Review and Assessment of Developmental Issues Concerning the Metal Parts Treater Design for the Blue Grass Chemical Agent Destruction Pilot Plant (2008)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Summary The United States is in the process of destroying its concerning the MPT design. The full statement of task for chemical weapons stockpile. In 1996, Congress mandated the Committee to Review and Assess Developmental Issues that the weapons at two sites, Blue Grass Army Depot in Concerning the Metal Parts Treater Design for the Blue Kentucky and Pueblo Chemical Depot in Colorado, would Grass Chemical Agent Destruction Pilot Plant is given in not be destroyed by incineration and that the Department of the Preface. Defense should demonstrate and select alternative methods. Contracts to design, build, operate, and close both pilot In 1999, Congress also passed Public Law 105-261, which plants were awarded to Bechtel International, teamed with required that the Under Secretary of Defense certify “in writ- Parsons Engineering. For PCAPP, Parsons is a subcontrac- ing to Congress” that the alternative technology would “be as tor to Bechtel. For BGCAPP, Bechtel and Parsons formed safe and cost effective for disposing of assembled chemical a joint venture called the Bechtel Parsons Blue Grass Team munitions as is incineration of such munitions. . . .” (BPBGT) and are teamed as prime contractors. These con- The Assembled Chemical Weapons Alternatives tractors tailored the specific design of the respective facilities (ACWA) program was established in response to these man- to the content of each site’s stockpile. dates. Because the selected alternatives at each site would be new applications of existing technologies, the Army the Metal Parts Treater and the designated the facilities used to implement the alternatives Munitions TreatMENT Unit as pilot plantsthe Blue Grass Chemical Agent Destruc- tion Pilot Plant (BGCAPP) and the Pueblo Chemical Agent Originally the MPT was planned for use at both Destruction Pilot Plant (PCAPP). BGCAPP and PCAPP to decontaminate metal parts. The The Program Manager for Assembled Chemical Weap- current PCAPP design now calls for a munitions treatment ons Alternatives is overseeing the efforts of systems con- unit (MTU) to decontaminate projectile and mortar casings. tractors to develop and test equipment to be used in the The MPT uses radiant and convection heating in an enclosed designs for constructing the two disposal pilot plants. Among metal cylinder to raise metal parts to 1000°F for a duration the first-of-a-kind equipment under development for the of 15 minutes in order to decontaminate these materials. BGCAPP are two metal parts treaters (MPTs), which would Steam is used as a carrier or sweep gas to remove vapors and be used primarily for the treatment of washed-out metal particulates released during heating. The intention is also to munitions cases from which the agent has been drained. The use the MPT to treat contaminated secondary waste before MPTs could also be used to treat secondary waste gener- its off-site disposal. ated during the destruction operations and waste materials The two MPTs being developed for the BGCAPP are generated during facility closure operations. During recent testing, results have shown the heat-up times of trays of mu- The stockpile at the Blue Grass Army Depot includes two nerve agents nition casings to be longer than expected. Another issue that and a blister agent in various types of munitions. The agents are nerve agent has developed involves problems in sealing the MPT as the sarin (GB) (C4H10FO2P), nerve agent VX (C11H26NO2PS), and Levinstein temperature inside is increased. The Program Manager for mustard agent (H) (C4H8Cl2S). The nerve agents are in M55 rockets and 8- inch or 155-mm projectiles. The mustard agent H is in 155-mm projectiles. Assembled Chemical Weapons Alternatives requested that The stockpile at the Pueblo Chemical Depot consists of 105-mm boxed the National Research Council (NRC) form a committee to cartridges, palletized 155-mm projectiles, and 4.2-inch mortars, all of which review ongoing testing to investigate and determine causes are filled with one of two forms of mustard agent: distilled mustard agent for the longer-than-expected heat-up times and other issues (HD) or mustard agent HT.  

 Review and Assessment of Developmental Issues Concerning the Metal Parts Treater Design designated as first-of-a-kind equipment because they are to gather data for further effort with the CFD model. unique, are being designed for this particular application, and •  erform test runs and cycles of components to make ob- P have never been used in an actual process. A small-scale test servations of critical design parameters that apply to the unit called the Technical Risk Reduction Program (TRRP) FOAK unit under design—particularly those that affect the risk of scale-up to the full-scale unit. These include, MPT was fabricated to demonstrate the operation of the MPT but are not limited to projectile paint debris generation and concept for the decontamination of metal parts and waste. In accumulation, thermal expansion stresses and deformation this report, the phrase “first-of-a-kind” or “full-scale MPT” points, interferences, Gaussian field measurements and lo- is used to describe the full-scale equipment, and the term calized heating effects, and wall temperature distribution. “TRRP MPT” is used to describe the three-quarter-scale version. The TRRP MPT testing used an off-gas treatment sys- The PCAPP design calls for an MTU to decontaminate tem that included a catalytic oxidizer unit rather than a bulk the projectile and mortar casings. The MTU, an adaptation oxidizer unit (more accurately called a flameless thermal of a metal annealing oven, is a continuous-belt muffle-type oxidizer) and did not include the venturi scrubber. Thus, the oven with material-handling equipment at the feed and flow of off-gas from the MPT enclosures was demonstrated, discharge ends. Modifications include new feed and exit but not the off-gas treatment system configuration or equip- sections and a muffle large enough in cross section (10 in. ment that will be provided for the full-scale MPT. The off-gas high by 30 in. wide) to accommodate 155-mm projectile treatment system bulk oxidizer unit is also considered to be bodies riding on the metal conveyor. The muffle section is a first-of-a-kind system. also long enough to ensure that all parts of the munitions TRRP MPT testing was performed using surrogates of reach 1000oF for at least 15 minutes at the operating speed all munitions metal parts and waste feed streams anticipated of the metal conveyor. for the two BGCAPP full-scale MPTs. All feed streams were The committee was not charged with evaluating the tested. However, the BPBGT terminated the waste stream MTU in detail. What this report presents is a technical testing before the completion of all planned tests because description and evaluation of the MPT (Chapters 2-4) plus it was believed that sufficient data to design the full-scale an evaluation of the technical feasibility of replacing the MPT had been obtained. All feed streams were tested to the MPT with an MTU and supplemental decontamination units extent allowed by existing permits at the Parsons fabrication and autoclaves such as those being designed and tested for facility. The permitting limitation prevented testing of the PCAPP (Chapter 5). energetics batch hydrolyzer waste with energetics remnants and halogenated materials. Assessment of Metal Parts Treater Testing During testing, the TRRP MPT unit experienced re- Activities curring operating problems, such as mechanical failures and munitions bodies taking longer than expected to reach The MPT concept has been subjected to testing in the necessary high temperatures as estimated by computer TRRPs, with most of the pertinent testing conducted under modeling. the Bechtel TRRP 05c test plan (BPBGT, 2007d). This test- The committee grouped the MPT test results into three ing has used the three-quarter-scale TRRP MPT. areas for review and evaluation: (1) mechanical issues and The testing objectives as given in the TRRP 05c test plan (2) secondary and closure waste treatment issues, which are were as follows (BPBGT, 2007d): assessed in this section, and (3) results of thermal testing, modeling, and predicted throughput of the MPT, which are •  emonstrate reliable mechanical performance of all parts D assessed further below. and functions of the MPT design, including seals, doors, bearings, and projectile jamming. •  emonstrate BGCAPP-specific design improvements D Mechanical Issues such as: projectile orientation, steam-injection orientation, gas take-off orientation, and tray design to improve heat- New Door Closure Mechanism and Seals up. •  alibrate the computational fluid dynamics (CFD) model C Difficulties with getting an acceptably tight closure on of the test unit on VX 155-mm projectiles to serve as a the air lock and main chamber doors for the TRRP MPT basis for first-of-a-kind (FOAK) full-scale unit modeling. have resulted in a change in the design of the door closure Inherent in this objective is the necessary demonstration mechanism and seals for the full-scale MPT. Instead of the that the MPT can heat all parts of materials fed to it to J-type sliding closure mechanism used on the TRRP MPT, 1000oF for at least 15 minutes at a rate that meets expected the door for the full-scale MPT will be moved against the feed rates during operation. closure face by using a two-direction cam design recom- •  emonstrate treatment of simulated energetics batch hy- D drolyzer (EBH) rocket warhead debris. mended by a commercial oven contractor. In addition, the •  emonstrate limited secondary-waste treatment options D seal material design has been altered to give the equivalent of two gaskets between the door and closure face. 

SUMMARY  Bearings for the Conveyor Rollers or (2) if the waste has been subjected to thermal treatment at 1000°F for 15 minutes. The second criterion, formerly called The Graphalloy® bearings for the conveyor rollers in the treatment to 5X, was a requirement for off-site shipment until main chamber experienced galling and other wear failures June 2004 when the criteria on WCL were introduced. The attributed to oxidation/corrosion at the main chamber operat- BPBGT plans to heat all secondary waste to 1000°F for at ing temperature. Three different bearing materials were eval- least 15 minutes in the MPT. At that temperature, in the very uated: “improved” Graphalloy®, Stellite, and Deva (Deva- low oxygen activity environment of the MPT, many second- Mogul sintered metal). All materials experienced wear, and ary waste materials will pyrolyze, leading to the formation the BPBGT concluded that the “improved” Graphalloy® of chars and tars. By lowering the temperature to ~500o F for bearings were acceptable, although they exhibited some 1 to 2 hours, six nines (99.9999 percent) agent destruction pitting. During full-scale MPT testing, the BPBGT intends and removal efficiency should be achievable, and char and to reconsider Stellite bearings that are interchangeable with tar formation should be greatly reduced. the Graphalloy® bearings. The BPBGT has also developed maintenance protocols that shorten replacement times for bearings as much as possible. Results of Thermal Testing, Modeling, and Predicted METAL PARTS TREATER Throughput Heating Zones TRRP testing of secondary waste treatment in the MPT was conducted at the Parsons fabrication facility in Kenne- The full-scale MPT will use two heating zones in the wick, Washington, in May 2007. The test results were gen- main chamber. Each will be capable of about 450 kW of in- erally favorable. However, the range of waste types treated duction heating. The TRRP MPT used one 600-kW induction was narrow, and the total amount of waste treated was small. heater. The use of two-zone heating in the full-scale MPT Thus, the committee recommends that the BPBGT perform main chamber should improve heating rates and control of more comprehensive testing during systemization, using heating. It is unclear whether maintenance on one MPT will waste representative of that encountered during closure as be possible while the other is in operation. If not, the avail- well as various types of secondary waste from operations and ability of the MPTs would be reduced, since both would maintenance. Heat-up times for projectiles located in the trays have to be shut down if either required in-room maintenance. passing through the TRRP MPT were measured and predicted TRRP testing and CFD modeling showed that certain areas to demonstrate that the metal parts at all locations in the tray of some projectiles were heating more slowly than the rest could be heated to 1000°F for at least 15 minutes within a of the projectiles. This slower heating required longer heat- total time duration that supported the design production rate. up times to achieve a uniform 1000°F for 15 minutes for The experimental data and the modeling focused on the tem- all projectiles. After its review of the chamber design, the perature-time profiles for projectiles at specified locations in BPBGT concluded that the slow heating resulted from “shad- the tray. The key elements of this effort were as follows: owing” of parts of the projectiles during the radiant heating process. The shadowing is being addressed by redesign of •  hermocouple measurements of the surface tempera- T the projectile trays, the superheated steam inlet header, and ture were made on a limited number of projectiles at the off-gas outlet header. specific locations in the tray as they passed through the MPT and were heated. Secondary and Closure Waste Treatment Issues •  redictions of the temperature distribution in the pro- P jectiles were made by using a CFD thermal modeling Waste to be treated in the MPT includes the washed program that was compared with a previous model munitions bodies from the munitions washout system, solid used in initial PCAPP MPT testing. However, all residues from the energetics batch hydrolyzer, and secondary model results must be validated with experimental and closure waste. Agent-contaminated waste will be treated data, as discussed elsewhere in this report. by chemical decontamination. When chemical decontamina- tion cannot be used, waste (e.g., agent-contaminated pallets) will be treated in the MPT before off-site disposal. Secondary Experimental Temperature Measurements waste that is not agent-contaminated is not expected to be Type K thermocouples encased in stainless steel sheaths processed through the MPT. were mounted to the top and bottom of three projectile cas- In general, secondary waste can be shipped off-site safely if it meets one of two criteria: (1) if analysis shows by EPA’S toxicity characteristic leachate procedure (TCLP) applied to the levels less than the applicable waste control limits (WCLs) residuals from the metal parts treater. The WCL may also, or additionally, be based on agent concentration in the air space above the containerized WCLs and the analytical methods required to demonstrate that they waste treatment residuals. Minimum required levels are typically 1 STEL have been achieved vary by state. In general, the WCL is defined as 20 (short-term exposure limit )—0.0001 mg/m3 for GB, 0.00001 mg/m3 for parts per billion (ppb) for GB and VX and 200 ppb for HD, as determined VX, and 0.003 mg/m3 for HD. 

 Review and Assessment of Developmental Issues Concerning the Metal Parts Treater Design ings in order to allow the monitoring of temperature-time •  ffective thermal treatment of secondary waste with- E profiles in the prototype TRRP MPT. Selection of the loca- out excessive fouling of the duct work leading to the tions on the casings was aided by use of the CFD thermal bulk oxidizer; model to identify potential cool spots. •  uccessful integration of the MPT with its flameless S thermal oxidizer (i.e., the bulk oxidizer) and cyclone; and Temperature Prediction by Computational Fluid Dynamics •  omplete destruction of energetic materials in the C Thermal Modeling waste stream of the energetics batch hydrolyzer The BPBGT used a mathematical model for comparison without adversely affecting the MPT. with the TRRP experimental measurements and to predict the performance of the full-scale MPT for BGCAPP. The Finding. The current range of heat-up times of munitions in BPBGT model gives spatial and temporal temperature the MPT should not affect the overall schedule of BGCAPP behavior of the parts being processed in the MPT. The pur- operations. pose of the modeling was to show that the MPT design was adequate for treating munitions at 1000°F for 15 minutes •  eat-up times in the TRRP tests are close to target H while meeting operational and schedule requirements and and appear to be capable of being improved by rais- that the design could guide the scale-up and the testing of ing the wall temperature of the full-scale MPT. the full-size unit. Comparison with the experimental mea- •  FD modeling predicts correct trends in temperature- C surements was used to validate and modify the model. The time profiles and locations of cold spots and should improved model appears to be fairly rigorous. The code is be useful in guiding the design and testing of the based on a nonlinear solver that (1) is accurate to second full-scale MPT. order in space and time; (2) globally and locally conserves •  he processing rate of projectile bodies in the MPT is T mass, momentum, and energy; and (3) allows a choice of not on the critical path of the process throughput. The finite-element shape function. The model handles multimode design calls for two MPTs. The second is intended heat transfer, including graybody radiation with view factor to be used for secondary waste, but it could also be computation. used for treating munitions bodies in an emergency. Finding. The MTU could be substituted for the MPT Comparison of the metal parts treater and at BGCAPP; however, it would be necessary to do the munitions treatment unit FOR BGCAPP following: The committee reviewed the applicability of the MTU as an alternate method for decontaminating munitions bodies •  se supplemental decontamination units and auto- U and secondary waste at BGCAPP. As noted above, the MTU claves to treat secondary waste, is currently planned for installation at PCAPP for the thermal •  ind another means of treating the detonators in the F decontamination of 155-mm and 105-mm projectiles and 4.2- M417 rocket fuzes, in. mortars that have been drained of mustard agent and passed •  odify the MTU design to accommodate 8-in. M through a high-pressure wash. Table S-1 compares various projectiles, operating requirements and features of the MPT and the MTU •  odify the footprint of the building to accommodate M and identifies changes that would be required for the MTU to be the units, and used at BGCAPP. The same table appears in Chapter 5, which • Modify the existing permits. gives additional supporting and clarifying information. Finding. For BGCAPP, the TRRP testing did not address a method to thermally treat the fuzes and a limited number of General Findings and Recommendations igniters from contaminated propellant that are not decom- Finding. The full-scale MPT as currently designed for posed in the energetics batch hydrolyzers. BGCAPP can decontaminate projectile bodies and second- ary and closure waste, and it will be able to achieve its target Recommendation 6-1. The BPBGT needs to develop a throughput rates provided that the BPBGT is able to resolve method to collect and pop the igniters and fuzes that will not the following issues: adversely affect the operation of the MPT. •  uccessful implementation of new designs for door S Recommendation 6-2. To reduce the technical risks in treat- closure and seals, for roller bearings on conveyors, ing secondary waste in the MPT, the BPBGT should continue and for the superheated steam header; to strive to send secondary waste off-site whenever possible and minimize the use of halogenated materials. 

SUMMARY  Recommendation 6-3. To reduce the load on the off-gas still required. These changes and improvements may require treatment system, the BPBGT should consider obtaining several iterations before satisfactory results are achieved. permits that allow the use of the Airborne Exposure Limit Guidelines to operate the MPT at lower temperatures for the Recommendation 6-4. A test plan should be prepared for all thermal treatment of secondary waste whenever possible. design changes identified for the MPT but not verified in the TRRP tests. This test plan should be conducted at the fabrica- Finding. TRRP testing demonstrated the validity of using tion facility and should include time for the repeated trials an MPT for thermal decontamination. It also identified needed to arrive at acceptable performance for the overall many changes and design improvements that will be neces- full-scale MPT. A similar test plan should be prepared for sary to achieve an acceptable throughput rate and control testing the integrated full-scale off-gas treatment system for maintenance and operating costs. Further testing with varied the MPT at BGCAPP. This test plan should include the test- secondary waste materials including halogenated waste is ing of a full range of secondary, energetics batch hydrolyzer, and closure waste at the full-scale MPT design rates. TABLE S-1  Comparison of the Metal Parts Treater and the Munitions Treatment Unit Characteristic MPT at BGCAPP MTU at PCAPP Changes for MTU Use at BGCAPP Status of testing Prototype (~3/4 scale) Full-scale unit completed Full testing would be required, using demonstrated on surrogate acceptance tests at the the energetic dregs and agent or munitions and waste streams. manufacturer’s facility with appropriate surrogate material. Two full-size MPTs will be built surrogate munitions. Only small design changes from the and tested at the manufacturer’s PCAPP MTU are required. Using the facility. MTU would save extensive testing of the MPT with secondary waste. Treatment of secondary waste in the SDUa and autoclaves has already been performed at the other sites. Feed streams 4.2-inch mortars and base plates None 97,106 (HD) Applies only to PCAPP, no mortar rounds at BGCAPP. 105-mm projectiles None 383,418 (HD) Not applicable to BGCAPP. 155-mm projectiles 15,492 (H) 299,534 (HD) A new permit will be needed to use 12,816 (VX) the MTU at BGCAPP. 8-inch projectiles 3,977 (GB) None Muffle height must be increased for use at BGCAPP. The MTU currently has internal height of 9.75 in. and width of 30 in. M55 rockets; undissolved 51,716 (GB) None Squibs and fuzes must be thermally fragments, including 17,757 (VX) decomposed (must be popped). Some undissolved squibs and fuzes Method for treating squibs and fragments are combustible and may from hydrolysis of rocket fuzes in the MPT is to be tested. require inert gas and special baskets if warhead and rocket motor treated in the MTU or, alternatively, segments in EBHsb they may be treated in the SDU or autoclave if WCLc guidelines for off- site disposal are used. Secondary waste Thermal treatment in MPT using Treatment in SDU or autoclaves. SDU and autoclave use based on special carrying trays. approach used at ABCDF.d Use at BGCAPP may require special permitting and higher temperatures for GB- and VX-contaminated waste. Only limited testing on various waste types has been performed in the MPT. Closure waste Thermal treatment in MPT. Treatment in second SDU. See comments for secondary waste. Agents destroyed   Mustard Yes Yes continues 

 Review and Assessment of Developmental Issues Concerning the Metal Parts Treater Design TABLE S-1  Continued Characteristic MPT at BGCAPP MTU at PCAPP Changes for MTU Use at BGCAPP   GB and VX Yes None SDU and autoclave would be required with the MTU for GB- and VX- contaminated waste streams. This method of decontamination may require higher treatment temperatures to achieve acceptable treatment times to meet WCL guidelines for off-site disposal. Number of units 2 MPTs with one for projectile 2 MTUs with only one ordinarily BGCAPP might require only 1 MTU processing and one for waste required to meet PCAPP because the number of projectiles is streams and as a backup spare. processing rates. an order of magnitude less than at PCAPP; MTU availability expected to be higher than that of the MPT. MDBe footprint ~70 ft long × 40 ft wide × 20 ft ~100 ft long × 30 ft wide × 20 ft MTU would require changes high high in BGCAPP MDB layout to accommodate longer processing Total direct footprint for 2 MPTs Total direct footprint for 2 MTUs length and SDU and autoclave units plus 50 percent of MPT/washout is 5,100 square feet plus 216 and to provide for collection bins for support room plus MPT cooling square feet for collection bin receiving treated metal parts from room is 4,640 square feet. enclosures. MTU discharge chute. Post-treatment agent clearance In exit air lock before leaving level In treated munition collection bin MTU use at BGCAPP would require A area. in Level D area. permit change for use of current MTU discharge configuration. Atmosphere in unit Nitrogen in air locks and Air flowing from both ends of Change in Kentucky permit would be superheated steam in main muffle to off-gas duct exit from required for use of MTU and SDU or chamber. muffle system. autoclave. Off-gas treatment system Flame arrestor, cyclone, bulk Flame arrestor, filter media, bulk Same components, with different gas oxidizer, venturi scrubber, and oxidizer, venturi scrubber, and flow rates and sizes. reheater. reheater. Method of heating 2 induction coils at 450 kW each Resistance heaters at 600 kW None plus 75 kW resistance heating for steam superheater. Method of operation Batch Continuous Would have to change from a batch stream to a continuous stream. Char and tar buildup Expected, but the design allows for Secondary waste not processed in Secondary waste not processed in from secondary waste addressing tar and char buildup. MTU. MTU. treatment Method of control of Doors and seals on air locks Curtains on munitions cold feed Curtains on munitions cold feed end atmosphere in main attached to main chamber. end and cooling section exit of and cooling section exit of muffle. treatment unit muffle. Overall availability 83 percent estimated using spare 91 percent estimated for single Both estimates based on using both (percent of time the MPT. MTU; no estimate given with treatment units. system will be operating) spare MTU. Munitions throughput rates 4.2-inch mortars None for BGCAPP ~60/hr 105-mm projectiles None for BGCAPP ~60/hr 155-mm projectiles ~40/hr ~40/hr ~40/hr 8-inch projectiles ~15/hr None MTU should be capable of modification to achieve BGCAPP 8- inch projectile processing rates. aSDU,supplemental decontamination unit. bEBH,energetics batch hydrolyzer. cWCL, waste control limit. dABCDF, Aberdeen Chemical Agent Disposal Facility. eMDB, munitions demilitarization building. SOURCE: Adapted from BPBGT, 2007c.