Click for next page ( 25


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 24
3 Implementation of Supercritical Water Oxidation at Blue Grass The supercritical water oxidation (SCWO) process 7.  ptimization—integrated plant operations, as if ac- O at the Blue Grass Chemical Agent Destruction Pilot Plant tual chemical agent were being processed, conducted (BGCAPP) will process the agent and energetics hydroly- before start of agent operations. sate streams resulting from agent and munition destruction. Three SCWO units have been designed and constructed for At this writing, the systemization process has not yet use at BGCAPP. One of these was used for first-of-a-kind begun. Also, at this writing, precommissioning activities are (FOAK) testing. All of the units will be shipped to BGCAPP scheduled to begin in March 2015, with the system demon- site sometime in mid-2013 but are not scheduled to begin stration to conclude June 2020.2,3 processing stimulant hydrolysate until at least 2017 (BPBGT, Procedures are in place to hand off elements of the over- 2012).1 all BGCAPP system from construction to commissioning/ The SCWO systemization process is intended to dem- start-up. This process is organized by dividing portions of the onstrate that the plant, procedures, and personnel are ready overall BGCAPP system into quality data packages (QDPs), for operating the SCWO system. Systemization is planned and the BGCAPP project has a schedule and process for this to proceed through seven sequential subphases. These are hand-off for all of the different QDPs. Some of these QDPs (BPBGT, 2012) the following: are scheduled for turnover in 2014, others not until 2016. The final optimization stage includes 20 weeks for train- 1. Presystemization—planning and document preparation; ing and preparing the plant workforce. The same personnel 2.  onstruction support—review of the actual plant C involved in systemization will be used in plant operations against plant design documents; to take advantage of the accumulated system knowledge 3.  recommissioning—input/output testing for major P and experience. After all the subsystems have been system- equipment (e.g., electrical and mechanical systems ized into the overall BGCAPP system, a final operational checks for each piece of equipment); readiness review will be performed to verify that the people, 4.  ommissioning—instrument calibration, logic con- C paperwork, and plant equipment are ready to support agent troller interface verification, alarm operation verifica- operations (BPBGT, 2012). tion, etc.; Figure 3-1 shows a SCWO block flow diagram. The 5.  ystem start-up—attention shifts to the system level S fuel, isopropanol, and hydrolysate flows are continuous and and integration of the subsystem components, inter- operate until the unit is shut down. A complete operating lock checks, standard operating procedure validation schedule is still under development. and revision as necessary, and safety checks; The hydrolysate storage tanks hold the output streams 6.  ystem demonstration—functional demonstration S from the neutralization of agent and energetics. These tanks of system capabilities including start-up, normal op- can hold about one third of the total output expected from eration in automatic and manual modes, processing the neutralization reactors. This total output is roughly 50 simulant where applicable; and 2Ron Hawley, plant general manager, URS Corporation, “Systemization Overview,” presentation to the committee on February 4, 2013. 1The BGCAPP schedule has undergone continuous modification, includ- 3Joe Novad, deputy program executive officer, Assembled Chemical ing during the course of this study. The committee has given the best dates Weapons Alternatives (ACWA), “Program Executive Office Assembled available to it at the time of this report. The findings and recommendations Chemical Weapons Alternatives Update,” presentation to the Committee on in this report do not depend on the exact dates quoted. Chemical Demilitarization on April 9, 2013. 24

OCR for page 24
IMPLEMENTATION OF SUPERCRITICAL WATER OXIDATION AT BLUE GRASS 25 Off-gas to filter Off-gas duct heater Flush water HP air HP air Gas heater compressors manifolds analyzers Hydrolysate HP hydrolysate heater Wet feed pump addiƟves module (H 2 SO 4 , HCl , NaOH) REACTOR MODULE FEED MODULE Solid feed addiƟves (Sulfur, NaCl) HP gas/ Blend SCWO liquid HP water pump reactor separator Agent tanks Preheater hydrolysate HP fuel pump EnergeƟcs hydrolysate HP quench pump LP gas/ liquid separator Batch holding tank Effluent heat exchanger TOC analyzers Off-spec tank Emergency Liquid effluent to water relief tank recovery system FIGURE 3-1  SCWO block flow diagram showing only the SCWO system. SOURCE: Program Executive Officer, ACWA, Presentation to the Supercritical Water Oxidation Technical Review on April 22, 2013. Figure 3-1 million lb.4 If each of the three SCWO units processes and the neutralization of hydrolysate to a specific pH will ho- 760 lb/hr of material (1,000 lb/hr nominal rate, in operation mogenize any minor variations in the SCWO feed that might at the defined 76 percent availability) on a continuous basis have arisen as batch-to-batch differences. Also as discussed (365 days/yr, 24 hr/day), then 2.5 years would be required in Chapter 2, the FOAK testing showed satisfactory SCWO for SCWO treatment of all of the hydrolysate. If only two unit performance when several parameters (elemental sulfur, SCWO units operate at one time, then 3.8 years would be HCl, hydrolysate flow rate) were varied individually within required to treat all of the hydrolysate. the small ranges anticipated during operation. Although these tests provided useful information, vary- ing each parameter individually may not have been sufficient SCWO FEED COMPOSITION to reveal the limitations of the SCWO system. Variations in The effectiveness of the SCWO system depends on the more than one feed component at a time might have resulted composition and flow rate of the feed stream, which makes in a more challenging environment. System contractor per- a thorough understanding of the feed composition and its sonnel reported that they could not afford to test all possible potential variability essential for ensuring that the SCWO combinations, and the committee concurs. Computer simula- system attains the destruction efficiencies required by permit. tion, however, might allow estimating the combined effects To date, however, there are no data on the potential variability of these variables based on their individual effect (see the in the SCWO feed composition coming from the blend tank. final section of this chapter, “Overall System Operations and There is also no indication of what constitutes an off-spec Computer Model,” for a more detailed discussion). feed. As is discussed in Chapter 2, however, the mixing of As noted in Chapter 2, there were problems of sulfur multiple batches of both energetics and agent hydrolysate melting and agglomeration with the original heaters used during FOAK testing. These were successfully addressed 4John Barton, chief scientist, Bechtel Parsons Blue Grass Team (BPBGT), by using a lower wattage electric heater (BPBGT, 2013). However, steam heat will be used at BGCAPP, and the “BGCAPP Process Overview,” presentation to the committee on January 7, 2013. committee is unsure how this will affect the potential for

OCR for page 24
26 ASSESSMENT OF SUPERCRITICAL WATER OXIDATION SYSTEM TESTING FOR THE BGCAPP sulfur melting and agglomeration. Also, the elemental sulfur Maintenance additive will likely be less soluble in the hydrolysate than According to committee discussions with BGCAPP in water (Boulegue, 1978). The results of FOAK testing staff, the current operating plan is to have two SCWO units confirmed that the issue of sulfur insolubility was adequately operating at a time, with the third as a spare. It is, however, addressed.5 From a systemization standpoint, because of conceivable that all three units could be operating at once differences between the FOAK and operational settings, BG- if unforeseen downtime requires makeup SCWO operation. CAPP staff need to check that the sulfur will be sufficiently The system contractor developed safety protocols for the agitated so that it behaves chemically in the same manner as operation of SCWO units during FOAK testing. While there a solution with sulfuric acid. were exceptions during FOAK testing, the general operating philosophy was that personnel were not supposed to be in the Finding 3-1. Compositional variations within the energet- SCWO bay while the SCWO unit was operating. There has, ics and agent feedstocks are dampened by the blending of however, been only limited discussion of the safety issues multiple batches and post-hydrolysis neutralization to a associated with operating three SCWO units in parallel at common pH range. It is unclear, however, how the SCWO BGCAPP during systemization, conducting system main- system would respond if it received a feed with a composi- tenance between systemization and plant operations, and tion outside the range tested during FOAK testing. conducting maintenance while SCWO units are operating. The committee is specifically concerned about plans to Recommendation 3-1. The range of acceptable feed com- perform maintenance on one SCWO unit while the adjacent positions for the SCWO unit, which must be achieved in the unit is operating. The unexpected release of steam due to hydrolysate blend tank, should be clearly specified. These failure of an operating unit, for example, might pose a safety compositional parameters should be based on the range of hazard to a worker doing maintenance on an adjacent unit. compositions actually used and verified during testing of the Based on information provided to the committee, such main- SCWO unit. tenance activity would be inconsistent with the safety prac- tices used during FOAK testing. Nevertheless, the presence Finding 3-2. There is uncertainty as to whether the agita- of plant personnel performing maintenance in the SCWO tion and heating methodologies used during FOAK testing bays during SCWO operations is anticipated to be normal accurately represent the conditions that will be present in the practice at BGCAPP. The committee does not believe that full-scale hydrolysate storage tanks. ensuring the safety of these workers is adequately considered in the existing safety planning. Recommendation 3-2. The full-scale agitation/circulation The system contractor believes that the design of the system at the Blue Grass Chemical Agent Destruction Pilot SCWO units makes performing maintenance while an Plant should be evaluated upon systemization. Care should adjacent SCWO unit is operating safe. The contractor that be taken to avoid conditions where sulfur can melt on the will operate the SCWO units at BGCAPP has stated that no heater surfaces, and vigorous agitation should be applied significant safety analyses have been conducted to date, but near the bottom of the tank. The blended sarin (GB) hydro- are planned.6 There are Lexan panels mounted on the SCWO lysate should be sampled to confirm that the agitation and process equipment skids. The system contractor states that heating conditions are sufficient to maintain uniform feed these panels will be capable of containing emissions from composition. minor leaks and even a major pipe failure, protecting person- nel. There are also blow-off panels on the tops of the equip- SAFETY ment skids to direct the force from any catastrophic failures upwards, away from personnel (though there are questions In the course of its work, the committee encountered two about the debris environment and how dangerous that might things that raised safety concerns. The first involves plans to be to personnel in the area).7 Even so, the system contractor have personnel in the SCWO bay performing maintenance did state that personnel access to the SCWO bay should be while adjacent SCWO systems are operating. The second restricted during operations, and that there should not be any involves the venting of SCWO process exhaust into the long-term presence of personnel, such as foot traffic between SCWO process building. 6Personal communication between Steven Mantooth, Blue Grass Chemi- cal Agent Destruction Pilot Plant program office, and James Myska, NRC study director, on February 7, 2013. 5Kevin Downey, advanced process system manager, General Atomics, 7Dan Jensen, advanced process system program manager, General Atom- “SCWO FOAK Test Status,” presentation to the committee on February ics, “SCWO System Testing and Lessons Learned,” presentation to the 4, 2013. Committee on Chemical Demilitarization on April 10, 2013.

OCR for page 24
IMPLEMENTATION OF SUPERCRITICAL WATER OXIDATION AT BLUE GRASS 27 SCWO units.8 This still leaves the committee concerned agent disposal facilities and identified primary causal fac- about personnel performing maintenance activities adjacent tors that led to the incidents. It then went on to discuss the to an operating SCWO unit. identification and use of process safety metrics as leading If BGCAPP does carry through on plans to have work- indicators to prevent safety incidents. The committee also ers perform maintenance activities in close proximity to wrote about process safety metrics that are used in the chemi- operating SCWO reactors, supplemental safety steps could cal industry. The present committee believes the 2011 report improve worker safety. Additional barriers could be placed would be a useful resource in addressing safety concerns between the workers and an operating SCWO reactor. Such raised by the SCWO system. barriers could also capture debris. Barrier materials and their placement must be tailored to the characteristics of the Process Exhaust debris threat. During FOAK testing, gas effluents from the single Finding 3-3. There is uncertainty regarding which parts of SCWO train operated were vented to a holding tank the SCWO system may have to shut down during parts (e.g., outside the process building. Process design descriptions liner or thermowell) replacement. There are also concerns outlined in the FOAK testing report indicate that for the about worker safety since at least two SCWO units will be BGCAPP facility the gas effluents from the SCWO units running simultaneously and they are located close to each will be discharged into the emergency relief tank room. other in the process building. After release to the airspace in that room, the gas effluents from the SCWO units are to be drawn through the carbon Finding 3-4. Current plans permit maintenance personnel filters of the SCWO process building ventilation system to be in the SCWO bay while adjacent SCWO units are for eventual release to the atmosphere (BPBGT, 2013). It operating. The committee is concerned that existing safety thus appears to the committee that SCWO process gas ef- planning and procedures may be inadequate for this situation. fluents are to be released into the general SCWO processing If current plans are implemented, the committee believes ad- building atmosphere via the emergency relief tank room, ditional worker protection would be warranted. where they could impact any personnel in the area. The committee is concerned that, depending on the volume of Recommendation 3-3. The safety protocols used during SCWO gas effluent, the general SCWO processing building first-of-a-kind testing need to be adapted and applied to atmosphere might be either high in carbon monoxide and/ operation and maintenance of the three parallel units at the or low in oxygen. Blue Grass Chemical Agent Destruction Pilot Plant. This includes not conducting maintenance adjacent to operating Finding 3-5. The discharge of process gas effluents into the supercritical water oxidation reactors. general atmosphere of a building is not a good design prac- tice. Such discharge can pose any number of safety risks to Recommendation 3-4. If the safety protocols used during personnel depending on the composition and volume of the first-of-a-kind testing are not adopted for SCWO operations gas effluent. and maintenance, the Blue Grass Chemical Agent Destruc- tion Pilot Plant (BGCAPP) project staff should conduct Recommendation 3-5. The Blue Grass Chemical Agent detailed operational risk and safety analyses before allowing Destruction Pilot Plant project staff should evaluate the personnel to conduct maintenance activities in close proxim- SCWO process building design to ensure that the quality ity to operating SCWO reactors. BGCAPP staff should also of the air inside the emergency relief tank room is safe for investigate additional worker protection, such as supplemen- personnel. Remedial action could include the discharge tal barriers between workers and operating SCWO reactors, of SCWO gas effluent through the appropriate offgas which could also capture debris. treatment system rather than to the atmosphere inside the building, air quality monitoring inside the emergency relief While the committee was not in a position to address room with appropriate alarm settings, or other approaches safety in any real depth, it does want to call the reader’s that would ensure the room and building air quality is safe attention to the NRC report Assessment of Approaches for for personnel. Using Process Safety Metrics at the Blue Grass and Pueblo Chemical Agent Destruction Pilot Plants (NRC, 2011). The Recommendation 3-6. If process exhaust is to be not committee that wrote the 2011 report conducted a review of vented through the appropriate offgas treatment system, an process safety incidents that had occurred at other chemical operational risk and safety evaluation should be conducted on the venting of process exhaust into the SCWO process building. 8Kevin Downey, advanced process system manager, General Atomics, “SCWO System Testing and Lessons Learned,” presentation to the Com- mittee on Chemical Demilitarization on April 10, 2013.

OCR for page 24
28 ASSESSMENT OF SUPERCRITICAL WATER OXIDATION SYSTEM TESTING FOR THE BGCAPP PERSONNEL ISSUES AND KNOWLEDGE TRANSFER from the training contractor have never been to the SCWO AND RETENTION system contractor site to observe the SCWO system much less to observe actual operation.12 This would have been an This section discusses the training of the BCGAPP staff opportunity to attempt to capture operational experience for who will operate and maintain the SCWO system, the trans- the purposes of training the BGCAPP staff. fer of knowledge and expertise from the system contractor It should be noted that SCWO technology in general SCWO operators to the BGCAPP SCWO operators, and the is fairly mature and the system contractor supplying the retention of that knowledge at BGCAPP during the several SCWO to BGCAPP has been in the business of selling sys- years between SCWO unit delivery to BGCAPP and SCWO tems for some time. The SCWO units designed and built for operation to treat hydrolysate. While BGCAPP personnel this application are based on a significant amount of prior shadowed the system contractor SCWO operators during experience and testing, and debugging efforts have been FOAK testing, the BGCAPP personnel were not permitted conducted on the actual units that will be installed and op- to conduct any hands-on work.9 Expertise is built by actu- erated at BGCAPP. Additionally, the workforce at the plant ally doing a job. The system contractor operators will not is composed in part of personnel who have worked at other move to BGCAPP to operate the SCWO units there. Thus, chemical demilitarization facilities. While the technologies there is a risk that the accumulated expertise of the system they managed were different, they bring with them a great contractor operators will be lost. Given the long duration of deal of experience in the disposal of chemical munitions systemization and number of years that will pass from when and a strong safety culture. The combination of these factors the SCWO units are delivered to BGCAPP and when they provides a path that should allow the operators to be trained begin to process hydrolysate, there is a very real chance that appropriately and the units to be operated safely. The com- knowledge, know-how, and key process metrics could be mittee’s concerns are not that it will be difficult to train the lost. A SCWO process operation manual designed specifi- operators but that much of the operational expertise, the art, cally for use during the interim between SCWO unit delivery of running the SCWO system smoothly will be lost in the to BGCAPP and SCWO system operation with hydrolysate transition from system contractor staff to BGCAPP staff. could be a valuable way to ensure preservation of knowledge. Training is another area of concern. It will be criti- Finding 3-6. The committee is concerned about how the cal to train operators who will not have had experience in know-how, knowledge, and experience obtained by the sys- running SCWO units. The committee visited the FOAK tem contractor SCWO system operators and engineers, and test site during its first meeting and observed the control by the Blue Grass Chemical Agent Destruction Pilot Plant systems and spoke with the operators. The discussions and staff who shadowed the system contractor operators, will be observations indicated to the committee that training people preserved during the multiyear hiatus between equipment in the basic operation of the SCWO units will not be par- delivery and systemization. ticularly challenging.10 A detailed high-level framework and approach has been proposed to train SCWO operators and, Recommendation 3-7. Plans should be made for periodi- based on the accumulated chemical industrial expertise of cally providing refresher training opportunities for SCWO the committee, seems to be well thought out. The details of plant operators and personnel during downtime and sys- implementing this high-level plan have not yet been worked temization activities. This training should include actual out. The approach and framework include knowledge train- system operations to the extent possible. A manual should be ing (classroom instruction) and skills training (training lab, developed for SCWO system operations and used routinely process control simulator, and on-the-job-training) along as part of training and maintaining the readiness of the plant with demonstration of knowledge retention (written and personnel and the process equipment and systems throughout oral exams) and demonstration of skills (measuring job systemization. Such a manual is described more fully in a performance). Operators will be trained on all aspects of the later section in Chapter 3. SCWO system, including start-up, normal operation, manual system operation, response to alarms, shutdown, and emer- gency shutdown. Similarly, SCWO process maintenance EFFECTS OF AGING AND STORAGE ON COMPONENT workers will be trained, including hands-on learning using a OPERABILITY mockup SCWO vessel. 11 Another contractor, different from The committee’s biggest maintenance concern is the the SCWO system contractor, is responsible for training the long downtime between equipment delivery and installation BGCAPP staff. The committee was told that staff members and the eventual operation of the system in a (semi)continu- ous mode. Preliminary schedules shared with the commit- 9Discussions between the committee and BGCAPP and General Atomics tee showed potential years-long delays between delivery of staff during the first meeting on January 7-9, 2013. 10Ibid. 11Doug Plummer, BPBGT training manager, GP Strategies Corporation, 12Discussions between the committee and BGCAPP and General Atomics “SCWO Training Plan,” presentation to the committee on February 4, 2013. staff during the first meeting, January 7-9, 2013.

OCR for page 24
IMPLEMENTATION OF SUPERCRITICAL WATER OXIDATION AT BLUE GRASS 29 equipment and installation and operation of that equipment. erations. Additional extended periods of idleness are likely. For simple equipment, like a stainless steel tank, the delay The committee has concerns about whether the equipment, is not a concern. For complex mechanical equipment with either as a working system or as individual components, will electronics, such as an air compressor, these delays are a perform as designed and expected after sitting idle for such concern. Electronics can become damaged, components can a long time. corrode, and other age-related failures can occur. The com- mittee believes there is significant schedule risk if complex Recommendation 3-9. In addition to planned systemiza- equipment sits for years exposed to changes in temperature tion activities, plans should be made for regularly testing and humidity before installation and operation. Both the op- or operating the various SCWO and water recovery system eration of long-idle equipment and knowledge preservation components. These plans should be included in the interim of the operators are concerns. SCWO operations manual for the interim period between delivery and operations, described in the next section. Finding 3-7. Systemization will use equipment that is to be acquired years before its operation. SCWO OPERATIONS MANUAL FOR INTERIM PERIOD Recommendation 3-8. A detailed maintenance schedule Operations procedures and operation manuals have been should be created that leaves ample time to test and evaluate provided by the system contractor for the SCWO system all subsystems of the SCWO system. The schedule should when it is in full operation. However, to the committee’s best include ample training time for operators to familiarize knowledge, such procedures have not been codified for sys- themselves with the equipment. tem monitoring and maintenance during the period between equipment delivery and systemization, during systemization, The most critical issue in the SCWO systemization plan and between systemization and full plant operation. Also, is the 6 yr or more delay between receipt of the SCWO units to the committee’s best knowledge, there are no integrated at BGCAPP and when the units will be expected to handle operating procedures for the combined SCWO and water the hydrolysate under operational conditions. This delay is a recovery system (WRS). serious issue and, to the best of the committee’s knowledge, In particular, for the interim period between system- one never before encountered in the start-up and operation ization and operation, the committee has concerns about of any commercial chemical processing unit. A detailed personnel training, safety, equipment resilience and reli- procedure has been proposed to deal with both personnel and ability, prevention of corrosion and other system degrada- equipment issues associated with this delay. This procedure, tion possibilities such as that of the reverse osmosis (RO) which is part of the systemization plan, includes testing the membrane (see Chapter 4), and compressor maintenance (see mechanical and electrical operability of each subsystem, Chapter 2). These concerns include the need for continual testing and calibrating all instruments, validating standard personnel training; cycling and rinsing the system to prevent operating procedures, demonstrating that individual system corrosion, the potential for solutions freezing in system components operate as an interdependent whole, and operat- lines, and precipitation in the lines; maintenance of the high- ing the entire plant with simulated hydrolysate to ensure that pressure compressors; and treating the RO membranes to everything is in working order.13 Because this delay is un- prevent degradation during this period. Many of these chal- precedented, at least in the experience of the committee, the lenges may differ from what the operations manual provided committee cannot judge the procedure’s likely effectiveness. by the system contractor says about full plant operation of It was also unclear to the committee whether the equipment- the SCWO. related activities associated with systemization (i.e., those The methods for individually handling each of these items outlined above) would be done just once or whether factors have been discussed and in most cases have been sat- they would be performed periodically. If they are done just isfactorily resolved. However, no comprehensive schedule or once, then there remains the possibility that process equip- manual has been prepared as a guide to implementing these ment will sit idle, conceivably for years, prior to, during, and procedures for the interim period between systemization after the systemization phase. testing and plant operation. A manual for the periods between equipment delivery Finding 3-8. SCWO system process equipment (e.g., reactor and systemization, during systemization, between systemiza- and peripherals, compressor, and pumps) will be shipped to tion and operations, and for the integrated SCWO and WRS the Blue Grass Chemical Agent Destruction Pilot Plant and could forestall extended downtimes. Such a manual would remain in storage for several years prior to start-up of the include a defined schedule for performing all of the identi- SCWO process for systemization and eventually for op- fied interim training and maintenance tasks on the SCWO and WRS. It would include standard operating procedures 13Ron Hawley, plant general manager, URS Corporation, “Systemization for the various system components, safety procedures to be Overview,” presentation to the committee on February 4, 2013. used during operation and during maintenance, maintenance

OCR for page 24
30 ASSESSMENT OF SUPERCRITICAL WATER OXIDATION SYSTEM TESTING FOR THE BGCAPP intervals, steps for corrosion prevention, steps for maintain- tanks during operations. A formal risk analysis would be ing RO membrane integrity, diagnostics and monitoring of needed to quantify the extent of the actual risks. potential system degradation, equipment testing intervals, A recent white paper (BPBGT, 2011) analyzes the and protocols to be used throughout the time period prior to results of earlier hydrolysis studies and calculates the ex- processing hydrolysate. It could include process metrics such pected concentrations of HCN/CN− resulting from the hy- as the acceptable range of SCWO feed concentrations for key drolysis of energetics in the Blue Grass stockpile. Cyanide components and the target performance values for the SCWO concentrations in the aqueous phase were measured for the units (e.g., total organic carbon and total hydrocarbons in the hydrolysates of all the various types of energetics and all had SCWO effluent). Creating the manual in electronic format positive detects. The hydrolysates showed concentrations would make all this information available to operators in a ranging from 1.7 to 705 mg/L (ppm). The level regarded searchable framework. It could also be made accessible on as safe is <0.2 mg/L. Gaseous hydrogen cyanide (HCNg) tablet computers, enabling portability and increasing utility. concentrations were measured above the hydrolysis solu- Electronic manuals are the industry standard. In addition tion with values ranging from nondetect to 0.08 mg/m3. The to assisting in the maintenance of equipment operability, measured HCNg concentrations in the off-gas were all well the manual could address the challenges of knowledge and below established exposure limits (5 mg/m3).14 experience transfer and retention during the time preceding Aluminum will also be present in the hydrolysate since SCWO systemization, throughout the SCWO systemization the rocket warhead bodies will be hydrolyzed. The aluminum time period, and during the time following SCWO system- must be removed before SCWO processing. Accordingly, ization, prior to the processing of hydrolysate. the basic hydrolysate will be neutralized with acid in the aluminum precipitation reactor to precipitate aluminum salts. Finding 3-9. A manual is needed that addresses the concerns Upon neutralization, CN− is converted to HCN in solution discussed above for the periods between equipment delivery and released into the gas phase above the solution. The spe- and systemization, during systemization, and between sys- cific CN−/HCN equilibrium in solution depends on the pH temization and operations, and including integrated SCWO of the solution. HCN has an acid strength (pKa) of ~9, so 99 and water recovery system operation. Creating the manual in percent of the cyanide in solution at neutral pH (7) will be electronic form would allow for maximum visibility across in the form of HCN. The concentration of HCN in solution a wide range of information in a searchable framework on versus the gas phase in the hydrolysate storage tanks (i.e., the portable platforms such as tablet computers. distribution of HCN between the liquid and the gas over the surface of the liquid) depends on pH, salt concentrations, the Recommendation 3-10. The Blue Grass Chemical Agent presence of a liquid/gas interface, mixing, and temperature. Destruction Pilot Plant project staff should prepare a manual In addition to the measurements cited above, the white specifically to be used in the interim period preceding SCWO paper calculated the estimated HCN concentrations in the systemization, throughout the SCWO systemization time gas phase above the hydrolysate at 77°F and 140°F for pH period, and during the time following SCWO systemization, ranging from 6 to 10. The results ranged from 8 to 3,200 mg/ prior to plant operation. This manual should also address m3 (gas) above the solutions, depending on the combination integrated SCWO and water recovery system maintenance of temperature and pH of the hydrolysate. The level of HCN and operation during these periods and should address all the that is immediately dangerous to life and health (IDLH) is concerns discussed above. The manual should be prepared 25 mg/m3.15 in electronic form. After neutralization of the hydrolysate to precipitate the aluminum salts, the gas over the hydrolysate/salt slurry in the aluminum precipitation reactor will be vented to carbon- CYANIDE containing canisters in the SCWO processing building. This Certain of the munitions types in the Blue Grass stock- is a closed system and limits opportunities for worker expo- pile contain energetic materials in the form of either propel- sure. The venting is through activated carbon filters. In the lant or burster charges. These materials include composition worst case calculated, a filter change-out every two weeks B4, tetrytol, and M28 propellant. The energetic materials or so would be required (BPBGT, 2011). will be treated with hot caustic (NaOH), and the resulting The neutralized solution, including any HCN/CN− hydrolysate will be strongly basic. Cyanide (CN−) is a minor present in the solution, will next be sent to the aluminum side product of hydrolysis of the energetics. The presence of filtration system, in the SCWO processing building, where cyanide or hydrogen cyanide (HCN) could pose contact and inhalation hazards should a worker touch the hydrolysate 14This is a Short-Term Exposure Limit (STEL) assigned by the National or breathe any vapors that might be above the liquid in the Institute for Occupational Safety and Health (NIOSH). It is a 15-minute hydrolysate storage tanks. This would be limited to main- time-weighted average (BPBGT, 2011). 15The IDLH is defined by NIOSH as the level of a chemical that is “likely tenance activities as no worker should be in contact with to cause death or immediate or delayed permanent adverse health effects or hydrolysate or the vapor space in the hydrolysate storage prevent escape from such an environment” (NIOSH, 2004).

OCR for page 24
IMPLEMENTATION OF SUPERCRITICAL WATER OXIDATION AT BLUE GRASS 31 it will be filtered to remove precipitated aluminum salts. The risk and safety analysis to identify the risks and mitigation aluminum filtration system is vented to the atmosphere, so strategies. any HCN gas released from the liquid during the filtration operation will be vented to the atmosphere.16 The solution Finding 3-11. Which methodology will be used to manage will then be pumped to the SCWO feed system tank, where the cyanide problem has not yet been determined. Whatever it will be combined with agent hydrolysates and stored as the methodology is used will change the chemistry of the SCWO blended feed for the SCWO. system feed and could impact the performance of the system. HCN is soluble in aqueous solutions, therefore, the HCN/CN− remaining in the hydrolysates will be transported Recommendation 3-13. The Blue Grass Chemical Agent in solution to the SCWO reactor from the blended hydroly- Destruction Pilot Plant project staff should identify an ef- sate tank and destroyed by hydrolysis and oxidation. Earlier fective cyanide management system that does not negatively experiments showed that the SCWO process destroys cya- impact the performance of the SCWO unit. nide to below detection levels (<10 µg/L in aqueous solution; <4 µg/m3 in gas). The process involves either oxidation to OVERALL SYSTEM OPERATIONS AND COMPUTER isocyanate and then hydrolysis to CO2, or hydrolysis to for- MODEL mate then oxidation to CO2. The BGCAPP project staff are investigating a variety To prepare for system operations, BGCAPP will take of mitigation strategies. These include oxidation strategies, various steps to simulate the actual operations. They have other chemical treatment strategies, physical separation strat- designed full-size indoor simulators for thermowell and liner egies, and the reengineering of hazard controls. This work is replacement to acquaint workers with actual operations and under way as of this writing and is not ready to be evaluated safety processes. Process control simulators are also being by the committee. Whatever strategy is eventually selected developed. One of the simulators will replicate the SCWO will have to be evaluated for impacts on the SCWO process. control room human–machine interface and allow control room operator trainees to manipulate simulated controls and Finding 3-10. Cyanide will be present in the energetics experience subsequent system responses— for example, the and blended hydrolysate streams. This must be managed to expected responses to opening a valve, initiating feed flow, protect worker safety. The testing done to date shows that and the like. This simulation will not, however, allow the the SCWO process will destroy any HCN or CN− present in exploration of unexpected operational conditions and their the liquid hydrolysate feed to undetectable levels. A formal impact on the system. There is no computer simulation model risk analysis would be needed to quantify the actual risks of the SCWO and the integrated SCWO/WRS system, nor is to workers. one currently planned.17 A computer simulation model imitates the operation Recommendation 3-11. If cyanide cannot be removed from of a real-world process or system over time. It is a tool to the hydrolysate, affected personnel should be informed that investigate virtually the behavior of the system under study. if exposure to hydrolysate occurs in an off-normal situation, Computer simulation is critical to understanding how a sys- the hydrolysate could have contained up to approximately tem will perform when in operation by allowing operators 7 ppm, or 300 µM, cyanide. Information about the possible to vary system parameters in the simulation model to see presence of cyanide should also be included in emergency how the system would react. This capability is invaluable response manuals or other documentation so that first re- for planning operations, testing alternative system settings, sponders are aware of the potential risk. Workers should be understanding system interdependencies, training personnel, trained in how to recognize exposure to cyanide and what understanding risks, and improving safety and emergency should be done while awaiting first responders. response readiness. In the context of this study, a computer simulation model Recommendation 3-12. The Blue Grass Chemical Agent would represent the operation of the SCWO unit and the Destruction Pilot Plant project management should quantify entire SCWO/WRS over time. By changing variables in such the amount of HCN gas that vents from the aluminum filtra- a simulation model, predictions could be made about system tion system and install an appropriate mitigation to avoid behavior in various operational circumstances and condi- worker/public exposure if the concentrations are above the tions. This could serve to familiarize operators with potential local air quality release limits. If cyanide cannot be removed bottlenecks in operations and to identify safety concerns. A from the hydrolysate, project management should conduct a computer model also offers a mechanism for examining ways to mitigate those concerns by simulating changes in opera- 16John Barton, chief scientist, Bechtel Parsons Blue Grass Team, 17Personal communication between Steven Mantooth, Blue Grass Chemi- “BGCAPP Process Overview,” presentation to the committee on January cal Agent Destruction Pilot Plant program office, and James Myska, NRC 7, 2013. study director, on February 7, 2013.

OCR for page 24
32 ASSESSMENT OF SUPERCRITICAL WATER OXIDATION SYSTEM TESTING FOR THE BGCAPP tions, processes, feed composition, maintenance and replace- SCWO process to a reasonable degree could aid in ment intervals, and the like before operations commence. achieving the project schedule by making operations BGCAPP could use such a computer simulation model to more efficient, helping with maintenance scheduling, optimize maintenance schedules, to forecast performance and avoiding unscheduled shutdowns of the SCWO for long-term operations, and to uncover and prepare for reactors. otherwise unexpected events. The results of these simula- tion scenarios could differ sharply from the 100-hr factory The development of a computer simulation model could acceptance tests that were conducted using simulant and can provide benefits for understanding and optimizing the com- prepare the operators for actual SCWO/WRS operations at bined SCWO/WRS system. On the other hand, chemical BGCAPP. plants routinely operate processes with complexity similar to In addition to understanding the operations of the entire SCWO without the benefit of computer simulation ­ odels.m SCWO/WRS system, the computer simulation model could The SCWO at BGCAPP will process only a few, highly be designed to study the interaction of multiple SCWO units characterized streams and has been shown through extensive so that intersystem dependencies (physical as well as opera- testing to successfully treat these streams. Also, the SCWO at tional) on schedules and maintenance and related safety and BGCAPP will only operate for a few years and then be shut risk issues could be understood and assessed. down and disposed of. Clearly, the effort to create a computer Critical input parameters for a computer simulation simulation model would need to be considered in the context model would come from the FOAK test results and later sys- of the benefit it would provide. temization of the SCWO units. The computer model could then simulate the long-term operations of the SCWO/WRS Finding 3-12. There is no computer simulation model system. A few potential benefits of a computer simulation that describes the integrated system operations, over model are these: time, of all SCWO system processes or of the entire SCWO/WRS. A model tailored for the SCWO and the •  wealth of process data have been obtained through A integrated SCWO/WRS systems would permit system numerous simulant tests of the SCWO unit and ancil- performance to be optimized. lary systems. The data reside in many different and often lengthy reports, which makes it difficult to ac- Recommendation 3-14. The Blue Grass Chemical cess them, even if one knows that some desired datum Agent Destruction Pilot Plant project staff should resides within a given report. A computer simulation consider investing in a simulation-optimization study model would allow the compilation of this informa- where external experts can help develop a computerized tion from diverse sources into a single comprehensive simulation-optimization model that describes the opera- depository that would then be easily accessible to all. tions, over time, for the entire combined supercritical •  here has been concern about knowledge retention, T water oxidation and water recovery system. especially considering the long time and change in personnel between the time of SCWO FOAK test- REFERENCES ing and BGCAPP operation with hydrolysate. The compilation of the existing knowledge and data into a BPBGT (Bechtel Parsons Blue Grass Team). 2011. Cyanide Formation Dur- ing the Caustic Hydrolysis of Energetic Materials—Potential Impact to computer simulation model would be a way to retain BGCAPP Unit Operations, Rev. 3, WP125, November 14. Richmond, that knowledge. Ky.: Blue Grass Chemical Agent Destruction Pilot Plant Program Office. •  afety protocols for workers performing routine S BPBGT. 2012. Systemization Implementation Plan (SIP), Rev. 0, August maintenance on one SCWO system while a second 30. Richmond, Ky.: Blue Grass Chemical Agent Destruction Pilot Plant system is operating in an adjacent space have not Program Office. BPBGT. 2013. Test Report for Supercritical Water Oxidation (SCWO) First- yet been developed. The computer simulation model of-a-Kind (FOAK) Test, April, Preliminary Draft. Richmond, Ky.: Blue would allow analysis of what-if scenarios, so that Grass Chemical Agent Destruction Pilot Plant Program Office. engineers could probe potential failure modes and NIOSH (National Institute for Occupational Safety and Health). 2004. design safety protocols to protect workers. NIOSH Respirator Selection Logic, Publication No. 2005-100. Cincin- •  computer simulation model could aid in optimizing A nati, Ohio: NIOSH. NRC (National Research Council). 2011. Assessment of Approaches for the operation of the SCWO system. The committee Using Process Safety Metrics at the Blue Grass and Pueblo Chemi- recognizes that this is a unique, short-term project cal Agent Destruction Pilot Plants. Washington, D.C.: The National and not an ongoing industrial operation that would Academies Press. span many years or decades. Still, optimizing the