Effluent Treatment Facility Waste Stream Stabilization Testing

Transcription

1 Effluent Treatment Facility Waste Stream Stabilization Testing Michael R. Silsbee, 1 Marisol Avila, 2 Boyd A. Clark, 1 Gary A. Cooke, 3 Mike D. Guthrie, 4 Gary L. Koci, 5 Richard J. Lee, 1 Larry L. Lockrem, 3 and Kristi J. Lueck 4 Abstract The U.S. Department of Energy Hanford Site, the location of plutonium production for the U.S. nuclear weapons program, is the focal point of a broad range of waste remediation efforts. This presentation will describe the development of cementitious waste forms for evaporated Hanford waste waters from several sources. Basin 42 waste water and simulants of proposed Waste Treatment and Immobilization Plant secondary wastes and Demonstration Bulk Vitrification System secondary wastes were solidified in cementitious matrices termed dry cementitious formulation. Solidification of these brines was difficult to deal with because of high sulfate contents. Two approaches were explored. The first was based on compositions similar to sulphoaluminate-belite cements. The main component of these cements is 4CaO 2Al 2 O 3 SO 4. When hydrating in the presence of sulfate, these cements rapidly form ettringite. The goal was to consume the sulfate by rapidly forming ettringite. Forming ettringite before the mixture has fully set minimizes the potential for deleterious expansion at a later date. These formulations were developed based on mixtures of calcium-aluminate cement, a glassy blast-furnace slag, class F fly ash, and Portland cement. A second approach was based on using high alumina cement like ciment fondu. In this case the grout was a mixture of ciment fondu, a glassy blast-furnace slag, class f fly ash, and Portland cement. The literature shows that for concretes based on equal amounts of ciment fondu and blast furnace slag, cured at either 2 C or 38 C, the compressive strength increased continuously over a period of 1 year. In this second approach, enough reactive calcium aluminate was added to fully consume the sulfate at an early age. The results of this study will be presented. Included will be results for expansion and bleed water testing, adiabatic temperature rise, microstructure development, and the phase chemistry of the hydrated materials. The results of toxicity characteristic leaching procedure testing on samples of dry cementitious formulation spiked with selected metals will also be presented. 1 The RJ Lee Group, Inc., Monroeville, Pennsylvania. 2 Center for Laboratory Sciences, Pasco, Washington. 3 CH2M HILL Hanford Group, Inc., Richland, Washington. 4 Fluor Hanford, Inc., Richland, Washington. 5 Fluor Federal Services, Richland, Washington.

2 Objective This study was designed to provide waste stabilization make-up parameters for use in the design of a chemically bounded cement-type waste stabilization process. Technical support provided included mixing, testing, and reporting of values for a composite solid waste form. The goal was to develop and test a cementitious material formulation to stabilize simulated aqueous wastes. The simulated wastes used have compositions similar to three waste streams: (A) concentrated Basin 42 waste water, (B) Waste Treatment and Immobilization Plant (WTP) secondary waste and (C) Demonstration Bulk Vitrification System (DBVS) secondary waste. Testing of small-scale casting of the solid waste form included compressive strength, expansivity, bleed water, and leachability of selected metals. Introduction The Effluent Treatment Facility (ETF) in the 2E Area of Hanford is considering converting several liquid waste streams from evaporator operations into a solid cementitious material waste form. The cementitous material/waste mixture will be poured into monoliths and sealed for land disposal at the Hanford Site. The ETF sought assistance in the development, mixing, and testing of a solid waste form composed of cementitious material mixed with several cold simulant wastes. Laboratory testing of cementitious material/simulant mixtures was required to demonstrate the viability of cementitious material to stabilize Basin 42, WTP secondary stream, and DBVS secondary streams. The resulting waste form will be solidified in monoliths. Additional testing of radiologically contaminated actual Basin 42 waste samples was done at the 222-S Laboratory to confirm the conclusion reached during the simulant testing. Preparation of Simulant Mixtures The simulant mixtures used are listed in Tables 1 through 3. In the case of the Basin 42 simulant the resulting mixture was a slurry with a significant amount of undissolved CaSO 4 (probably as the mineral, gypsum). The WTP brine simulant was prepared at an elevated temperature to dissolve the components and some precipitate formed on cooling. The simulants were prepared at two solids salt concentrations, 25% and 4% total solids by weight (or the highest practical level). It was not possible to make the DBVS simulant at the 4% level. For the final round of simulant testing, the simulants were spiked as follows: Arsenic 5 mg/l Lead 5 mg/l Barium 1, mg/l a Mercury 2 mg/l Cadmium 1 mg/l Selenium 1 mg/l Chromium 5 mg/l Silver 5 mg/l a May need to be optional due to levels.

3 Table 1. Basin 42 Simulant. Compound 25% Total Solids (gm/kg) 4% Total Solids (gm/kg) CaOH Na 2 SO NaOH MgCl Mg(NO 3 ) NaCl.. NaNO (NH 4 ) 2 SO NH 4 OH KOH Water 75 6 Table 2. WTP Brine Simulant. 25% Total Solids Compound (gm/kg) 4% Total Solids (gm/kg) NaCl.3.5 Na 2 CO NaNO NaNO Na 3 PO Na 2 SO NaOH (NH 4 ) 2 SO Water 75 6 Table 3. Bulk Vitrification Scrubber Simulant. 25% Total Solids (gm/kg) 4% Total Solids (gm/kg) Compound NaOH Na 2 SO NaCl NaNO Na 2 CO Water 75 6 NaOH Na 2 SO

4 Selection of Cementitious Material Formulations The first trial cementitious material formulations were based on the concept of adding just enough calcium aluminate cement (SECAR 6 51) to react with the sulfate in solution to form ettringite (Ca 6 Al 2 (SO 4 ) 3 (OH) 12-26H 2 O). This series of formulations were based on the concept of simulating the compound 4CaO 2Al 2 O 3 SO 4. 4CaO 2Al 2 O 3 SO 4 is the main phase in a family of cements known as sulphoaluminate cements. This second series utilized Ciment fondu as a replacement to the normal cementitious material components to arrive at the desired chemistry. The balance of the solid fraction was modified cementitious material dry reagent formulation (DRF) (i.e., a cement, fly ash, slag mix). The simulants were warmed to 1 F prior to mixing. After mixing, the trial mixtures were checked at 24 and 48 hours to measure the amount of bleed water and relative hardness of the mixtures noted. The formulations were adjusted if necessary and another round of mixtures prepared. The iterative process continued until mixtures promising enough to be cast for 28-day specimens were identified. Table 4 shows the formulations selected for further study. Table 4. Formulations Selected for Further Study. Simulant DBVS WTP Basin 42 Solids in simulant (wt%) Brine (wt%) Secar 51 (wt%) DRF (wt%) Brine/solid Compression Test ASTM C 39 Final compression tests were performed on each formulation selected after mixing and curing. Compression test specimens were evaluated after the initial mixing trial; only mix formulations identified as viable candidates were tested. The relevant test method is ASTM C 39, Standard Test Method for Compressive Strength of Cylindrical Concrete Specimens, with modifications for sample size. Samples of 2-in. diameter x 4-in. length were tested. The formulations tested are the same as those used for the Toxicity Characteristic Leaching Procedure (TCLP) [U.S. Environmental Protection Agency (EPA) SW-846, Method 1311] testing and adiabatic curing. The results of the compressive strength testing are shown in Figures 1 through 3. Densities calculated by measuring and weighting the cylinders are shown in Figures 4 and 5. Figure 6 shows a sample made using Basin 42 simulants. Figures 7 and 8 are load 6 Secar is a registered trademark of Lafarge Calcium Aluminates, Inc., Chesapeake, Virginia.

5 deflection curves for Basin 42 samples made from 25 and 4 wt% curves, respectively. The crumbly appearance of the 25 wt% samples is reflected in their behavior during loading. The loading curve indicates a soft material that is slowly compressing as opposed to the brittle breakage shown in the 4 wt% sample. Figure 1. Compressive Strengths in Samples Prepared using Basin 42 Simulants and Cured for 1 Days. (The strengths given are the average of the strength determined for three 2-in.- diameter x 4-in.-length cylinders.) Compressive Strength (PSI) wt.% 4 wt. % Brine to Solids Ratio Figure 2. Compressive Strengths in Samples Prepared using Basin 42 Simulants and Cured for 28 Days. (The strengths given are the average of the strength determined for three 2-in.-diameter x 4-in.-length cylinders.) 35 Compressive strength (PSI) wt.% 4 wt. % Brine to solids ratio

6 Figure 3. Compressive Strengths in Samples Prepared using WTP and DBVS Simulants and Cured for 28 Days. (The strengths given are the average of the strength determined for three 2-in.-diameter x 4-in.-length cylinders.) Compressive strength (PSI) % WTP (B/S =.9) 25% WTP (B/S=.9) 25% DVBS (B/S =.9) Figure 4. Density for Samples Prepared using Basin 42 Simulants and Cured for 28 Days. Density Brine to solids ratio 25 wt.% 4 wt. %

7 Figure 5. Density for Samples Prepared using WTP and DBVS Simulants and Cured for 28 Days. Density % WTP (B/S =.9) 25% WTP (B/S=.9) 25% DVBS (B/S =.9) Figure 6. Sample of 1-Day-Old Basin Wt% waste form (Brine/S olids =.8) shown as removed from mold. (The top and bottom crumbled when removed. Nodules and blemishes were found on surface.)

8 Figure 7. Illustration of Load vs. Time Curve for Sample Basin 42 25wt% (Brine/S olids =.8) Maximum Load = 218 pounds Load (lbs) Recorder s Time (seconds) Figure 8. Illustration of Load vs. Time Curve for Sample of Basin 42 4 wt% (Brine/S olids =.6) Maximum Load = 6,123 Pounds Load (lbs) 5 Load (lbs) :1 : :1 :2 :3 :4 :5 1: Recorder min:s Time (min:secs) Evaluation of Adiabatic Effects Adiabatic curing simulates the temperature rise expected during curing of full-scale monoliths. This task evaluates the potential for the development of excessive temperatures during curing. One of the most promising mixtures from each of the simulant formulations was selected for evaluation of the effect of adiabatic curing on the solidified product. The formulations tested are the same as those used for strength testing and TCLP testing. Curing the mixtures adiabatically simulates the temperature profile expected when cast in large scale monoliths. After curing the solid formulations were

9 sampled and characterized. This procedure allowed a qualitative characterization of the effect of adiabatic curing on the expansion behavior of the mixtures to be made. Figure 9 shows the experimental setup used to simulate the environment that the waste forms will see when cast into larger monoliths. The waste form is cast into a stainlesssteel thermos and then placed into a water bath. A thermocouple inserted into the center of the thermos allows the temperature rise occurring in the waste form to be monitored. As the temperature of the waste form rises, the temperature of the water bath is increased at a matching rate. This mimics the environment that will be found in the center of a large monolith of the waste form. Figure 9. Experimental Setup for Adiabatic Temperature Cure. Illustrated is the sample thermal vessel and thermocouples. The vessel has an approximate 2-in. 3 volume for the simulated grout-waste mix. 7 Figures 1 through 12 show the temperature rise that occurred in three waste forms. The Basin 42 4 wt% waste form hydrated slowly, only starting to gain temperature after several hours. The maximum temperature was not reached for 4+ hours. The WTP 4 wt% sample hydrated much faster. The maximum temperature was reached at approximately 1 hours. The DBVS 25 wt% sample also showed a several-hour induction period. The waste form reached a maximum temperature at ~24 hours. In all cases the maximum temperature reached was near 7?C. EPA SW-846, Method 1311 The EPA SW-846 Method 1311 test is the TCLP from a size-reduced sample. TCLP testing was performed on selected formulations after 28 days of curing. The formulations tested are the same as those used for strength testing and adiabatic curing. The performance of the test is based on land disposal restrictions mandated by both Washington State and the Code of Federal Regulations. In this testing, the samples were crushed to <3/8 in. Preliminary testing indicated that the samples should be leached using glacial acetic acid (extraction fluid 2) as the extraction solution. 7 LAUDA is a registered trademark of Dr. R. Wobser GmbH & Co. KG Ltd. FR Germany.

10 Figure 1. Temperature Rise Observed During Adiabatic Curing of Basin 42 4 wt% Waste Form. Time vs. Temperature Adiabatic Test Temperature, C Adiabatic Grout Water Bath Time, hours Figure 11. Temperature Rise Observed During Adiabatic Curing of WTP 4 Wt% Waste Form Time vs. Temperature Adiabatic Test 2 Adiabatic Grout Water Bath Time (hours) Figure 12. Temperature Rise Observed During Adiabatic Curing of DBVS 25 Wt% Waste Form. Temperature (C) Time vs. Temperature, Adiabatic Test 3 Water Bath Adiabatic Grout Time (hours)

11 The EPA SW-846 test determines the characteristic leaching of metals from a crushed (-3/8 in.) sample. The leach solution was analyzed for antimony, arsenic, barium, beryllium, cadmium, chromium (total), lead, mercury, nickel, selenium, silver, thallium, and zinc. The acceptance levels for inorganic elements are from Washington State Department of Ecology. Table 5 shows the result of the leach testing. Bleed Water and Expansion Test ASTM 94 98a The bleed water and expansion test measures bleed water and expansion of the cementitious material during the process of hydration for a period of 24 hours. The relevant test method is ASTM C94 98a, Standard Test Method for Expansion and Bleeding of Freshly Mixed Grouts for Preplaced-Aggregate Concrete in the Laboratory, with modifications to the size and shape of the test cylinder to allow these castings to be recovered for other tests. Excessive bleed water (>5% by depth at a ph of 9 or greater) may imply excessive free liquid at the end of the 28-day curing period. For cementitious material processing this could require the use of adsorbents either in the cementitious material, on the surface of the cementitious material in the storage container, or removal and treatment of the free water as a waste product. Figure 13 contains the results of the expansion and bleed-water testing. In all cases the amount of bleed water was <2% and in most cases it was zero. The waste forms typically showed either zero expansions or a slight shrinkage. Crystal and Microstructure Analysis The three cast samples (Basin 42 4 wt%, WTP 4 wt% and DBVS 25 wt %) were examined by X-ray diffraction (XRD) (powder) analysis and scanning electron microscopy/energy dispersive spectroscopy (SEM/EDS) analysis. The crystalline phases for each of the three samples are listed in Table 6. The SEM/EDS analysis in most cases confirmed the presence of the phases identified by XRD (Figures 14 through 16) and also detected other (noncrystalline) phases. In the Basin 42 samples, ettringite is the major hydration product. The ettringite formed from the reaction between gypsum in the brine and the monocalcium aluminate in the SECAR 51. In the WTP samples a sodium-rich analog of ettringite was the principal hydration product. The DBVS samples produced hydration products that were largely amorphous in character. Additionally, a white layer was observed on the surface of the WTP 4 wt% sample after the sample had been exposed to air. This white lay er was determined by XRD to be composed of Na 2 SO 4 (two different crystalline phases) and ettringite (Table 7). The XRD results were verified by SEM/EDS analysis.

12 Table 5. Results of TCLP Testing. Basin 42 (mg/l) WTP (mg/l) DBVS (mg/l) (mg/l) EPA a (mg/l) 25 wt% Total Solids 4 wt% Total Solids 25 wt% Total Solids 4 wt% Total Solids 25 wt% Total Solids Beryllium < <.2.72 <.2 Spikes (mg/l) Chromium < Nickel < Zinc < Arsenic < Selenium < Silver < Cadmium < Antimony < Barium < , Mercury < Thallium < <.6 <.6 <.6 Lead < <.2 < Uranium Cesium a Title 4, Code of Federal Regulations, Part , Universal Treatment Standards.

13 Figure 13. Bleed Water and Expansion Test..4.2 Combined Expansion % Time (min.) 25 wt% 25wt% 4wt% Basin 42 4wt% Basin 42 25wt% Table 6. Composition of the Crystalline Part of the Cast Samples. Chemical Formula Compound Name Basin 42 (4 wt%) Ca 6 Al 2 (SO 4 ) 3 (OH) 12 26H 2 O Ettringite Major Ca 2 Al(Al,Si)O 7 Gehlenite Minor SiO 2 Quartz Minor Ca(SO 4 ) (H 2 O) 2 Gypsum Trace Al(OH) 3 Gibbsite Trace BaSO 4 Barium sulfate (barite) Trace WTP (4 wt%) NaCa 4 Al 2 O 6 (SO 4 ) H 2 O Sodium calcium aluminum sulfate Major hydrate Ca 6 Al 2 (SO 4 ) 3 (OH) 12 26H 2 O Ettringite Minor Na 2 SO 4 Thenardite Minor Ca 2 Al(AlSi)O 7 Gehlenite Minor Ca(CO 3 ) Calcite Trace DBVS (25 wt%) SiO 2 Quartz Major CaCO 3 Calcite Major Ca(SO 4 )(H 2 O) 2 Gypsum Trace

14 Table 7. Composition of the White Crystals Growing in the WTP 4 wt% Cast Sample. Chemical Formula Compound Name WTP (4 wt% - white crystals on surface) Na 2 SO 4 Thenardite Major Na 2 SO 4 Sodium sulfate Minor Ca 6 Al 2 (SO 4 ) 3 (OH) 12 26H 2 O Ettringite Trace Figure 14. Selected SEM Images from the Basin 42 4 Wt% Sample: (A) Ettringite, (B) Gypsum, and (C) Barium Sulfate. A B C

15 Figure 15. Selected SEM Images from the WTP 4 Wt% Sample: (A) Na/Al/Si/Ca Sulfate Phase and (B) Ettringite. A B Figure 16. Selected SEM Images from the DBVS 25 Wt% Sample: (A) Na/Al/Si/Ca Particle and (B) a Silicon O xide Particle. A B

16 Summary When prepared, the simulants all had a syrupy texture. It was not possible to prepare a DBVS simulant at a loading higher than 25 wt%. The Basin 42 simulant immediately precipitated a white solid (gypsum). Precipitation occurred in other simulants with aging. In general, all the formulations exhibited an optimum mixing when using a brines/solids ratio in the.8-.9 range. In many cases a small amount of bleed water was noticed shortly after mixing. The bleed water was typically reabsorbed within a few hours. In all cases the brines with the higher loading gave higher strengths. These formulations were selected for further testing. None of the formulations tested showed excessive bleed water. The highest bleed water found was 1.36%. All of the formulations tested showed zero expansions and a slight shrinkage. All of the formulations tested reached temperatures higher than 7 C. This is in the temperature range at which ettringite could be expected to start to break down. If this occurs there is potential for delayed ettringite formation that could lead to cracking and deterioration of the physical properties at later ages. All of the formulations tested exceeded the 4 CFR limits for chromium. This is likely due to Cr 6+ being used as the spike. Others (Hoang 21) have noted that the only way to stabilize Cr 6+ containing wastes is to pretreat the waste to reduce the chrome to Cr 3+. Basin 42 formulations also failed to meet the 4 CFR limit for cadmium. In all cases the brines were spiked at levels of at least 1 times the 4 CFR levels. In some cases the spikes were more that 8 times the levels. Even with the high spike levels, all of the grouts met the Washington Administrative Code (WAC) , Dangerous Waste Regulations. References 4 CFR , Universal Treatment Standards, Code of Federal Regulations, as amended. ASTM C 39/C 39M -99, 1999, Standard Test Method for Compressive Strength of Cylindrical Concrete Specimens, American Society for Testing and Materials, Easton, Maryland. ASTM C 94-98a, 1998, Standard Test Method for Expansion and Bleeding of Freshly Mixed Grouts for Preplaced-Aggregate Concrete in the Laboratory, American Society for Testing and Materials, Easton, M aryland. Hoang, T. T., 21, Testing Stabilization/Solidification Processes for Mixed Waste, WM 1 Conference, February 25-March 1, 21, Tucson, Arizona. WAC , Dangerous Waste Regulations, Washington Administrative Code, as amended.