Assessment of Supercritical Water Oxidation System Testing for the Blue Grass Chemical Agent Destruction Pilot Plant (2013)

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 Army Element, Assembled Chemical Weapons The Program Executive Office (PEO), ACWA, of the Alternatives (ACWA), is responsible for managing the con- Department of Defense requested that the National Research duct of destruction operations for the remaining 10 percent Council (NRC) review and evaluate the results of the FOAK of the nation’s chemical agent stockpile, stored at the Blue tests conducted on one of the SCWO units to be provided Grass Army Depot (Kentucky) and the Pueblo Chemical to BGCAPP. In addition, the NRC was asked to provide Depot (Colorado). Facilities to destroy the agents and their recommendations for systemization of the SCWO system, associated munitions are currently being constructed at these an assessment of actions taken in response to the 2012 NRC sites. The Blue Grass Chemical Agent Destruction Pilot Plant report The Blue Grass Chemical Agent Destruction Pilot (BGCAPP) will destroy chemical agent and some associated Plant’s Water Recovery System (Letter Report), and how energetic materials by a process of chemical neutralization systemization of the SCWO system and WRS might be af- known as hydrolysis. The resulting chemical waste stream fected by recommendations from that prior report. is known as hydrolysate. To accomplish this task, the NRC established the Com- Among the first-of-a-kind (FOAK) equipment to be mittee on Assessment of Supercritical Water Oxidation Sys- installed at BGCAPP are three supercritical water oxidation tem Testing for the Blue Grass Chemical Agent Destruction (SCWO) reactor systems. These will be used to further treat Pilot Plant (the SCWO committee). This report presents the the agent and energetics hydrolysates generated by hydro- SCWO committee’s background discussions, findings, and lysis of sarin (GB) and VX nerve agent, possibly mustard recommendations pursuant to these tasks. The committee’s blister agent, and the energetics associated with M55 rocket statement of task was as follows: warheads and with various projectile munitions. The SCWO systems will subject the hydrolysate feeds to very high tem- The National Research Council will establish a committee peratures and pressures, breaking down their organic content for an assessment of First-Of-A-Kind factory acceptance test into carbon dioxide, water, and salts. These particular hydro- performance results of the General Atomics SCWO system lysate feeds present unique non-agent-related challenges to design and recommendations for testing during systemiza- tion at BGCAPP: subsequent processing via SCWO due to their caustic nature and issues of salt management. •  xamine test reports and results as made available within E The potential problems with corrosion and salt buildup the timeframe of this study with respect to verification of prompted ACWA to require extensive FOAK testing of the the functionality of tested SCWO system components, SCWO process prior to shipment, installation, and system- operability under normal and abnormal conditions, and ization at BGCAPP. Ten months were allocated for testing performance in meeting throughput and other operational SCWO Unit 1 to verify the full-scale design and to continue requirements; prior work in furtherance of mitigating performance risks. •  rovide an evaluation of how well the test data supports P These tests used hydrolysate simulants representative of conclusions reached; the actual agent and energetics hydrolysates that will be •  rovide recommendations on SCWO systemization test- P processed at BGCAPP. During this testing, not only were ing at BGCAPP inclusive of durability testing; and •  iscuss systemization testing objectives and concepts in D corrosion and salt buildup examined, but changes in chemi- consideration of committee findings from the BGCAPP cal additives to the SCWO feed and in SCWO operating Water Recovery System (WRS) study. conditions were tested. 1 

2 ASSESSMENT OF SUPERCRITICAL WATER OXIDATION SYSTEM TESTING FOR THE BGCAPP After chemical neutralization in the main part of the mustard hydrolysate is the least corrosive. After the SCWO BGCAPP plant, the agent and energetics hydrolysates will process, the SCWO effluent will be cooled by quench water be sent to hydrolysate tanks, where they will be stored until recycled from the WRS. This is to manage the salt buildup they are blended for further processing in the SCWO reac- problem. Finally, the temperature and pressure are let down tors. Aluminum will also be removed from the energetics and the gaseous and liquid effluents are separated. At let- hydrolysate prior to blending. Also, chemical additives are down, the salts dissolve into the water. added to allow the flow of salts through the reactors. In the The liquid SCWO effluent, water with dissolved salts SCWO reactors the blended hydrolysate will be mixed with and possibly other solids (depending on the hydrolysate supercritical water, air, and a fuel and exposed to conditions ­being processed), will be passed through a WRS for purifica- of roughly 1200°F and 3,400 psia. The resulting oxidation tion. Coagulant and prefilters will be used to remove solids will convert elements to their most stable oxidized state. from the effluent. The resulting salt-containing water will Thus, carbon is oxidized to carbon dioxide, hydrogen to be passed through a reverse osmosis membrane to filter out water, sulfur to sulfates, and so on. However, salts such as the salts. At least 70 percent of the effluent will be passed sodium sulfate will not dissolve in supercritical water and through the membrane and emerge as permeate with a level can block outlet orifices and coat the walls if not properly of total dissolved solids of less than 500 mg/L. What does managed. not pass through the membrane will become a concentrated The supercritical hydrolysate streams in the reactor can brine that will have to be disposed of off-site. A portion of be very corrosive. Thus, a sacrificial titanium liner is inserted the permeate will be recycled to the SCWO reactors as de- into the reactor to protect the outer pressure vessel walls. The scribed above. Figure S-1 depicts the combined SCWO and titanium liners will have to be replaced periodically owing to WRS systems. corrosion. Sarin (GB) hydrolysate is the most corrosive and The SCWO committee, after extensive study of test Offgas HP air HP air to filter compressor manifolds Condensate from Offgas other processes duct Agent hydrolysate heater EnergeƟcs hydrolysate HP fuel pump Ipa pump Gas Salt analyzers Solid feed addiƟves S Hydrolysate Hp hydrolysate HP hydrolysate Acids blend tanks pump Pump Wet feed addiƟves NaOH HP water pump TOC Hydrolysate blend/ HPQuench pump quench pump analyzer SCWO Batch holding batch holding effluent tank tank heaƟng skid Feed Module tank LP gas- Lp gas SCWO liquid Liquid reactor separator Separator Emergency relief tank HP gas- Hp gas - liquid Liquid Off -spec separator Separator tank Reactor Effluent heat exchanger Module Antiscalant Coagulant RO units (3) RO RO permeate storage From SCWO effluent tanks To RO permeate tanks SCWO Multimedia Canister High- effluent filters filters pressure tank (6) (3) pump RO reject pumps Recycle pump To RO reject tanks FIGURE S-1  Flow diagram of SCWO system and reverse osmosis WRS. SOURCE: Adapted from Dan Jensen, advanced process system program manager, General Atomics, “SCWO System Equipment and Layout,” presentation to the committee, January 7, 2013; and NRC, 2012. Cooling tower and Water steam softener blowdown Spent regenerant to RO reject tank 

SUMMARY 3 documents, briefings by contractor and BGCAPP staff, plant modification that would lengthen the thermowell life during visits, and much deliberation arrived at the following find- operations and help minimize maintenance-related reactor ings and recommendations. They are arranged in the order shutdowns. in which the appear in the report. Finding 2-5. As noted above, the corrosion of thermowells is greater for higher feed velocities than for lower feed veloci- FINDINGS AND RECOMMENDATIONS ties. Another way to address the corrosion of thermowells Chapter 2 of this report reviews and assesses the results could be by reducing the feed velocity. of FOAK testing. The committee had discussions with the BGCAPP project staff and the SCWO system contractor and Finding 2-6. Any new alloy used for the thermowells, or any reviewed a significant amount of information, including the coatings applied to the thermowells, would have to be quali- preliminary FOAK test report. These are the findings and fied before use in the SCWO system. There is no guarantee recommendations resulting from this. that corrosion performance would improve. Using a lower processing temperature for sarin (GB) processing and using an alternative geometry for the titanium cap would likely Corrosion necessitate further testing to ascertain whether these changes Finding 2-1. The FOAK testing was adequate to establish would affect SCWO processing operations. the expected operating lifetime for the titanium liners. This expected operating lifetime is sufficient to enable continued Finding 2-7. Based on the test data, the committee believes safe operations and adequate operational throughput except that availability goals can be met with grade 2 titanium for VX hydrolysate. thermo­ ells of appropriate wall thickness. w Finding 2-2. The 300-hr liner change-out interval recom- Recommendation 2-3. The corrosion issue with thermo­ mended for the operational GB hydrolysate campaign is wells should not be resolved by attempting to qualify another adequate if the operating temperature remains at the recom- material. Grade 2 titanium should be used as planned. Other mended 1150°F and the flow rate is 1,000 lb/hr. The 300-hr potential strategies for extending thermowell life (i.e., coat- change-out interval is not adequate if a higher operating ings, alternative geometries, reduced operating temperature, temperature (1200°F vs. 1150°F) is reached in the reactor for and hardened titanium alloys) need not be explored since significant lengths of operation. It also may not be adequate simply increasing the wall thickness of the thermowells will for flow rates above 1,000 lb/hr. extend their operational life and thus reactor uptime. Finding 2-3. The 400-hr liner change-out interval recom- Elemental Sulfur Additive mended for the VX nerve agent campaign is not supported by the worst-case corrosion data in the draft FOAK preliminary Finding 2-8. The elemental sulfur additive did not have any test report. adverse impact on FOAK testing and successfully resolved the issues it was intended to. Recommendation 2-1. The Blue Grass Chemical Agent De- struction Pilot Plant staff should shorten the liner change-out Recommendation 2-4. Elemental sulfur additive should be period for VX processing from the 400 hr recommended in used instead of sulfuric acid additive. the test report until the corrosion rates under actual operating conditions are verified. Two hundred hours would be a better Variation in Feed Composition initial change-out period. Finding 2-9. The limited feed variations tested during FOAK Finding 2-4. Corrosion thinning of the thermowells (ac- testing did not result in any negative impacts to the SCWO companied by mechanical failure when minimal material process. Variations other than those tested during FOAK remains) is a critical failure path for the reactors and is ex- testing could be a concern and are addressed in Chapter 3. pected to be the most frequent maintenance issue associated with continuous operations. Functionality During FOAK Testing Recommendation 2-2. The diameter of the thermowells High-Pressure Air Compressor should be increased to at least 0.75 in. from the current diameter of 0.50 in. by increasing the wall thickness. The Finding 2-10. There was a much higher failure rate for the Blue Grass Chemical Agent Destruction Pilot Plant project high-pressure air compressor than would be expected from a staff should use existing corrosion data to estimate the life- mature system. These failures repeatedly interrupted FOAK time of these new thermowells. This operation is a simple testing. The multiple issues with the air compressor, which 

4 ASSESSMENT OF SUPERCRITICAL WATER OXIDATION SYSTEM TESTING FOR THE BGCAPP is a single, critical component of the SCWO system, are of ization Implementation Plan and had extensive discussions concern to the committee. with the BGCAPP project staff. The following findings and recommendations resulted from this work. Recommendation 2-5. The Blue Grass Chemical Agent Destruction Pilot Plant project staff should review the main- SCWO Feed Composition tenance and start-up plan for the air compressors so that they are maintained to industry standards. The project staff should Finding 3-1. Compositional variations within the energet- use the time between compressor delivery and systemization ics and agent feedstocks are dampened by the blending of to run the compressors, to discover any problems similar to multiple batches and post-hydrolysis neutralization to a those which occurred during FOAK testing, with the goal of common pH range. It is unclear, however, how the SCWO obtaining reliability comparable to that of the supercritical system would respond if it received a feed with a composi- water oxidation reactors. tion outside the range tested during FOAK testing. Finding 2-11. The oil-water separator appears to present the Recommendation 3-1. The range of acceptable feed com- risk of a critical single-point failure that could shut the entire positions for the SCWO unit, which must be achieved in the SCWO process down. hydrolysate blend tank, should be clearly specified. These compositional parameters should be based on the range of Recommendation 2-6. In addition to the action recom- compositions actually used and verified during testing of the mended in Recommendation 2-5, the Blue Grass Chemical SCWO unit. Agent Destruction Pilot Plant project staff should consider mitigation strategies such as having an additional oil-water Finding 3-2. There is uncertainty as to whether the agita- separator (OWS) in parallel that could run should the main tion and heating methodologies used during FOAK testing OWS require maintenance. Additionally, a spare OWS could accurately represent the conditions that will be present in the be kept on hand. full-scale hydrolysate storage tanks. Recommendation 3-2. The full-scale agitation/circulation High-Pressure Quench Water system at the Blue Grass Chemical Agent Destruction Pilot Finding 2-12. The processing of the modified blended H Plant should be evaluated upon systemization. Care should hydrolysate simulant recipe at 1275°F proceeded smoothly be taken to avoid conditions where sulfur can melt on the with little corrosion and complete oxidation. heater surfaces, and vigorous agitation should be applied near the bottom of the tank. The blended sarin (GB) hydro- Recommendation 2-7. The Blue Grass Chemical Agent lysate should be sampled to confirm that the agitation and Destruction Pilot Plant supercritical water oxidation system heating conditions are sufficient to maintain uniform feed test plan should ensure that the recipe for the blended H composition. hydrolysate is the same as the recipe used in the FOAK tests. Safety High-Pressure Hydrolysate Feed Maintenance Finding 2-13. Particulates in the blended hydrolysate simu- lant, which was not filtered, caused the check valves to stick Finding 3-3. There is uncertainty regarding which parts of during testing. the SCWO system may have to shut down during parts (e.g., liner or thermowell) replacement. There are also concerns about worker safety since at least two SCWO units will be Maintainability running simultaneously and they are located close to each Finding 2-14. The overall first-of-a-kind testing objectives other in the process building. for the supercritical water oxidation system to be used at the Blue Grass Chemical Agent Destruction Pilot Plant were Finding 3-4. Current plans permit maintenance personnel met. The processes and subsystems, inclusive of mainte- to be in the SCWO bay while adjacent SCWO units are nance activities, performed individually and collectively operating. The committee is concerned that existing safety to meet test objectives and exceed the target of 76 percent planning and procedures may be inadequate for this situation. availability. If current plans are implemented, the committee believes ad- ditional worker protection would be warranted. Chapter 3 of the report considers systemization of the SCWO reactors at BGCAPP. In addition to the preliminary Recommendation 3-3. The safety protocols used during FOAK test report, the committee also reviewed the System- first-of-a-kind testing need to be adapted and applied to 

SUMMARY 5 operation and maintenance of the three parallel units at the system operations to the extent possible. A manual should be Blue Grass Chemical Agent Destruction Pilot Plant. This developed for SCWO system operations and used routinely includes not conducting maintenance adjacent to operating as part of training and maintaining the readiness of the plant supercritical water oxidation reactors. personnel and the process equipment and systems throughout systemization. Such a manual is described more fully in a Recommendation 3-4. If the safety protocols used during later section in Chapter 3. first-of-a-kind testing are not adopted for SCWO operations and maintenance, the Blue Grass Chemical Agent Destruc- Effects of Aging and Storage on Component Operability tion Pilot Plant (BGCAPP) project staff should conduct detailed operational risk and safety analyses before allowing Finding 3-7. Systemization will use equipment that is to be personnel to conduct maintenance activities in close proxim- acquired years before its operation. ity to operating SCWO reactors. BGCAPP staff should also investigate additional worker protection, such as supplemen- Recommendation 3-8. A detailed maintenance schedule tal barriers between workers and operating SCWO reactors, should be created that leaves ample time to test and evaluate which could also capture debris. all subsystems of the SCWO system. The schedule should include ample training time for operators to familiarize themselves with the equipment. Process Exhaust Finding 3-5. The discharge of process gas effluents into the Finding 3-8. SCWO system process equipment (e.g., reactor general atmosphere of a building is not a good design prac- and peripherals, compressor, and pumps) will be shipped to tice. Such discharge can pose any number of safety risks to the Blue Grass Chemical Agent Destruction Pilot Plant and personnel depending on the composition and volume of the remain in storage for several years prior to start-up of the gas effluent. SCWO process for systemization and eventually for op- erations. Additional extended periods of idleness are likely. Recommendation 3-5. The Blue Grass Chemical Agent De- The committee has concerns about whether the equipment, struction Pilot Plant project staff should evaluate the SCWO either as a working system or as individual components, will process building design to ensure that the quality of the air perform as designed and expected after sitting idle for such inside the emergency relief tank room is safe for personnel. a long time. Remedial action could include the discharge of SCWO gas effluent through the appropriate offgas treatment system Recommendation 3-9. In addition to planned systemiza- rather than to the atmosphere inside the building, air quality tion activities, plans should be made for regularly testing monitoring inside the emergency relief room with appropri- or operating the various SCWO and water recovery system ate alarm settings, or other approaches that would ensure the components. These plans should be included in the interim room and building air quality is safe for personnel. SCWO operations manual for the interim period between delivery and operations, described in the next section. Recommendation 3-6. If process exhaust is to be not vented through the appropriate offgas treatment system, an SCWO Operations Manual for Interim Period operational risk and safety evaluation should be conducted on the venting of process exhaust into the SCWO process Finding 3-9. A manual is needed that addresses the concerns building. discussed above for the periods between equipment delivery and systemization, during systemization, and between sys- temization and operations, and including integrated SCWO Personnel Issues and Knowledge Transfer/Retention and water recovery system operation. Creating the manual in Finding 3-6. The committee is concerned about how the electronic form would allow for maximum visibility across know-how, knowledge, and experience obtained by the sys- a wide range of information in a searchable framework on tem contractor SCWO system operators and engineers, and portable platforms such as tablet computers. by the Blue Grass Chemical Agent Destruction Pilot Plant staff who shadowed the system contractor operators, will be Recommendation 3-10. The Blue Grass Chemical Agent preserved during the multiyear hiatus between equipment Destruction Pilot Plant project staff should prepare a manual delivery and systemization. specifically to be used in the interim period preceding SCWO systemization, throughout the SCWO systemization time Recommendation 3-7. Plans should be made for periodi- period, and during the time following SCWO systemization, cally providing refresher training opportunities for SCWO prior to plant operation. This manual should also address plant operators and personnel during downtime and sys- integrated SCWO and water recovery system maintenance temization activities. This training should include actual and operation during these periods and should address all the 

6 ASSESSMENT OF SUPERCRITICAL WATER OXIDATION SYSTEM TESTING FOR THE BGCAPP concerns discussed above. The manual should be prepared in a simulation-optimization study where external experts in electronic form. can help develop a computerized simulation-optimization model that describes the operations, over time, for the entire combined supercritical water oxidation and water recovery Cyanide system. Finding 3-10. Cyanide will be present in the energetics and blended hydrolysate streams. This must be managed to Chapter 4 discusses systemization of the WRS and protect worker safety. The testing done to date shows that integrated SCWO and WRS systems. It does so bearing the SCWO process will destroy any HCN or CN− present in in mind the 2012 NRC report The Blue Grass Chemical the liquid hydrolysate feed to undetectable levels. A formal Agent Destruction Pilot Plant’s Water Recovery System. risk analysis would be needed to quantify the actual risks The chapter also takes into account the results of the FOAK to workers. t ­esting of the SCWO system and how possible changes in the makeup of the SCWO effluent from what was consid- Recommendation 3-11. If cyanide cannot be removed from ered in the 2012 report might affect systemization of the the hydrolysate, affected personnel should be informed that WRS. if exposure to hydrolysate occurs in an off-normal situation, the hydrolysate could have contained up to approximately BGCAPP Actions in Response to the 2012 NRC Report 7 ppm, or 300 µM, cyanide. Information about the possible presence of cyanide should also be included in emergency Pretreatment response manuals or other documentation so that first re- sponders are aware of the potential risk. Workers should be Finding 4-1. Although stable, fouling-free performance can- trained in how to recognize exposure to cyanide and what not be guaranteed ahead of time, the Blue Grass Chemical should be done while awaiting first responders. Agent Destruction Pilot Plant project staff appear to have taken appropriate engineering design measures to safeguard Recommendation 3-12. The Blue Grass Chemical Agent the reverse osmosis membranes from excessive fouling Destruction Pilot Plant project management should quantify conditions. the amount of HCN gas that vents from the aluminum filtra- tion system and install an appropriate mitigation to avoid Reverse Osmosis Membranes worker/public exposure if the concentrations are above the local air quality release limits. If cyanide cannot be removed Finding 4-2. The WRS committee’s concerns about the from the hydrolysate, project management should conduct a storage of the RO membranes between delivery and sys- risk and safety analysis to identify the risks and mitigation temization and their chemical cleaning appear to have been strategies. adequately addressed. Finding 3-11. Which methodology will be used to manage Materials of Construction the cyanide problem has not yet been determined. Whatever methodology is used will change the chemistry of the SCWO Finding 4-3. The materials of construction planned for use system feed and could impact the performance of the system. in the WRS now appear to have been adequately tested in representative environments. It is now possible to have con- Recommendation 3-13. The Blue Grass Chemical Agent fidence that the planned materials will perform adequately Destruction Pilot Plant project staff should identify an ef- for the amount of time that the water recovery system will fective cyanide management system that does not negatively operate and in the anticipated environments. impact the performance of the SCWO unit. Systemization of the WRS Overall System Operations and Computer Model Reverse Osmosis Membrane Storage after Systemization Finding 3-12. There is no computer simulation model that describes the integrated system operations, over time, of all Finding 4-4. There is the potential for membrane degrada- SCWO system processes or of the entire SCWO/WRS. A tion between systemization and full plant operation unless model tailored for the SCWO and the integrated SCWO/ care is taken to preserve the membrane. WRS systems would permit system performance to be optimized. Recommendation 4-1. Blue Grass Chemical Agent Destruc- tion Pilot Plant operators should obtain, review, and follow Recommendation 3-14. The Blue Grass Chemical Agent manufacturer recommendations for membrane preservation Destruction Pilot Plant project staff should consider investing between systemization and start-up of plant operations. 

SUMMARY 7 Development of Water Balance may have a considerably higher level of TOC than envisioned in previous reports, and this could lead to more rapid and se- Finding 4-5. The operation of the WRS has not been opti- vere fouling of the reverse osmosis membrane than originally mized for use with the SCWO system. envisioned in the WRS design. The main concern is the need for more frequent and extensive cleaning (ideally between Recommendation 4-2. Blue Grass Chemical Agent Destruc- campaigns), which could lead to rapid membrane degrada- tion Pilot Plant staff should optimize permeate generation to tion and call for premature replacement. match supercritical water oxidation system quench demand.   Recommendation 4-4. Blue Grass Chemical Agent Destruc- Impact of Water Recycling on the SCWO System tion Pilot Plant operators should monitor the reverse osmo- sis membranes and system during systemization testing to Finding 4-6. The committee foresees no problems with us- determine if there is any evidence of premature membrane ing the water recovered via reverse osmosis in the SCWO degradation. If this is observed, then the membranes would system other than slight changes that may need to be made need to be replaced sooner. This is a minor operational issue to the monitoring parameters. that can be handled by keeping a spare set of membranes on-site at all times. Recommendation 4-3. The Blue Grass Chemical Agent Destruction Pilot Plant project staff should characterize the reverse osmosis effluent for each agent campaign. They REFERENCE should then reestablish normal operating conductivities NRC (National Research Council). 2012. The Blue Grass Chemical Agent when reverse osmosis permeate is injected (recycled) at the Destruction Pilot Plant’s Water Recovery System (Letter Report). Wash- bottom of the SCWO reactor. ington, D.C.: The National Academies Press. Integrated SCWO and WRS Operations Finding 4-7. Since the final effluent total organic carbon (TOC) level will be based on permit levels, the WRS feed