket processing line have been implemented at the TOCDF: To ensure the full drainage of rockets, an interlock has been added that requires a signal from sensors in the agent quantification tank to confirm that rockets have been properly drained before the rockets can be moved out of the drain station. An interlock constraint on feed rate of rockets has been added to the deactivation furnace to prevent loading in excess of that allowed in the RCRA permit. Changes for the projectile processing line include: Fail-safe proximity sensors ensure that the trays of projectiles remain on the tray conveyor system and that trays cannot proceed without this verification. The pick-and-place loader primary control has been moved from local control to the central control system. The local control instrument did not have the capability of monitoring manual operations. The control room monitors both instruments and manual operations. An interlock constraint on the feed rate of projectiles to the metal parts furnace has been added to

OCR for page 11
--> prevent loading in excess of that allowed in the RCRA permit. Changes for the bulk item processing line include: A bubbler device has been added to indicate the level of agent remaining in a munition or bulk container. An encoder on the bulk drain tube may be installed to determine the level of liquid left in the container. With improvements in operations of the load cells, this might be enough to ensure the proper drainage of containers. The Army is considering this improvement, and tests are being run. In addition, the Army is developing permit terms with the Utah Division of Solid and Hazardous Waste (DSHW) to deal with containers that cannot be fully drained because of gelled agent. An interlock constraint on the feed rate of bulk items has been added to the metal parts furnace to prevent loading in excess of that allowed in the RCRA permit. Changes to prevent problems from the manual override of interlocks include: Adding sensor fault trapping, which indicates to the control system when changes in critical sensors or interlocks occur. The implemented improvements have been satisfactorily tested during integrated plant runs as part of the TOCDF systemization process (MITRE, 1995). The changes made and tested at the TODCF address the problems identified during the JACADS OVT and should provide for substantially safer furnace feed performance than at JACADS. Residual Gelled Agent Determining the residual agent level and detecting residual gelled agent in ton containers and spray tanks (bulk containers) is essential to establishing a remote real-time indication of agent level while a container is being drained. The method of measuring the level of residual agent during JACADS operational verification testing required that personnel wearing demilitarization protective ensembles (DPEs) manually insert a dipstick into each container. An accurate measure of residual agent in bulk containers is necessary to ensure compliance with RCRA permit requirements. Also, the residual level is critical because it affects the processing rate of the metal parts furnace. Consequently, the Stockpile Committee recommended that this problem be addressed: Address all problems associated with residual gelled mustard, in particular, the use of suited personnel to perform functions that were intended to be automated. (OVT2-2G) Since completion of JACADS OVT, the agent collection and quantification system at the bulk drain station (BDS) for the TOCDF has been modified (GPS Technology, Inc., 1993). This modification was tested at the Chemical Demilitarization Training Facility, and the finalized design was implemented at the TOCDF. The TOCDF team is also updating operating procedures to incorporate revisions in the agent collection and quantification system into the process operating procedures. Modifications to the bulk drain station emulate the accepted design for the residual agent level sensing system of the multipurpose demilitarization machine. The modifications to the bulk drain station fulfill several other remote processing needs at JACADS. Measuring the residual agent level inside a bulk container is facilitated by a bubbler orifice at the lower end of a new suction tube inserted into the container. This presents a remote real-time indication of agent level while the suction tube is still in the bulk container at the bulk drain station, thus providing the operator with a discrete go/no-go reading. The sensor and connected instrumentation are remotely and automatically checked at the beginning and the end of each operation to ensure that the system is operating properly. This eliminates the need for manual dipstick measurements of agent remaining in a bulk container after drainage. If the bulk drain station is not in use, or an area ventilation system upset occurs, the system is automatically disabled and isolated. The systems for sensing the residual level of agents in bulk containers also alert operators to "pump out" problems, which are indicative of residual gelled agent. These systems should help to increase plant processing performance to the designed capacity, and the remote-automatic measurement procedure should substantially reduce the number of entries by personnel wearing demilitarization protective ensembles.

OCR for page 11
--> Environmental Permitting and Regulatory Requirements With respect to environmental permitting and regulatory requirements, the Stockpile Committee recommends that the Army: Establish and maintain close working relationships with permitting agencies, and support these efforts with careful analysis of operating parameters to ensure that permits provide for safe destruction of agent, adherence to regulatory requirements, and effective plant operations. (OVT2-3) The TOCDF staff has established a set of comprehensive activities to ensure timely and effective coordination with the Utah DSHW. This coordination has yielded positive results in identifying problems related to environmental permitting and compliance and in developing solutions to these problems. Some examples of these coordination efforts are: Technical responses. A quick response system has been established to respond to DSHW questions and requests for information. Typically, the response is provided by a member of the TOCDF team making a personal appearance at the DSHW offices in Salt Lake City. For example, during the review of the liquid incinerator #1 surrogate trial burn plan, daily meetings were held each afternoon to respond to comments and questions resulting from the DSHW review. Because of this level of coordination, the liquid incinerator #1 surrogate trial burn plan was completed on schedule. This approach is being used for other document reviews as necessary. Planning meetings. Planning and scheduling meetings attended by the TOCDF team and the DSHW were held on a monthly basis until March 1995. As the schedule became tighter, meetings were held approximately once a week. These planning and scheduling meetings are used to prioritize the efforts of both the TOCDF team and the DSHW personnel. TOCDF presence at the DSHW offices. In addition to face-to-face coordination meetings, members of the TOCDF team are present at the DSHW offices at least two days a week to help develop DSHW work schedules coordinated with TOCDF schedules, and to ensure that the documentation in the DSHW information library is current. DSHW presence at the TOCDF site. The DSHW established an office at the TOCDF site, which is staffed at least two days each week, and at other times as necessary (e.g., for the completion of required inspections). Once agent operations start, it is anticipated that the DSHW will maintain a presence on site. Training of DSHW personnel in chemical demilitarization. Representatives from the DSHW attended a training course (a short version of the control room operator course) at the Chemical Demilitarization Training Facility. After the course, DSHW personnel had a better understanding of how the systems work and, more importantly, why they work the way they do. The comprehensive sets of activities outlined above should ensure the safe destruction of agent, adherence to regulatory requirements, and effective plant operations. Environmental Compliance The Stockpile Committee believes that environmental compliance and the Chemical Stockpile Disposal Program should go hand in hand. This belief is reflected in the committee's clear recommendation regarding compliance: Establish programs, procedures, and management oversight to ensure continuing compliance with all environmental regulations. (OVT2-4) EG&G Defense Materials, Inc., the Army's implementing contractor, has the overall responsibility for ensuring that the TOCDF is in compliance with all applicable federal, state, and local environmental requirements. To meet this responsibility, EG&G has drafted an Environmental Compliance Plan (ENVCP) (EG&G, 1995a), a summary document that provides guidelines for general environmental compliance, background for the environmental impact statement, and an overview of the environmental requirements to which the TOCDF will adhere. The ENVCP is a flexible document that is modified, subject to regulatory approval, to accommodate specific phases of the project, e.g., systemization, operations and decommissioning, and chancing environmental regulations affecting specific phases of the project. To ensure timely and complete implementation of the ENVCP, the following roles and responsibilities have been assigned:

OCR for page 11
--> The Environmental, Safety, and Surety Manager is responsible for assuring implementation of all environmental procedures by the Environmental Compliance Group, including the maintenance and updating of all environmental procedures; notifying regulatory agencies and management personnel of EG&G Defense Materials, Inc. of major reportable deficiencies and completing all paperwork required by company, state, and federal rules; informing EG&G management of compliance status and trends. The Environmental Compliance Coordinator ensures that both manpower and material assets are available to execute all environmental inspections; ensures that reports are drafted on time and that follow-up inspections are executed for all deficiencies; is responsible for administering the TOCDF Environmental Assessment Program, a periodic evaluation—via audits, surveillance, or inspections—of TOCDF compliance with federal, state, Department of Defense, and Tooele Army Depot environmental permits, rules, and regulations. Environmental Inspectors oversee and provide technical assistance to operators of the regulated areas; conduct environmental compliance inspections in accordance with written procedures. Quality Assurance/Quality Control Manager assists the environmental manager with overall guidance for the proper implementation of the Internal Environmental Audit Program, including auditing hazardous waste transporters; ensures that the appropriate personnel are properly trained, qualified, and certified in auditing techniques and procedures and that environmental audits of hazardous waste transporters are planned, executed, and documented. All TOCDF employees report noncompliance with environmental requirements to their supervisors, no matter how insignificant an incident may appear. Reports are forwarded to the environmental office for evaluation against reportable event criteria. Representatives from the office of the Program Manager for Chemical Demilitarization (PMCD) are then notified; required notifications to the Tooele Army Depot or the Utah Division of Solid and Hazardous Waste are coordinated by the PMCD; telephone notification is then made, followed by a written report. Operational procedures for environmental inspections have been established for the following areas: the munitions demilitarization building; the personnel and maintenance building; the container handling building (CHB); the monitor support building; the chemical assessment laboratory; the outbuildings at the TOCDF; the residue handling area; the brine reduction area; the pollution abatement systems; TOCDF satellite accumulation sites; the process and utility building; and the 90-day hazardous waste storage sites. In addition, environmental procedures have been developed for hazardous waste transport audits; environmental self-assessment; environmental inspections; and incorporating environmental compliance requirements into operating procedures, project regulatory procedures, and drawings. The ENVCP ensures compliance with all environmental laws, rules, permits, and regulations by identifying shortfalls in environmental/operational design and execution. The environmental compliance activity accomplishes this through daily surveillance inspections and periodic audits of both on-site regulated activities and off-site treatment, storage, and disposal facilities that receive waste generated at the TOCDF. Databases to track nonconforming conditions are maintained and periodically examined to identify significant problems. Procedures are reviewed, and changes or new regulatory requirements are incorporated and "tagged." Opportunities for hazardous waste minimization and pollution prevention are also investigated. Implementation of the ENVCP and strict adherence to the established environmental procedures listed above should ensure continuing compliance with all environmental regulations. Overall Safety One major design improvement at the TOCDF over JACADS was the better incorporation of human factors in the design of the physical plant. Many of the poorer design aspects of JACADS, such as inaccessible manual valves and meters, especially in areas requiring workers wearing demilitarization protective ensembles, have been eliminated at the TOCDF. These design problems had been implicated in upsets at JACADS, such as operators not monitoring an agent flow gauge that was difficult to reach. The Stockpile Committee reacted in part to this situation by making the following recommendation:

OCR for page 11
--> Develop systems to improve overall management of safety. (OVT2-5) At the TOCDF, the committee noted that most places requiring manual access were within reasonable human reach and sight. The pipe work was clearly labeled as to flow direction and contents; good use was made of color coding of temporary connections; and computer console displays and video surveillance systems in the control room were designed to be more user-friendly. The workers interviewed at the TOCDF commented favorably on the quality of these improvements. Changes Resulting from Risk Assessment At the following recommendation of the Stockpile Committee, the Army undertook a site-specific quantitative risk assessment (QRA) at the TOCDF: Complete the risk assessment for the Tooele Chemical Agent Disposal Facility during the systemization period. (OVT2-6) Although the assessment for the TOCDF has not yet been completed, routine reviews of information from the risk assessment for campaigns 1 and 2 (SAIC, 1995a) have led to changes in the facility and in its schedules to reduce risk. These are described further in chapter 6, "Overview of Site-Specific Risk Assessment." Specifically, the following four changes are currently being implemented: Operation of the metal parts furnace feed airlock. The QRA suggested that flammable vapor might accumulate in the metal parts furnace feed airlock, especially for bulk agent containers. Venting the airlock to the furnace afterburner is being evaluated, as are procedural changes to limit the residence time for items in the airlock. A solution will be found prior to the start of agent operations in the metal parts furnace. Weteye bomb aluminum and agent interaction. The QRA identified the potential for interaction between molten aluminum and liquid agent in the metal parts furnace during processing of aluminum weteye bombs. Interaction could cause an explosion within the furnace. As a result of this finding, the order of the campaigns was changed. Ton containers instead of weteye bombs will be co-processed with GB rockets during the first campaign. Further review of the Science Applications International Corporation (SAIC) calculations and development of processing strategies that will avoid molten aluminum and agent interaction (SAIC, 1995a) are under way. Weteye bomb handling and inventories. The QRA found that weteye bombs are significant contributors to risk because they contain GB and are relatively thin-walled. Additional analysis in the QRA may indicate the number of weteye bombs that should be stored in the container handling building to minimize risk. Seismic anchorage of the liquid propane gas tank. The facility review of equipment fragilities during an earthquake indicates that the liquid propane gas tank anchorage has some seismic vulnerability despite its construction to seismic zone three requirements. The information provided by the QRA team will be used as a basis for evaluating the need for additional bracing for this tank. Risk reduction from these changes is a direct result of the risk assessment. The committee anticipates that additional improvements will be identified and implemented as the TOCDF risk assessment is completed early in 1996. Review of NRC Recommendations Regarding the Monitoring System at JACADS After reviewing the monitoring systems for the destruction of chemical agents and agent by-products at the JACADS during systemization, the committee issued a report entitled Review of Monitoring Activities Within the Army Chemical Stockpile Disposal Program (NRC, 1994b). This report included five general recommendations and ten specific recommendations. The specific recommendations addressed six issues involving plant-wide agent monitoring, and exhaust stack agent and agent destruction by-product monitoring, and four issues affecting the operation of the analytical laboratories supporting both agent and nonagent monitoring activities. These recommendations were motivated by the finding that "the monitoring system currently in use at JACADS should be improved prior to employment at sites in the continental United States" (NRC, 1994b). In this section the recommendations presented in the Monitoring report

OCR for page 11
--> are repeated and the Army's responses to these recommendations are reviewed and evaluated. In response to the general recommendations, the Army has addressed monitoring issues for both agent and agent destruction by-products more attentively and actively than was evident before the Monitoring report (NRC, 1994b) was issued. In general, when specific issues could be addressed with improved deployment of, or incremental improvement to, standard agent monitoring techniques or utilization of other commercially available analytical instrumentation, the Army has done a very good job. However, when the implementation of committee recommendations required significant research and development to create or identify new technology, the Army's response has been tentative and less effective. This is partially attributable to the fact that the Chemical Stockpile Disposal Program is a major defense acquisition program attempting to slow cost growth attributable to previously unbudgeted requirements. In addition, the agent monitoring program did not have experience in procuring and managing research and development and has been slow to master these new skills. General Recommendations for Agent/Nonagent Monitoring In the Monitoring report, the Stockpile Committee made five general recommendations. The first was: The Army should initiate a substantial program to upgrade the monitoring systems for continental U.S. sites. (MON-1) In response, the Army tasked senior personnel within the Environmental and Monitoring Division of their Chemical Demilitarization and Remediation Activity (now Program Manager for Chemical Demilitarization) to initiate and sustain a substantial program to upgrade monitoring systems for the U.S. sites, commencing with the TOCDF. The second general recommendation was: The Army should obtain expert help at both the systems design and the equipment selection levels, perhaps by engaging a contractor with extensive experience in monitoring of trace species and in advanced instrument development. (MON-2) The Army has accordingly retained contractors to help evaluate and select upgraded monitoring equipment (SAIC, 1995b); however, the contractors involved are generally competent only to assist in identifying standard, commercially available analytical and monitoring equipment. They lack the background necessary for deciding whether some new monitoring technology, which is only available for research, should be improved or adapted to meet the needs of the disposal program. The third general recommendation was: The Army should undertake whatever instrument development is necessary to ensure that improved instrumentation is available to the chemical disposal program in suitably rugged and operational forms. (MON-3) In response to specific committee recommendations, the Army has let one instrument development/demonstration contract for a basic agent monitoring technology (Fourier Transform Infrared system) that is not currently available commercially. However, additional research and development, either by the Army's laboratory programs or by outside contractors, has not been initiated yet so that additional technologies that might provide longer term upgrades in monitoring capabilities can be tested and promoted. The fourth general recommendation was: The Army should test and use new monitoring instrumentation at JACADS before such instrumentation is employed at Tooele. (MON-4) The Army has initiated both new deployment modes of existing monitoring technology and deployment of new commercial analytical instrumentation at JACADS to evaluate their effectiveness for use at the TOCDF. The fifth general recommendation was: The Army should plan to continually improve the monitoring system in areas where performance is presently limited by unavailability of suitable instrumentation. (MON-5) The Army has initiated activities to redress most of the deficiencies that have been identified in monitoring capability. The Stockpile Committee believes that the Army is committed to the continuous improvement of monitoring systems at the TOCDF and other U.S. disposal sites.

OCR for page 11
--> Specific Recommendations for Agent/Nonagent Monitoring In the Monitoring report, the Stockpile Committee made ten specific recommendations, of which the first six addressed agent/nonagent monitoring. The first recommendation in this group was: Add the capability for positive identification of chemical agent species (chemical speciation) to the agent detection system and analytical laboratories at all of the disposal facilities in order to reduce the occurrence of false positives. (MON-6) The present analytical system uses gas chromatography, which is sensitive to many chemically similar compounds leading to false positive signals. A single false positive requires shutdown of agent operations but does not by itself initiate the response appropriate for a major agent release. False positive signals disrupt plant operations and increase the potential for human error and equipment degradation. Using mass spectrometric detection in the analytical laboratory now provides improved discrimination between chemical species, as would using the next generation of reliable, rugged infrared spectroscopy-based field instruments with adequate detection limits throughout the plant. In response to the above recommendation, the Army has deployed gas chromatographic systems with mass spectrometric detectors (GC/MSD) at selected sites at JACADS and the TOCDF to identify the specific chemical species responsible for activating the Automatic Continuous Air Monitoring System (ACAMS) agent alarms. The new instruments are deployed in the analytical laboratories at each site and supplement the gas chromatographic/flame photometric detector (GC/FPD) analysis of Depot Area Air Monitoring System (DAAMS) absorption tube samples. These samples are used to determine whether an ACAMS alarm was triggered by agent release or was a false positive alarm triggered by another chemical. Procedures have been devised and tested for GC/MSD to analyze an additional DAAMS sample with every ACAMS alarm. This procedure facilitates the identification of the specific chemicals that frequently cause false positive ACAMS alarms. Once they have been identified, these chemical can be eliminated from the disposal facilities, thus reducing the level of false positive alarms. In addition, the Army is planning a field test in the analytical laboratory at the TOCDF of a gas chromatographic system with both mass spectrometric detectors and atomic emission detectors (AED). This GC/MSD/AED system ought to provide more information about the identity of substances that trigger false positive alarms. The Army plans to test this instrument at the TOCDF in June 1996 with various potential trigger substances. A more species-specific alarm to complement the current ACAMS system needs to be developed for both general plant and stack monitoring. The second specific recommendation was: Institute continuous monitoring for all agents present at each facility, including those in storage areas. (MON-7) With single-agent monitoring systems, there is a risk that release of a different agent from mislabeled munitions or leaky storage containers might go undetected. Single agent monitoring systems also increase plant downtime during agent changeover operations. There are several ways to conduct multiple-agent monitoring, including the simple installation of additional ACAMS at each site. These modifications should be weighed against possible increases in the false alarm rate. The Army's deployment of ACAMS set at the time-weighted average (TWA) exposure level for each agent (GB, VX, HD) in the agent unpack area at each site is a positive step. ACAMS for each agent are also deployed on the ventilation air carbon filter units and on the common incinerator exhaust stack. The Army has announced plans to implement a similar deployment in the agent storage areas as recommended by the committee. The third specific recommendation was: Reduce the time required for confirmation of false positives. (MON-8) Confirming ACAMS alarms requires retrieval and laboratory analysis of DAAMS sample tubes. The manual transport of tubes and the analysis can lead to delays of 15 to 20 minutes. Analysis times can be reduced by using analytical methods with greater agent specificity. But, manual transport delays, a significant component of overall delays, will still be a problem. A significant reduction of the time required to confirm ACAMS alarms would require development and deployment of a rapid, species-specific alarm system

OCR for page 11
--> alongside each ACAMS. At present, the Army has not yet identified monitoring technology to fulfill this function. However, the Army has initiated a development program for an advanced electronics architecture, dual detector technology ACAMS, which should provide agent detection with a three to five minute response time as well as detection redundancy. The fourth specific recommendation was: Evaluate the procedures for periodic testing of field sensors to ensure that false negatives are not possible if a significant release should occur. (MON-9) Potential failure modes in sample collection could produce false negative responses. Monitoring system challenge tests are now conducted, but a statistical analysis of test results could be used to identify common failure modes. Once these are identified, changes in monitoring equipment or its maintenance could minimize the number of false negatives. In briefings to the committee, the Army has indicated high confidence that its procedures for regular field challenges of ACAMS and DAAMS field monitors using agent samples in calibrated solutions are adequate to eliminate a significant number of false negatives. The current field challenge procedures were developed in 1991 in collaboration with the Department of Health and Human Services (DHHS). The Utah Department of Environmental Quality (DEQ) has questioned the extent to which ACAMS agent challenges test the air sampling component and the GC/FPD detector portion of the ACAMS system. The current field monitor challenge procedures test several levels of calibrated agent solutions on a daily basis. These procedures appear to be effective in identifying failing ACAMS. The Army must continue to analyze the results of these tests to identify common failure modes and to institute maintenance procedures to minimize field failures. The fifth specific recommendation was: Implement monitoring designed to provide more rapid response to high-level agent release. (MON-10) The Army is planning to deploy time-phased (staggered) ACAMS on the common incinerator exhaust stack at both JACADS and the TOCDF. This is a good first step toward an alarm system that will cut the alarm time in half in response to a major release. The response time of the standard ACAMS instrumentation lengthens when the ACAMS is operated at greater sensitivity. An ACAMS set for low detection levels with longer sampling times might lead to unnecessary delays in response to high levels of agent release. Supplementing low-detection-level ACAMS with high-detection-level ACAMS may be an acceptable way of improving detector response times. This recommendation and the Army's response are closely coupled to recommendations MON-8 and MON-9. To date, the Army has committed to deploy ACAMS in the unpack area at both JACADS and the TOCDF set at the high immediately dangerous to life and health (IDLH) level. These ACAMS will respond more rapidly than ACAMS set at normal levels. In addition, time-phased (staggered) ACAMS are being deployed in the common stack for faster ''stop feed'' control of furnaces. These systems will also be in place at the TOCDF before the start of agent operations. Both the Stockpile Committee and the Army recognize that new monitoring technology will be required to reduce the desired response times to a few seconds, rather than the few minutes possible with staggered ACAMS. The Army has contracted for the development and demonstration of a Fourier Transform Infrared (FTIR) multipass absorption technique, which should be capable of real-time (~1 second) detection of high agent release levels. The sixth specific recommendation was: Evaluate the benefits of more frequent analysis of facility stack gases for nonagent trace contaminants. (MON-11) Daily analysis of samples of flue gas for products of incomplete combustion, particulates, and metals may provide an important database to monitor incinerator operating conditions and the environmental impacts of the disposal system. Although analysis of these emissions is required only in trial burns, more frequent characterization would provide additional information about incineration performance and might provide further assurance to the local population. The schedule for the first 24 months of TOCDF operations is dominated by required trial burns. At least 16 trial burns (each consisting of multiple runs) are scheduled for the first two years. During these trial burns, samples will be collected and analyzed for products of incomplete combustion (PICs), particulates, heavy metals, and volatile and semivolatile PICs, such as chlorinated dioxins and furans. Data from the trial

OCR for page 11
--> burns will be used to modify the health risk assessment conducted for the TOCDF. This information and the revised health risk assessment will then be used to develop an enhanced monitoring program for compounds identified as potentially significant. This is appropriate for the TOCDF, even though JACADS exhaust analysis indicates that the level of products of incomplete combustion will be very low. The Army appears to be committed to reassessing the sampling plan after conducting a health risk analysis of the data from the TOCDF trial burns. Specific Recommendations for Laboratory Operations The remaining specific recommendations addressed laboratory operations. The first recommendation of this group was the seventh specific recommendation: Increase the automation of sample handling and laboratory operations to ensure better quality control and efficiency. (MON-12) The JACADS laboratory operations involve extensive handling of samples and manual data entry. Implementation of suitable and commercially available laboratory automation procedures could significantly enhance the efficiency and reliability of the laboratory. Observers of laboratory operations at JACADS raised concerns that the process for handling DAAMS tubes appeared to be unnecessarily prone to error. Several steps have been taken at the TOCDF to improve these procedures. First, although desorption from the 8-mm DAAMS tubes to 4-mm tubes used for GC/FPD analyses is still required, a heated desorber has been developed and tested to mechanize this process. Ten tubes at a time can be desorbed with greater consistency than was previously possible. Second, much better control is exercised over the DAAMS tubes themselves. They are now individually barcoded, and a database is maintained on all uses and analyses by tube number. This should reduce errors of misidentified tubes, and allow tracking of tubes for gradual changes in agent retention or other deterioration of performance. The eighth specific recommendation was: Give laboratory personnel a variety of tasks that ensure optimal attention and performance. (MON-13) Repetitive tasks tend to decrease an operator's ability to detect unusual analysis results. An effective way to keep operators alert would be to include daily analysis of products of incomplete combustion in exhaust stack emissions. Laboratory personnel can be assigned a variety of tasks to ensure that they remain alert and to provide breaks from repetitive tasks, particularly DAAMS tube analyses. One option is adding analysis of non-agent stack emissions. This analysis is now performed at the TOCDF, but the new task is not being used as part of a regular within-shift rotation for DAAMS analysts. A necessary condition for job enrichment is cross-training, which has not been completed in the laboratory operations because of the time pressures during systemization. There is some movement of operators between jobs on a longer-term basis, but this does not meet the requirement for varying tasks within a given shift. The operators themselves were recruited from the local area, as well as from other chemical demilitarization and chemical weapons programs (e.g., CAMDS, JACADS, and Dugway Proving Ground, Utah). They consider their training at the Chemical Demilitarization Training Facility as good preparation for their jobs at the TOCDF because there are only minor differences between the equipment and procedures at the training facility and the disposal facility. The ninth specific recommendation was: Give blind challenges to the laboratory. (MON-14) The previous quality control procedures involved regular agent challenges to the laboratory chromatographs. These are more a calibration exercise than a verification that agent will be detected. Additional, unexpected challenges were recommended. The TOCDF laboratory program, run by quality control personnel, includes randomly spiking DAAMS tubes with agent. Each challenge is chosen from a wide range of levels to prevent complacency among operators. Each operator receives at least one challenge per day. Both operators and their supervisors are blind to which tube is a challenge and to the level of the challenge. The tenth specific recommendation was: Perform a detailed error analysis of the laboratory system and procedures. (MON-15)

OCR for page 11
--> Reliance on manual data entry and sample handling in the laboratory is a source of potential errors. A detailed error frequency analysis may reveal the sources and reduce the rate of errors in laboratory operations. Reports of high error rates observed in JACADS laboratory operations prompted the committee to recommend an analysis to determine possible human errors and their consequences. At the TOCDF, a system hazard analysis has been conducted and presented in an Army report to the Utah DEQ as part of the RCRA permitting process, but this does not appear to cover errors in laboratory activities. Although there is evidence of a satisfactory response to detected errors, there is still a need for a system-oriented approach concentrated on potential errors. The Army contracted with MITRE to perform a laboratory error analysis during a five month period from May to September 1995. This analysis was scheduled to begin at JACADS and extend to laboratory operations at TOCDF. If performed adequately, this analysis should answer the Stockpile Committee's concerns. Summary of Responses to Monitoring Recommendations The Army has made considerable progress in responding to both the general and specific recommendations presented in the Monitoring report. Recommendations that could be addressed wholly or in part with existing commercially available instrumentation have generally been effectively addressed and implemented, with most solutions tested at JACADS prior to adoption at the TOCDF. The response to longer-range recommendations that require monitoring technology R&D has been more tentative. However, the monitoring improvements in the longer-range recommendations are not required for successful and safe operation at the TOCDF or other U.S. sites. They would, however, make operations easier, more efficient, and more reassuring to the public. The Army and Tooele County now have a direct line of communication for informing civilian emergency management personnel of any alarms or alerts concerning agent release or other emergency incidents. An isolated false positive alarm requires stopping agent feed operations on-site but does not require activation of emergency response operations. A release large enough to threaten the immediate response zone (IRZ) would be quickly obvious on-site. A large release is likely to trigger several ACAMS alarms or to trigger the same ACAMS alarm several times. "Multiple" alarms are not characteristic of the response to false positives, which trigger sporadic, single alarms. Therefore, the response appropriate for a major agent release should not be triggered by sporadic ACAMS false positive alarms. At other sites, where the disposal facility is in closer proximity to communities, the emergency response may have to be activated sooner following an initial alarm. Thus, the false positive problem may have to be solved prior to the start of agent operations at those sites. Committee visits to the TOCDF have included thorough discussions with laboratory and other monitoring personnel. Their training appears to be effective, and the equipment and operating procedures include significant improvements over those used during JACADS operations. The monitoring and supporting analytical laboratory capabilities in place at the TOCDF can support safe operations. Recommendation on Carbon Filtration After examining the treatment of stack gases emanating from the baseline incineration system, the Stockpile Committee, in Recommendations for the Disposal of Chemical Agent and Munitions (NRC, 1994c) found that: The Stockpile Committee finds the baseline system to be adequate for disposal of the stockpile. Addition of activated carbon filter beds to treat all exhaust gases would add further protection against agent and trace organic emissions, even in the unlikely event of a substantial system upset. If the beds are designed with sufficient capacity to absorb the largest amount of agent that might be released during processing, addition of these beds could provide further protection against inadvertent release of agent. Consequently, the committee made the following recommendation: The application of activated charcoal filter beds to the discharge from baseline system incinerators should be evaluated in detail, including estimations of the magnitude and consequences of upsets, and site-specific estimates of benefits and risks. If warranted, in terms of site-specific advantages, such equipment should be installed. (REC-13)

OCR for page 11
--> The principal focus of this recommendation was to consider whether charcoal filters might provide an additional safety factor at continental U.S. sites with relatively large nearby populations. Although the filters might reduce some nonagent emissions and could provide additional protection against plant upsets, the filters might also create additional risks if they caught fire, for example. The Army has initiated a review of various design options and related systems performance evaluations. If the evaluations are positive, the Army has chosen the TOCDF as the site for a demonstration unit; a primary reason for the selection of this site was the availability of an exhaust gas stream for testing purposes. This choice does not appear to be a prejudgment on whether a full-scale application of a carbon filter system would be warranted at the TOCDF or elsewhere.