Test Materials and Description The sludge used for the gas generation testing was taken from the KE Basin floor and fuel canisters in March and April 1999 by Duke Engineering & Services Hanford. A consolidated sampling technique was employed for collecting the material (i.e., sludge from several locations was combined to form "consolidated samples"). Three sludge samples were used: fuel canister sludge (KC-2/3); floor sludge collected from between slotted fuel canisters containing highly damaged fuel (KC-4); and floor sludge collected away from fuel canisters and away from areas known to contain high concentrations of organic ion exchange resin (KC-5). The canister sludge used in this testing (KC-2/3) was prepared by combining two consolidated sludge samples (i.e., KC-2, collected from canisters containing highly damaged fuel, and KC-3, collected from canisters containing moderately damaged fuel). Portions of these samples were sieved to separate particles greater than or "plus" 250 µm (P250) from particles less than or "minus" 250 µm (M250). This separation was made to mimic the separation operations that are planned for the retrieval of certain K Basin sludge types and to gain a better understanding of how uranium metal is distributed in the sludge. [The separation point for certain K Basin sludge types was subsequently changed from 250 µm to 500 µm (Pearce and Klimper 2000).] Fine uranium metal particles have a high surface area and will react rapidly. Larger uranium metal particles will react at a slower rate, since their surface area per unit mass of sludge is lower. For the testing described here, sludge samples were placed into four large-scale vessels (850 ml) and eight small-scale reaction vessels (30 ml). The gas pressure in the vessels was monitored continuously, and gas samples were collected intermittently and analyzed via mass spectrometry. The large-scale testing was initiated in August 1999, and three tests continued through June 2000. One large-scale test is still underway, and will continue into FY2001. These test vessels contained 70 to 440 grams of settled sludge held at ambient hot cell temperature (~32ºC). The large-scale-test conditions are expected to be prototypical of those that will be experienced during long-term storage of the sludge at T Plant (i.e., largescale tests serve as a mock-up for prolonged T Plant storage). The small-scale tests, which were conducted with about 15 grams settled sludge each, were initiated on October 6, 1999, and completed in June 2000. The small-scale tests were conducted at elevated temperatures (six tests at 80ºC, one at 60ºC, and one at 40ºC) to accelerate the reactions and provide conclusive gas generation data within a reasonable testing period. The temperature in all of the small-scale test vessels was increased to 95°C for a period at the end of the tests to force completion of the reactions. The table below shows the test matrix. Experimental Matrix for Large-and Small-Scale Gas Generation Tests KC-2/3 KC-4 KC-5 Temperature and Test Scale M250 µm P2...

Executive SummaryFive bench-scale tests (~50 ml each) were conducted in sealed, un-agitated reaction vessels using sludge samples collected from the K East Basin and particles/coupons of irradiated N Reactor fuel (also from the K Basins). These tests were designed to evaluate and understand the chemical changes that may be occurring under the hydrothermal conditions (e.g., 7 to 72 h at 185°C) of the Sludge Treatment Project (STP) corrosion process and the effects that any changes would have on sludge rheological properties. The two sludge formulations used in these tests were selected to represent nominal (vs. bounding) compositions of the two major sludge feed streams to the STP process. The scoping tests were not designed to evaluate engineering aspects of the process.The hydrothermal treatment affected the chemical and physical properties of the sludge. In each test, significant uranium compound phase changes were identified, resulting from dehydration and chemical oxidation and reduction reactions. Physical properties of the sludge were significantly altered from their initial, as-settled, sludge values including, shear strength, settled density, weight percent water, and gas retention.The high uranium content sludge (~70 wt% uranium) set up to form a very stiff solid that exhibited very high shear strength (120,000 to 170,000 Pa) after hydrothermal treatment at 185°C for 7 to 10 hours under static (unstirred) conditions. Shear strengths of untreated sludge range from about 270 to 8100 Pa. Also, the hydrothermal treatment reduced the water content in the settled sludge by about 20 wt%. The treated sludge was difficult to remove from the Teflon test vessels, with sludge firmly adhering to the Teflon vessel surfaces. The strength of the treated sludge was further evaluated by agitation in water. In a 600-ml beaker, with 400 ml water at ~30°C, the diameters of agglomerates were reduced by about 40 to 50% after 1 hour of agitation with a 5.08-cm (2-in.) diameter impeller rotated at a tip speed of 80 cm/s.Further static tests with lower uranium-content sludges (~16 wt% uranium) run 72 h at 185°C produced softer solids with 9,000 to 16,000 Pa shear strengths. While some sludge adhered to the Teflon liner, it was not tenaciously bound. The agglomerates from these tests were relatively weak and, in some cases, were disintegrated with a gentle stream of water. The water content in the settled sludge was not significantly affected by the hydrothermal treatment.Chemical phase alteration, observed by X-ray diffraction, scanning electron microscopy, and energy dispersive spectrometry, gave evidence that solids dissolution followed by precipitation was responsible for the increased strengths of the sludge products. The presence of organic ion exchange resin (OIER) and polymer flocculent (constituents known to be in K Basin sludge) in the lower uranium-content sludge did not appear to have a significant impact on the physical behavior of the post-treated sludge.Four irradiated uranium metal fuel coupons were included in on...

metal in simulated sludge, in the present testing, are similar to, and thus confirm, those in prior tests of irradiated uranium metal fuel particles from K Basins. The following table summarizes the hydrogen generation rates from uranium metal corrosion at 60°C for the grout formulations and reference tests. The rates are for hypothetical RH-TRU 55-gallon drums loaded with 14.6 liters of KW canister sludge in the respective reference sludge or grouted sludge waste form. This amount of sludge is at the 200-gram 239 Pu fissile gram equivalent (FGE) limit mandated by WIPP. The hydrogen generation rates at the FGE-limited loadings are compared with the maximum 3.65×10-8 moles/second hydrogen generation rate tolerated in a WIPP RH-TRU drum. Case Drum Hydrogen Generation Rate (moles H 2 /sec) (a) Factor Improvement Needed for WIPP Reference Tests Uranium Metal 1.80×10-5 493 Uranium Metal in Simulated Sludge 5.71×10-6 156 Portland Cement BNFL 8.57×10-6 235 Bentonite 4.00×10-6 110 Weakley 9.43×10-6 258 Cast Stone 7.37×10-6 202 Magnesium Phosphate Cement Tectonite 1.71×10-5 470 Tectonite-Bentonite 7.57×10-6 207 Drum Hydrogen Generation Rate Limit 3.65×10-8 1 (a) Drum H 2 generation rate based on the uranium metal surface area in 14.6 liters of nominal KW canister sludge drum TRU RH

This report was originally published in March 2001. In January 2004, a transcription error was discovered in the value reported for the uranium metal content of KE North Loadout Pit sample FE-3. This revision of the report corrects the U metal content of FE-3 from 0.0013 wt% to 0.013 wt%. The text also has been revised to more accurately describe the expected uranium metal reactions being measured and evaluated. Experimental Matrix for Series II Gas Generation Tests and Uranium Metal Estimates Sludge Sample Fraction Test Temp., °C Estimated Uranium Metal Content Test Description Sample ID Settled Sludge, g Start ~300 hr Final ~650 hr Wt % U Metal (Settled Sludge Basis) Estimation Technique KE Canister Sludge (Consolidated Sampler Used in Single-Pull Mode) Unmixed KC-1 KC-1 (whole) 23.8 60 95 0.065 Fission Product Gas Mixed KC-1 KC-1 (whole) 25.6 60 95 0.56 Fission Product Gas U Metal in KC-1 Plus 500 µm 14.2 80 95 3.7 Fission Product Gas U Metal in KC-1 Minus 500 µm 22.3 80 95 0.14 Fission Product Gas Single-Pull Core KE Floor and Pit Sludge Samples Main Floor (FE-1) FE-1 (whole) 27.0 90 95 0.022 Fission Product Gas North Loadout Pit (FE-3) FE-3 (whole) 21.2 90 95 0.013 Corrosion Gas Weasel Pit (FE-5) FE-5 (whole) 20.7 90 95 0.027 Fission Product Gas Dummy Elevator and Tech View Pit (FE-4 + FE-6) FE-4 (63%) + FE-6 (37%) 16.1 (10.2/5.9) 90 95 0.0052 Fission Product Gas Other Tests Organic Ion Exchange Resin Sludge KC-6 (whole) 19.1 90 95 0.026 Corrosion Gas 1996 KE Canister Sludge 96-06 19.5 80 95 0.90 Fission Product Gas v gas generation testing. The technique used to estimate uranium metal content (corrosion gas generation or fission product gas release by the corrosion) is also noted in the table. The uranium metal content is estimated based on fission product gas release, where available, because of its higher reliability, especially at low concentrations. Gas Generation Testing Gas production observed in the Series II tests with KE Basin sludge was generally lower than observed in similar Series I studies of KE Basin canister and floor sludges (Delegard et al. 2000). The lower gas generation rates can partially be attributed to the longer storage periods at hot cell temperatures experienced by the Series II sludge samples. All tests in Series II (conducted with sludge from samples KC-1, FE-1, FE-3, FE-5, FE-4/6, KC-6, and 96-06) produced CO 2 as the predominant gas product with lower concentrations of H 2 , except the KC-1 P500 test, which produced predominantly H 2 gas with lower CO 2 concentrations. Other gases released or produced were Xe, Kr, and hydrocarbons (methane, ethane, and higher hydrocarbons). Trace amounts of O 2 and N 2 present from atmospheric contamination also were consumed in each test. Similar gas generation and consumption was observed in the Series I tests and in gas collection tests from KE canister sludge (Makenas et al. 1997). The release of CO 2 in these tests may be due to the reaction of schoepite with calcite to form becquerelite and CO 2 , with some (but not all) of the CO 2 also...

No abstract