Loading...
HomeMy WebLinkAbout9a Chromate Removal Plant Part 1_20170726IWC 11-41 Chromate Removal at the Hanford Site Dean Neshem Jr. CH2MHILL Plateau Remediation Company Richland, Washington Peter Meyers ResinTech West Berlin, New Jersey Francis DeSilva ResinTech West Berlin, New Jersey KEYWORDS: chromate, groundwater, Hanford, hexavalent chromium, ion exchange. IWC 11-41 ABSTRACT Remediation of groundwater containing hexavalent chromium has been in progress at the Hanford site for more than 15 years. Although the early systems used type I (gel) strong- base anion exchange resin and off site regeneration, the most recent systems use a long -life single -use resin. Use of the long -life resin is projected to result in significant operating cost savings. This paper explores the history of the chromate remediation efforts at the site, a review of more than 15 years operating experience with strong -base resin and the operating results with the long -life resin. IWC 11-41 BRIEF HISTORY The Hanford site, located in eastern Washington state, was chosen in 1943 as part of the Manhattan project for production of plutonium for nuclear weapons (figure 1). The site was chosen because it was close to the established electrical infrastructure from the Grand Coulee dam, was sparsely populated, considered stable geologically, and had a large amount of water available from the Columbia River. Residents of the small communities of White Bluffs and Hanford were given 30 days to move from their homes. The initial reactor went from design to completion in 13 months. The B reactor was the worlds first plutonium production facility. It had a graphite core, 2004 process tubes and flowed 186,000 gallons per minute of cooling water. (Hanford Story video,2011). Cooling water taken from the Columbia river made a single pass through the reactor and was sent to concrete cribs and basins to allow heat and short lived radionuclides to decay prior to being returned to the river. Sodium dichromate was added to this water as a corrosion inhibitor. Naturally occurring (unenriched) uranium 238 was transmuted into plutonium 239, which was then extracted and used in the Fat Man bomb that was detonated over Nagasaki Japan on August 9, 1945 The successful completion of the Hanford works remains one of the boldest and remarkable engineering feats of all time. Although the use of nuclear weapons remains very controversial, there is little doubt that the use of Fat Man greatly reduced American casualties that would have occurred had an invasion of the Japan mainland been necessary. Over the years, additional reactors were built. A total of 8 single pass reactors were constructed and operated at Hanford between 1944 and 1971, all using the same once through cooling and sodium dichromate as a corrosion inhibitor. Much of the land surrounding the Hanford site is very porous. Leakage of concentrated sodium dichromate from underground piping and other accidental spillage resulted in substantial quantities of chromate finding their way into the soil and gradually moving downward toward the Columbia River. IWC 11-41 Washington Seattle Spokane Hanford Site land North Slope 100-H AREA 100-K AREA �1O HR.3 GW.1 �100-F AREA ')perable Unit ff 100 HR-3 GW aereble Unit ` 100 KR1 GW 100 FR-3 Gw' Operable Unit Operable Unit, 100 Areas Figure 1. Location of Hanford site CHROMIUM CONTAMINATION Hexavalent chromium on the Hanford site is primarily found near the reactor sites. It is estimated that a 2.0 km2 (0.8 mil has contamination at concentrations greater than 100 pg/L. (DOE CERCLA, 2011) Peak concentrations of 55,600 pg/L have been measured near the D and DR reactors. A substantial additional area has contamination below 100 fag/L. Plume maps are shown in appendix A. PILOT PLANT STUDIES Following the shift from plutonium production to site cleanup in the early nineties, a pilot plant was constructed to evaluate the use of ion exchange as a means of remediating the chromium in the groundwater on site. The initial studies used a variety of ion exchange resins available at the time and resulted in a determination that ion exchange could be used. A type I strong base anion resin was chosen for the full scale facilities to be constructed in the HR-3 and KR-4 operable units. Initially a 400 gpm plant was constructed (HR-3) to treat the largest area of contamination near the D and H reactors. A 200 gpm plant was also built near the K reactors with operation commencing in 1997. Control of the contaminant plumes is accomplished using a combination of extraction wells and injection wells. Contaminated groundwater is extracted from near the Columbia river and pumped to ion exchange facility for treatment. Following treatment the water is sent to a series of injection wells upgradient of the plume. This approximates a closed loop system where the injection water drives the contaminants toward the extraction wells. EXPANSIONS - As more information about the size of the contamination was gathered a series of expansions and continued testing of additional technologies have been implemented. Pump and treat capacity in the K reactors was expanded from 200 gpm in 1997 to 1100 gpm in 2009 with the addition of two additional facilities. Additional pump and treat capacity around the D and DR reactors began in 2004 with a 50 gpm pilot scale facility (DR-5) that is able to regenerate resin in vessel. Another innovative treatment technology was also employed in this area. It is known as the ISRM (In -Situ Redox IWC 11-41 Manipulation) barrier. This barrier consisted of a closely grouped network of wells that had sodium dithionite injected into them to create a reducing environment, converting hexavalent chromium to trivalent chromium and greatly reducing the amount of chromium reaching the Columbia river. A portion of the barrier failed to stop the flow of chromium prompting the expansion of pump and treat capacity to 600 gpm beginning in late 2010. The latest expansion of capacity is near the H reactor. A new facility is replacing the original IX system, increasing capacity from 300 to 800 gpm. Table 2. 100 Area IX System Description Influent Chromium Number of 0esO Flaw Conoantratlon System Tr - (Umin[ppm]) 1pa/l) statue KX 5 2.2i1 (EX) 60 AM,. KR4 3 1.136 (300) 30 Ace - KW 2 757(200) 150 AcM UR-5 1 189(50) 1.13W-3,3W Aclire° DX 4 2.371 (600) S00 Unger mnsin,atgnp HR-3 3 1.136 (300) 50 AEGve HX 8 3,028(8001 50 UnaarconsUuckv a. Shatdawe Spnnp 2011_ b. Complete October 2010. c Shvtdcwn Spnnp W12. E. Complete W ber 2011. HANFORD VESSEL DESIGN Ion exchange vessels for chromium treatment have employed the following design philosophy, the only exception being the DR-5 pilot scale plant. The system is modular, utilizing groups of four vessels or "trains" rated at 100 gpm. To expand capacity additional trains are added rather than changing the size of the vessels. Each vessel has an 80 ft3 capacity for ion exchange resin. Each of the four vessels in a train are in active operation in a lead-lag1- lag2-polish configuration. Resin is removed when the lead vessel is reaches capacity for chromium or the polish vessel effluent is greater than 10 fag/L. The four bed system has proven to be a predictable and forgiving system. Consistency in the vessel design between facilities allows for operational flexibility, since the operators can switch between facilities without extensive retraining. Over the years several modifications or changes were implemented to aid the dependability of the system. Initially the lead vessel was operated in a sacrificial mode. This was due to concerns that uranium and other radionuclides may accumulate on the resin presenting a transportation risk. During this time the lead vessel would be left in service for IWC 11-41 six months prior to removal, sampling and disposal. The remaining vessels would be rotated between lead, lag, and polish positions based on chromium breakthrough. Evaluation of sample results showed this as an unnecessary precaution and operation shifted to rotating all four vessels allowing for increased loading of the resin prior to regeneration. Containers appropriate for offsite transport of the resin for regeneration and a system to load/unload the resin efficiently were also implemented to minimize the need to handle the resin and to speed the changing of resin from the vessels. Metering pumps to add sulfuric acid for pH adjustment of the influent water were also developed to maintain influent water at pH 7 were implemented to minimize mineral scaling that would develop on the resin bed and distributors that would cause additional strain on the system pumps. RESIN REGENERATION The primary method used for treatment of spent strong base anion resin at the Hanford site has been off site regeneration. This method is simple for plant operation but leads to concerns over shipping and treating resin off site. Extensive sampling and analysis is required to ensure that the resin meets all shipping requirements. This requires a large investment in resin due to the long turnaround time as the resin is sampled, shipped, regenerated, shipped back and placed into service. Several variations of onsite regeneration have been tested over the years. Regeneration at a facility onsite was performed using sodium sulfate leaving the resin in the sulfate form (as opposed to the chloride form normally received from offsite). This resin had approximately 80% of the capacity as the resin in the chloride form. Additionally, the facility did not have the capacity to regenerate all of the resin needed to eliminate offsite regeneration. A pilot scale facility (DR-5) was also operated from 2004 to 2010 that used in vessel regeneration. This process used sodium chloride to strip the chromium from the resin. Sodium dithionite was used to convert the hexavalent chrome (anionic) to trivalent chrome (cationic). Hydrochloric acid was used in a second step to condition the resin. Phosphoric acid was used on the regenerant wastewater solution to precipitate the chrome, followed by sodium hydroxide to increase the pH and aid precipitation and settling. The precipitated chrome was separated by a filter press. BENCH SCALE TESTING In 2008, plans to expand the pump and treat capacity near the D and H reactors were being finalized. With this impending expansion, the decision to re- evaluate the ion exchange resin and regeneration process was made. A test skid was procured that allowed for six resins to be tested simultaneously using a one inch diameter column with a bed IWC 11-41 depth of up to 48 inches. The columns had individual feed tanks and pumps allowing for pH and flowrate to be varied individually. Figure 2. Resin test skid. Beginning in March of 2009, Seven different resins were tested. They included several strong base resins (including the resin's currently being used onsite), a styrenic weak base resin, as well as granular and bead forms of an epoxy polyamine resin. SeM[rnt< stymne St,— EPoq Epm:y pulyem— diveryWevame drviaylbmarre palyemer Shape h}sheriraf head %,h—I bead G.em.lm SPh—.d heed Acuvegml) Q-WM-Y'-- QPa---) P-M-tary emmamum amine P'rcgsac-Y amine 'type (mvn nchange) TYPe t straps Type 1 sEmB W ase nk b base I.. Wrak bax ilPxr1-1 pB 6to 14 Il ro 11 16.5 16.5 .hbilty to be resenveted? Y. Yc. No Nu . The pressurt. amp is rNrnmipcd ai ike ic+n4M1+nuac in ae raw 1er4wa b Ibe bahwa� Elax rrte is dCermieM. e�x rbe xnegenrme in rbe mw6ekm• ii The initial testing showed results similar to the operating facilities for the strong base resins at pH around 7, confirming the validity of the test skid as a method of evaluating resin performance. �- � Resin Evatl p2 Figure 7. Breakthrough Curves for the Second Resin Evaluation Additional tests were performed with the epoxy polyamine resins operated at reduced pH (5) showed exceptional affinity for chromate. F•tgum 2. Breaftruugh Curves for the First Resin Evaluation The initial test column of this resin lasted through the entire duration of the initial test series (almost nine months) without nearing capacity. The performance of this test column shifted the focus of the testing to learn more about this resin and evaluating if the switch to this resin would be the right choice for use at Hanford. Additional columns with much smaller bed depths were tested in an attempt to reach capacity of the resin. None of the tests IWC 11-41 were able to reach the total capacity of the resin, an idea about when the vessels would begin to allow chrome leakage into the lag vessels was determined. EPDXY POLYAMINE RESIN Making the switch to a weak base epoxy polyamine resin would require some significant changes to the way ion exchange resin is handled on site. This type of ion exchanger is granular rather than spherical. PHYSICAL PROPERTIES Polymer Structure Epoxy Polyarnme Functional Group R - N - (CH,)2+ Ionic Form as Shipped Free Base Physical Form Tough, Uniform Granules Screen Size Distribution 1.2 to 30 Mesh Nominal +12 mesh (US Sid.) C 25% -50 in (US Sid-) < 1% PH Range 1 to 14 Uniformity Coefficient Approximately 20 Stability Insoluble In Acids, Bases and Most Solvents Water Retention 52% to 62% Approximate Shipping Weight 38 lbs./cu. R. Swelling (free base to salt form) Less than 10% Total Capacity (free base form) >2.7 meq/mL Salt Splitting Capacity Approx. 10% Of Total The facility would require systems to deliver acid and caustic to adjust the pH of the groundwater being treated. These systems would also require bulk deliveries of chemicals due to the increased use rates needed to reach lower pH's. Additionally the type of coating inside the vessels was changed to one more appropriate to use in acidic conditions. Cathodic protection was also added to the design of new vessels to add an additional layer of protection against corrosion. The granular shape (instead of the typical bead shape) of the best performing resin generates increased differential pressure across the resin potentially increasing pumping costs and requiring reevaluation of the sizing of the feed pumps. Concerns about loading and unloading of the resin, resulted in a test to load resin into and out of one of the existing vessels to determine if the existing sluicing mechanisms would work. The test was successful, and not discernible difference was noted by the operators performing the work. New vessels also have had four air ports added to the bottoms of the vessels to aid the removal of the resin. Since the resin cannot be regenerated for reuse, a disposal path is required. To dispose of the resin onsite it must meet certain requirements to be accepted at the disposal facility onsite. Samples of the resin from testing were analyzed to determine if any additional treatment would be required for disposal. The results of these tests showed that the resin would potentially require stabilization (i.e. grouting) to meet TCLP for total chrome. The initial sample indicated stabilization would be required. Later samples, that were taken from resin samples with even higher chrome loading met requirements for disposal onsite. Additionally these tests showed that more than 90% of the chrome on the resin was in the trivalent form rather than hexavalent form. This helped to confirm that the exceptional capacity of the resin is due to more than simple ion exchange. Testing showed the resin IWC 11-41 had more than an order of magnitude greater capacity than any of the other resins tested. (Neshem, 2010) We believe the mechanism is ion exchange followed by reduction with the resin matrix and precipitation inside the resin matrix. This hybrid property appears to be both flow and pH sensitive in that the rate at which chromate is reduced to trivalent chrome is slower it higher pH. pH Testing — An effective operating range for the new range was determined through a set of bench scale tests. Very small volumes of resin were added to beakers of highly contaminated groundwater and stirred to ensure mixing. Hexavalent chromium concentrations were measured daily and water was replenished until the resin was unable to remove additional chrome. The results shown in figure 3 indicated that the resin capacity decreased substantially when the pH was greater than 6. Additional testing with larger resin volumes and samples measured every minute to more closely simulate normal operation showed a decrease in the rate of chrome removal at pH's above 6.5 as well Table 5. Results of SIR-700 pH Evaluation Chromium Solution (pH) Absorption. (mg) Relative Capacity 5.0 57,0 1.00 5.5 44.3 0.78 6.0 43-6 077 6.5 24.3 0.43 7.0 17.5 0.31 8.0 11.2 0.20 Figure 3. pH optimization testing INITIAL RESULTS WITH THE NEW EPDXY POLY AMINE RESIN Additional column testing was performed with both sulfuric acid and with hydrochloric acid to determine if there was an advantage in operating capacity to using one acid over the other. Very little difference was noted. a .eae �mw+ naa vo— ' Figure 4. Canparlson ofSIR-700 Performance Using Different Acid Types A life cycle study was performed to determine the present worth of various options. The single use epoxy polyamine resin was much less expensive than off site regeneration of strong base resin, very similar to the present worth of on site regeneration but without the complicated operation that on site regeneration would require. IWC 11-41 Table G. OX System Option Estimated Cost Comparlwit EaHrnaw C.w Baseline RegenaralEonti oftim P.9- attan 0."a In-Veasel Regeneretton` peeeneratm.- SlnplaOse Raaina Capital coat 10.76 1.73 992 8.15 7.33 Mneal O&M wait 333 2.3S 221 1.58 1.84 Of—yde-1 39.60 28.15 2912 21.85 2325 a. In miltliona of tlallats. 5. Oowez 21K ream with o8slte reyenenallon. e. P—te0 wait fer the Ok lacillty mel— 56 peeoent o'tntal reixteal regeneration faculty cods d.0 Injectlon oftmaled.9anaratbn WBet—., With questions about the epoxy polyamine resin answered, the decision to use the resin in the newest pump and treat facility was made. The resin was loaded into the vessels and operation began without pH adjustment due to schedule constraints and positive short term testing in the test skid. Operation without pH adjustment occurred for about 100 bed volumes over the course of several weeks. More than a week after the start of operation with pH adjustment chrome leakage was detected in the effluent of the lead vessel. 1200000 1000000 800000 600000 400000 200000 0 12/2 Train A tGW Total, gallons ♦ Lead Cr, ppb 12/4 12/6 12/8 12/10 12/12 12/14 70 60 s0 40 30 20 10 0 12/16 Lead vessel effluent values peaked several days later before stabilizing near 10-20 pg/L (influent was about 325 pg/L). During this time the effluent required little to no caustic to return the pH back to neutral. Initial operation stripped the pH buffering capacity of the resin leaving few active sites available for ion exchange. A week after the peak in lead column effluents, a much smaller peak in the lag-1 vessels as the resin was re -acidified. Once the caustic use increased to the expected levels the chromate leakage stabilized and was limited to the lead vessel only. After 2 months of operation the leakage from the lead vessels dropped below detectable levels (5 pg/L) and has stayed below detection since. As of July 2011, the resin has processed about 40,000 BVs, removed more than 138 kg (304 Ibs) and is not showing any signs of reaching capacity. After several months of operation, differential pressure across the resin beds began increasing to higher levels than expected increasing with each subsequent vessel in series. Inspection of the resin beds showed a layer of resin fines had built up on the top of each of the lag-1, lag-2, and polish vessels. Backwashing of the vessels, since they had been exposed to little or no chromate returned pressures to normal. In the future the resin will be backwashed more thoroughly when loaded to help minimize this problem. Additionally when the new facility near the H reactor is started, a change to three vessels in service at any one time is being made to reduce pumping costs since the resin is proving to be so much more effective than the strong base resins previously used. COST EFFECTIVENESS Along with resin testing, life cycle cost estimates were prepared for the facility built near the D reactors as shown in Figure 4. This estimate used the strong IWC 11-41 base resin with offsite regeneration as a baseline and compared this with the most effective strong base resin regenerated both on and off site, and the single use epoxy poly amine resin. Resin life for the epoxy poly amine resin was assumed to be 40,000 bed volumes (very conservative) due schedule constraints requiring the estimates to be prepared in parallel to the resin testing. These life cycle costs showed that the epoxy polyamine resin to be the most cost effective option saving almost $20 million over the 11 year lifespan of the facility. This is roughly equal to the cost of construction of the facility. Figure 4. Estimated life cycle costs With the successful implementation of epoxy poly amine resin at the D area facility, the next step is to evaluate the conversion of the remaining facilities near the K reactors to this resin. To do this a test is being implemented at the smallest facility to determine whether the existing simple acid injection system is capable of maintaining a desirable pH for use with the new resin. Monitoring of the vessel wall thickness will determine the suitability of the liner for use in an acidic environment. Operation with smaller volumes of resin per vessel is also being evaluated as an alternate method to reduce pressure loss across the resin bed. It is also beneficial to increase the headspace for backwashing since the vessels do not have an upper distributor that would prevent resin carry over in the event backwashing was too vigorous (newly constructed facilities have upper distributors that match the lower distributors to eliminate this concern. Leachables from epoxy polyamine resin The epoxy polyamine resins are included in the CFR's that relate to the use of ion exchange materials in food grade applications. Preliminary extraction and analysis indicate there should be no problem certifying this type of resin to ANSUNSF 61 at some future date, should customer demand dictate that this is necessary. NMI :MAN0201WAI Several significant challenges still need to be solved at the Hanford site. A pilot plant was also done for removal of technetium from groundwater. The results of this pilot plant were favorable, however no extensive studies were performed to optimize. Perhaps the biggest remaining challenge is that of removing radioactive iodine 1129 Radio -iodine was released in IWC 11-41 significant quantities over the years the plutonium facility operated, much of it airborne. As such there is significant contamination in much of the groundwater underlying the site. Perhaps the biggest challenges still await. The more that is learned about the site, the more surprises there are. IWC 11-41 References: The Hanford Story video, (2011) http://www.hanford.gov/c.cfm/video/tags.cfm/The_Hanford_Story Neshem, D (2010). Resin Evaluation and Test Report to Support DX Treatment System. SGW-41642, Rev. 1 Hanford Site CERCLA Five -Year Review, (2011), DOE-RL-2011-56 IWC 11-41 Appendix _r 100-D -__ 1OO-H 100-N f 100-K 11O0-F 10 1 � I r 200-West 1 i 13 I I � 1 � 1 200-East kw�r---------------- rconstituents ` 6 d Chromium COWS = 100 uglLj A Uranium (QWS � 30 ugfI-) � Technetium-99 (D1NS = 900 pCilL) Nitrate (OWS = 45 mgQ Strontium-90 (DM = 8 pCilL) lodina-129 (DWS= 1 pCK) Tritium (UWS = 20,000 pCLfL) ® Carbon Tetrachloride (DWS = 5 ug1L) Inner Area Boundary Outer Area Boundary Site Boundary Basalt Above Water Table 0 Columbia River 0 3 5 9Km I I I I (J 1.5 3 4.5 Mi syPo9d 400 Area 5W 300 Area rM IWC 11-41 Hanford site groundwater contamination plumes In The Upper Unconfined l • Mcnitoring Well Fall2009 Hexavalent Chromium Plume p g Extraction We] i Cr >= 20 ❑gIL and < 50 ug" ¢`I ® Injection Well Cr — 50 u & and < 200 uglL + Aquifer Tube Cr — 200 uglL and < 500 uglL ISRM Barrier Wells Cr >= 500 uglL and < 1.000 ug1L Transfer Line Cr >= 1400 uglL and < 2,000 uglL -� 1 • 96 523 Waste Sites M Cr>= 2,000 uglL and c 5,000 uglL %f Pump and Treat Buildings _ Cr >= 5,000 ug1L Area Boundary (] Columbia Mver R 100 200 300 M • DB 7D DD-17 3 } 0 .53 350 700 1.050 Ft Q..i09aa'.; cs27B :7 a1 08-546 Jr OB-54A D855 rd •Des / AT-D-iEl �, �j�/// • D8-73 ,(L 36-o+ , D8-88 �C AT-0-2-rvn • DB .Q a3'.81-4n } • D8-B ['ti pins.. ��♦` D5 13 0537 / AT-D-1.0 / / D5-A4 //r • D5-126 I 4 GS u' •p5 t4 �D5-35 5DS-125 RCdOK-1$.Q t 04-83 D`-33 Redm-260 - D'I's DS-15 Q4.39� D4-42 95-123 .0515 p0-69-2 i- RedoK-3-4..1 + 2D4-00 __ pa-37 65 3d; R¢GDD-41 0 * pq�8 0 D4 34 • DC - 3A D5-t 03 DD.-01.3 D4-32 DD-42-4+ �493 04-22 D4.11 ,��� 0"' DD-033+ 04-84 • 28 OS-179 D5.720 D4'13� \ 041 • D4.15 DS .� p5-104 DU-m4 D4-2a Da-31 Q5121 CN D4.eo D4-92 Q1-99 53 Da.25 O5-9z2 DS-102 C 2]1 C62P0 D4-&5 • C5-1' T'D4-19 D5.40 p5-98 Q4-20 /DD-49-3 680-6JI D-111 D32 D2-6 • Q2-11'//i /i//// / i / / / /t � i/r/ /fi �"j//i✓//i D area chromium plume IWC 11-41 Yr Tta JpP.r J,K�emm.e . fAwtom0 Vh9 HLMeq zaeHi4FexPnr Cl+Hmlya Rlwra j aiP.r.�VrPa sa,waa,n Feemis ®Cr.,pwL wa.wwh jl + ATde'T✓b F I1—n ]al--iNWL. " tlSRM Hnnlar Wain LJCr+. ,QaarH�ad-6W uyL ii�Tmrn.Lrn �Gsa9liq^'•P+1:i.JpH VpL r—� iLgie.Sin � G• 'A^q sre T+ee..a.^utlp .. i.CJY Wl noc306a yK a..ea.ry.rT -G�.;amrge ap-Slaa uJL CpwnW �tlr rf + a e` a c + Thy n �1 s .loo.pL. M TFo Jppw V,uonflmd , ���� Weeuaini Clromlun PHen A EniCwilriN ®G�, AaY�uticaJ,yL J IM'� VMi �Cw a..•n WL N•itla yL + M+id Tale OG�•Na'r9�L errJ<ypp �pn 9rPyfx Lore � tr e• ,,OHO Wti W ' 20pp rqe T -Pura ub TuJ bup.p -Ga.;HOH W+^i SHW uqH F¢Min Prw ml® v , Pa 100-DI+ .... D and H area chromium plumes IWC 11-41 V a K Area chromium plumes