3

Supercritical Water Oxidation Process for the Treatment of VX Hydrolysate

MATERIALS OF CONSTRUCTION AND OTHER ISSUES

The subject of SCWO for treating VX hydrolysate was discussed extensively in Using Supercritical Water Oxidation to Treat Hydrolysate from VX Neutralization (the SCWO report) (NRC, 1998). The treatment of VX hydrolysate is significantly different and more complex than previous treatments using SCWO. Because of the multitude of outstanding challenges with this process, the committee recommended that a pilot-scale test (i.e., an engineering-scale test, [EST]) be done to define operating characteristics and to demonstrate sustained continuous operation. A number of critical issues identified in the recommendations of the SCWO report have not been sufficiently addressed in the ADP (60-percent design package).

Although progress has been made on two of the most critical issues, the selection of materials of construction (MOC) and demonstration of salts/solids handling in and downstream of the SCWO reactor, but has not yet been demonstrated during the EST. A number of key issues, including demonstration of a robust pressure let-down system, worker safety (see below), and regulatory, analytical, and disposal issues associated with salts and other process effluents, have yet to be addressed. A contract has been awarded to General Atomics, Inc., for an EST to be carried out with VX hydrolysate in 2000. The results will have to be carefully evaluated to ensure that lessons learned are incorporated into the design and operation of the full-scale SCWO process. The Army should consider design contingencies or alternative hydrolysate management options in case adequate process throughput cannot be achieved with the current design.

The selection of appropriate MOCs is a key requirement for the successful treatment of VX hydrolysate by SCWO (see Finding 2 and Recommendation 2 of the SCWO report [NRC, 1998]). SCWO treatment of VX hydrolysate involves highly corrosive and erosive environments that change drastically as the hydrolysate passes through the reactor and its peripheral components (inlet, outlet, and pressure let-down valves). As the oxidation of VX hydrolysate proceeds, the reaction mixture undergoes dramatic changes in pH (from strongly basic initially to slightly acidic as acids are formed), temperature, salt content, and dielectric constant. Based on available data, the committee believes that the hydrolysate becomes a changing, multiphase mixture consisting of dense fluids, salt solutions, and precipitated solids.

The treatment of VX hydrolysate puts very stringent requirements on the MOCs, which must be extremely durable and reliable. Furthermore, the use of unusual MOCs may create challenges for reactor fabrication. Consequently, the importance of tests for specifying MOCs to be used in the full-scale system cannot be overstressed. To date, no demonstrated material has provided reliable operation over a sufficient period of time; however, recent results indicate that platinum and platinum alloys may be viable materials. The Army recently provided the Stockpile Committee with preliminary draft evaluations of the results of two tests: static coupon tests in small laboratory-scale autoclaves (batch testing) with hydrolysate surrogate1 (the SWEC test) (Ballinger, 1999; Jensen et al., 1999); and General Atomics continuous coupon testing with VX hydrolysate and hydrolysate surrogate (the General Atomics MOC test; three campaigns from July 31, 1999, to October 4, 1999, totaling 500 hours) (Jensen et al., 1999).

SWEC Test Results

Screening tests were performed in a small laboratory-scale Alloy 625 autoclave (at 3,500 psi) charged with a small volume of liquid hydrolysate surrogate, a headspace of gas (oxygen or an inert gas), and coupons supported on yttriastabilized zirconia (YSZ) buttons (for galvanic isolation from

1

Hydrolysate surrogate is comprised of 50.5 wt percent deionized water, 15.3 wt percent sodium salt of isethionic acid, 13.2 wt percent dimethyl methylphosphonate, 9.8 wt percent diethanolamine, 9.0 wt percent sodium hydroxide, and 2.2 wt percent isopropyl alcohol (Landry, 1999).



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 12
Integrated Design of Alternative Technologies for Bulk-Only Chemical Agent Disposal Facilities 3 Supercritical Water Oxidation Process for the Treatment of VX Hydrolysate MATERIALS OF CONSTRUCTION AND OTHER ISSUES The subject of SCWO for treating VX hydrolysate was discussed extensively in Using Supercritical Water Oxidation to Treat Hydrolysate from VX Neutralization (the SCWO report) (NRC, 1998). The treatment of VX hydrolysate is significantly different and more complex than previous treatments using SCWO. Because of the multitude of outstanding challenges with this process, the committee recommended that a pilot-scale test (i.e., an engineering-scale test, [EST]) be done to define operating characteristics and to demonstrate sustained continuous operation. A number of critical issues identified in the recommendations of the SCWO report have not been sufficiently addressed in the ADP (60-percent design package). Although progress has been made on two of the most critical issues, the selection of materials of construction (MOC) and demonstration of salts/solids handling in and downstream of the SCWO reactor, but has not yet been demonstrated during the EST. A number of key issues, including demonstration of a robust pressure let-down system, worker safety (see below), and regulatory, analytical, and disposal issues associated with salts and other process effluents, have yet to be addressed. A contract has been awarded to General Atomics, Inc., for an EST to be carried out with VX hydrolysate in 2000. The results will have to be carefully evaluated to ensure that lessons learned are incorporated into the design and operation of the full-scale SCWO process. The Army should consider design contingencies or alternative hydrolysate management options in case adequate process throughput cannot be achieved with the current design. The selection of appropriate MOCs is a key requirement for the successful treatment of VX hydrolysate by SCWO (see Finding 2 and Recommendation 2 of the SCWO report [NRC, 1998]). SCWO treatment of VX hydrolysate involves highly corrosive and erosive environments that change drastically as the hydrolysate passes through the reactor and its peripheral components (inlet, outlet, and pressure let-down valves). As the oxidation of VX hydrolysate proceeds, the reaction mixture undergoes dramatic changes in pH (from strongly basic initially to slightly acidic as acids are formed), temperature, salt content, and dielectric constant. Based on available data, the committee believes that the hydrolysate becomes a changing, multiphase mixture consisting of dense fluids, salt solutions, and precipitated solids. The treatment of VX hydrolysate puts very stringent requirements on the MOCs, which must be extremely durable and reliable. Furthermore, the use of unusual MOCs may create challenges for reactor fabrication. Consequently, the importance of tests for specifying MOCs to be used in the full-scale system cannot be overstressed. To date, no demonstrated material has provided reliable operation over a sufficient period of time; however, recent results indicate that platinum and platinum alloys may be viable materials. The Army recently provided the Stockpile Committee with preliminary draft evaluations of the results of two tests: static coupon tests in small laboratory-scale autoclaves (batch testing) with hydrolysate surrogate1 (the SWEC test) (Ballinger, 1999; Jensen et al., 1999); and General Atomics continuous coupon testing with VX hydrolysate and hydrolysate surrogate (the General Atomics MOC test; three campaigns from July 31, 1999, to October 4, 1999, totaling 500 hours) (Jensen et al., 1999). SWEC Test Results Screening tests were performed in a small laboratory-scale Alloy 625 autoclave (at 3,500 psi) charged with a small volume of liquid hydrolysate surrogate, a headspace of gas (oxygen or an inert gas), and coupons supported on yttriastabilized zirconia (YSZ) buttons (for galvanic isolation from 1 Hydrolysate surrogate is comprised of 50.5 wt percent deionized water, 15.3 wt percent sodium salt of isethionic acid, 13.2 wt percent dimethyl methylphosphonate, 9.8 wt percent diethanolamine, 9.0 wt percent sodium hydroxide, and 2.2 wt percent isopropyl alcohol (Landry, 1999).

OCR for page 12
Integrated Design of Alternative Technologies for Bulk-Only Chemical Agent Disposal Facilities the autoclave). Three separate conditions were tested to simulate reaction zones in a continuous SCWO reactor: Zone 1: heat-up/mixing zone—surrogate, 690°F, inert gas (argon) Zone 2: reaction zone—surrogate, 1,200°F, oxygen Zone 3: cool-down/quench zone—end product surrogate, CO2, O2, 700°F In addition to the Alloy 625 of the autoclave itself, the materials or material junctures examined were: nickel (Ni) 201, platinum, platinum to Hastelloy C-276 weld in Zone 3, the YSZ supports, and arc-plasma sprayed YSZ on zirconium (thermal cycling only). The nickel, YSZ supports, and YSZ coating on zirconium exhibited poor performance (high corrosion, fracture, and delamination, respectively) in Zone 1 and Zone 2. Platinum was deemed to be satisfactory in Zones 1 and 3 but exhibited corrosive attack, including pitting and the formation of a platinum-sulfur compound, at the grain boundaries in Zone 2. The platinum/Hastelloy C-276 weld was determined to be satisfactory in Zone 3. The platinum coupons in experiments for Zone 2 resulted in only a 0.05 percent weight gain after 160 hours of exposure. However, photomicrographs indicated two forms of attack: (1) formation of a reaction product on the grain boundaries; and (2) significant pitting of the surface. Annealed samples underwent the same forms of attack (a total of three samples were studied). The reported formation of a platinumsulfur compound from oxidized sulfur (from the sulfate product of the sulfonic acid moiety in the hydrolysate surrogate) in an oxidizing environment is somewhat surprising, although the report claims it was consistent with thermodynamic considerations (General Atomics, 1999a). Although these preliminary results are discouraging, the chemical and physical environments in the system used for these batch tests (i.e., headspace in the batch reactor, different phases, and variable temperatures) may be very different from those in a continuous SCWO reactor. Additional fundamental work and continuous reactor testing appear to be necessary. For example, oxygen depletion resulting from the corrosion/oxidation of the wall of the autoclave could account for the presence of reduced sulfur species, including hydrogen sulfide. The Army and the contractor will need a more thorough understanding of the processes in the autoclave and of the reactivity and thermodynamics of sulfur species and platinum. Furthermore, because the presence of platinum-sulfur compounds implies platinum oxidation, with the consequent potential for solubilization, the effluent from the reactor should be analyzed for platinum. General Atomics Test Results The preliminary results of recent MOC tests conducted by General Atomics under continuous conditions in a SCWO reactor are more encouraging (General Atomics, 1999b). This series of three sequential MOC coupon tests used a hydrolysate surrogate and VX hydrolysate provided by the Assembled Chemical Weapons Assessment (ACWA) program. The VX hydrolysate used in the tests was higher in NaOH concentration and had higher trace amounts of chlorine and fluorine than the hydrolysate that will be produced from the NECDF stockpile. The goal of the three tests was to recommend MOCs for the SCWO reactor. Coupons made from a range of materials were exposed to representative SCWO conditions to determine corrosion rates and mechanisms. Because the amount of available VX hydrolysate was insufficient to support a test with a duration of 500 hours, one test with VX hydrolysate was run for 167 hours, and two tests with hydrolysate surrogate were run for 167 hours to expose the coupons to a total of 500 hours of operation. The MOC tests were carried out in a 4-in diameter by 6-ft long reactor that had been used during ACWA demonstration tests (General Atomics, 1999b). High-pressure oxygen was used as the oxidant. Coupons of various materials were attached to a coupon tree placed along the reactor length. In this way, coupons could be placed in different regions of the reactor along its length, including a quench zone at the end of the reactor. The coupons were subsequently examined under a microscope to determine microstructural effects and thicknesses as a function of location in the SCWO reactor. The following materials were tested: noble metals (platinum, Pt-20%Ir, Pt-10%Rh), Au-8%Pd, Rh) platinum-plated metals, ceramics (magnesium-stabilized zirconia [MSZ], YSZ) nickel-steel alloys (C276, I617, I625, I690, HA25, HA188) refractory tantalum and niobium (for quench zone and downstream components only) The preliminary results of the tests are shown in Table 3-1. The results described in the General Atomics report can be summarized as follows (Jensen et al., 1999): platinum and Pt-20%Ir and Rh showed the least corrosion with no evidence of intergranular corrosion, suggesting them as prime candidates for MOCs Pt-10%Rh and Au-8%Pd exhibited intergranular attack as a corrosion mechanism platinum-plated coupons exhibited delamination in all but the quench zone, indicating that higher quality platinum plating or other liner materials will be necessary in zones where the coating components have complex shapes Ni-based steels, including Ni/Co alloys, showed much greater corrosion than the Pt alloys in zones other than the quench zone YSZ and MSZ ceramic coupons displayed coating

OCR for page 12
Integrated Design of Alternative Technologies for Bulk-Only Chemical Agent Disposal Facilities TABLE 3-1 Preliminary Results of Corrosion Testing with Noble Metals and Superalloys NOBLE METALS Maximum Corrosion (mils/year) Material Inlet Heat-up Section High-Temperature Reaction Section Quench Section Pt 55 47 8 Pt-20%Ir 68a 21 5 Pt–10%Rhb 68 81 5 Au–8%Pdb > 500 > 500 16 Rh 39 c c SUPERALLOYS Maximum Corrosion (mils/year) C276 (Ni-steel) c 2,360 26 I617 (Ni-steel) c 1,940 c I625 (Ni-steel) c 2,070 26 I690 (Ni-steel) 2,100 2,180 131 HA25 (Ni-steel) 2,500 2,200 26 HA188 (Ni-steel) c 1,550 c Tantalum c c 52 Niobium c c 1,940 Notes: a Nozzle deterioration caused direct impingement that may have caused higher corrosion b Displayed intergranular corrosion c Not tested Source: Adapted from Jensen et al., 1999. spallation and corrosion in the reaction zones all metals and alloys except niobium exhibited little corrosion in the quench zone Corrosion rates of the platinum and Pt-based alloys ranged from 50 –65 mils/year, based on micrographs revealing uniform recession (i.e., no intragranular attack) (Jensen et al., 1999). On the basis of these preliminary estimates and the absence of undesirable corrosion mechanisms, such as intergranular attack, General Atomics has recommended a platinum liner (20 mils thick) for the EST and a Pt-20%Ir alloy liner (12 mils thick) as a backup (Burchett, 1999; Jensen et al., 1999). MOC testing and performance evaluation will have to be an intrinsic part of the EST program, both for the reactor and for downstream components exposed to aggressive environments (e.g., pressure and temperature reduction components). The selection and reliability of all primary components will have to established by a thorough demonstration of process performance. Current plans for the NECDF call for installing two full-scale SCWO reactors (both sized for 100 percent of the planned throughput of VX hydrolysate) and one uninstalled reactor in reserve. All three reactors will have platinum liners 20 mil thick (Jensen et al., 1999). However, a thickness of 20 mil is substantially thinner than the required liner thickness based on test results for 100-percent throughput of VX hydrolysate for a single reactor, with no conservative design margin. This change in thickness would significantly reduce the capacity margin provided for in the original design. Because of the high level of uncertainty associated with the performance of MOCs, a highly conservative design margin, either by an increase in liner thickness or through reserve reactor liners, will be essential. Another factor to be considered is the time required to fabricate additional liners if the initial supply is depleted. Recently, General Atomics claimed it was able to fabricate 20-mil thick liners of the required diameter for the reactor. General Atomics plans to “float ” a precious-metal liner in a cylindrical Hastelloy pressure vessel and use cooled elastomeric O-rings that have performed satisfactorily on other SCWO systems to form the SCWO reactor. The annular space between the liner and the vessel wall will be monitored for leaks to indicate when change-out of the liner is required. No decision has been made yet on whether to use platinum or Pt-20%Ir. Because they have markedly different mechanical properties, these two liner materials may require significantly different fabrication methods. Platinum is relatively weak and very ductile; Pt-20%Ir is less ductile but 10 times stronger. Final selection of the liner material for use at the NECDF was scheduled for early 2000. Fabrication

OCR for page 12
Integrated Design of Alternative Technologies for Bulk-Only Chemical Agent Disposal Facilities times for the liner and reactor assembly have been reported to be at least two to three months. Production of hydrolysate for the EST was begun in 1999 to ensure that sufficient VX hydrolysate would be available for the EST planned for the spring of 2000. In summary, parallel with plans for further MOC tests, the Army has elected to move forward with a design that will incorporate a liner in the SCWO reactor system that may or may not be used in a sacrificial capacity. Batch testing suggests that significant corrosion occurs even with precious-metal liners. The preliminary analyses of the General Atomics tests are useful for identifying materials that might be viable but are not sufficient for making a final selection, which will require reconciling the significant disparities between the General Atomics and SWEC test results. In addition, the drastic differences in the performance of platinum materials during batch and continuous tests should be explained and verified. Although oxygen-depleted zones may have been present in the batch tests, they might also be present in a full-scale SCWO system processing VX hydrolysate because of poor mixing. Therefore, this issue must be resolved. The EST should provide further information on corrosion kinetics and mechanisms, as well as a demonstration of the viability of liner fabrication and the reliability of MOCs. In short, the Army must have a more fundamental and practical understanding of MOCs. Using a liner, or an entire reactor, in a sacrificial capacity may become a viable alternative, but this decision must be based on an understanding of reactor lifetime and the development of maintenance, monitoring, and change-out/ replacement procedures and schedules. Considerable testing, including pilot-scale tests, will be necessary to address these issues. Corrosion of the metal components in the reactor will cause these metals to appear in the effluent from the SCWO reactor and, ultimately, in the salt stream from the evaporator. The ADP design does not describe how these metals, some of which are regulated (e.g. chromium, nickel), could affect the disposal of the salt streams, which consist largely of sulfate and phosphate salts. The Army will have to determine the amounts and effects of metals in the SCWO effluent as a result of corrosion, including their effects downstream of the SCWO reactor and how their presence will affect handling and disposal, overall effluent quality, monitoring, and regulatory and permitting issues. Operational Issues Problems with salt/solids management, including the potential for plugging and difficulties in pressure let-down, was thoroughly described in the SCWO report (NRC, 1998) but have not been resolved in the ADP as currently modified. Further development work and testing will be necessary to address this important issue. The tests carried out by General Atomics to demonstrate high destruction efficiencies of VX hydrolysate indicated potentially significant operational problems with the handling of solids in the SCWO reactor and with the robustness of the pressure let-down system (NRC, 1998). General Atomics has since carried out two additional tests with a modified pressure let-down system to improve system reliability (General Atomics, 1999b; Jensen et al., 1999). The new design calls for gas/liquid separation after quench at high pressure and prior to pressure let-down. Test results have shown satisfactory continuous operation for up to 40 hours without plugging or corrosion/erosion of the pressure let-down system (General Atomics, 1999b). However, the system did not operate without interruption for extended periods during demonstration tests for the ACWA program. The interruptions were attributed to plugging by entrained solids in the aqueous reactor effluent from corrosion of the liner materials that did not redissolve during temperature and pressure let-down. A platinum liner was not available for the test because of schedule constraints. Recommendation 3-1a. The Army should develop criteria and a schedule for resolving design and operational issues raised in the 1998 report, Using Supercritical Water Oxidation to Treat Hydrolysate from VX Neutralization, that have not yet been resolved for supercritical water oxidation operation at Newport. These issues include materials of construction, fabrication methods, system plugging, pressure letdown, and the duration of successful continuous pilot-scale operations. Recommendation 3-1b. The Army should pursue the testing of materials of construction for treating VX hydrolysate by supercritical water oxidation (SCWO) more aggressively to finalize materials selection, design, and fabrication methods for critical components, including the SCWO reactor, inlet, and pressure let-down system. This testing should clearly define mechanisms and rates of corrosion and erosion under the range of anticipated process conditions. An independent panel of experts in materials of construction should evaluate testing to date and identify further needs to ensure that the reliability of the SCWO system is adequate to meet the processing objectives. SAFETY ISSUES Several safety issues for processing VX hydrolysate by SCWO must still be resolved. First, the SCWO reactor operates at 1,200°F and 4,000 psig pressure. Protective measures should be put in place to protect personnel in case of a rupture of the reactor. The system should be anchored and barricaded to protect workers and the environment. Second, current plans call for high-purity oxygen to be used as the oxidant being supplied to the reactor at high pressure, which will require appropriate safeguards. The evaluation of safeguards for using pure oxygen should include compatibility

OCR for page 12
Integrated Design of Alternative Technologies for Bulk-Only Chemical Agent Disposal Facilities with MOCs, both during normal operating conditions and potential failure modes (e.g., exposure of SCWO reactor components at high temperature and pressure if leakage of the primary liner occurs). The current ADP calls for carbon filtration of the off-gases from the SCWO reactor. The purpose of the carbon beds is not clear to the Stockpile Committee. SCWO off-gas enriched in oxygen relative to air (as a result of excess oxygen input to the reactor) would create a potential fire hazard. If carbon filtration is required to remove trace organic species from the SCWO off-gas prior to emission, added safeguards may be required to prevent carbon fires that could result from reaction of the carbon with unreacted oxygen in the SCWO off-gas. If the carbon beds are intended to remove traces of agent, they appear to be superfluous because no agent should be present in the off-gas after hydrolysis and the subsequent reaction of the hydrolysate in the SCWO reactor. In the committee's opinion, treatment by carbon filtration to guard against the emission of chemical agent could increase complexity and risk without providing any benefit. Trace constituents in SCWO effluent gases can be quantified during the EST. Recommendation 3-2. For worker protection and secondary containment, the final design package for the Newport facility must include the physical hazard controls (e.g., protective barricades) common to industrial operations involving high pressure and stored energy. Systems must be designed to minimize leaks, plugging, and ruptures of the supercritical water oxidation reactor and associated plumbing and protective barriers. Secondary containment equipment will also be necessary, including safety systems for handling high-purity oxygen at high pressure, such as protection against downstream fires and explosions caused by contact between combustible materials (e.g., activated carbon) and oxygen-enriched gas streams under normal and potential upset conditions.