DATA SUMMARY REPORT SMALL SCALE MELTER TESTING OF HLW ALGORITHM GLASSES MATRIX1 TESTS VSL-07S1220-1 REV 0 7/25/07

Eight tests using different HLW feeds were conducted on the DM100-BL to determine the effect of variations in glass properties and feed composition on processing rates and melter conditions (off-gas characteristics, glass processing, foaming, cold cap, etc.) at constant bubbling rate. In over seven hundred hours of testing, the property extremes of glass viscosity, electrical conductivity, and T{sub 1%}, as well as minimum and maximum concentrations of several major and minor glass components were evaluated using glass compositions that have been tested previously at the crucible scale. Other parameters evaluated with respect to glass processing properties were +/-15% batching errors in the addition of glass forming chemicals (GFCs) to the feed, and variation in the sources of boron and sodium used in the GFCs. Tests evaluating batching errors and GFC source employed variations on the HLW98-86 formulation (a glass composition formulated for HLW C-106/AY-102 waste and processed in several previous melter tests) in order to best isolate the effect of each test variable. These tests are outlined in a Test Plan that was prepared in response to the Test Specification for this work. The present report provides summary level data for all of the tests in the first test matrix (Matrix 1) in the Test Plan. Summary results from the remaining tests, investigating minimum and maximum concentrations of major and minor glass components employing variations on the HLW98-86 formulation and glasses generated by the HLW glass formulation algorithm, will be reported separately after those tests are completed. The test data summarized herein include glass production rates, the type and amount of feed used, a variety of measured melter parameters including temperatures and electrode power, feed sample analysis, measured glass properties, and gaseous emissions rates. More detailed information and analysis from the melter tests with complete emission chemistry, glass durability, and melter operating details will be provided in the final report. A summary of the tests that were conducted is provided in Table 1. Each of the seven tests was of nominally one hundred hours in duration. Test B was conducted in two equal segments: the first with nominal additives, and the second with the replacement of borax with a mixture of boric acid and soda ash to determine the effect of alternative OPC sources on production rates and processing characteristics. Interestingly, sugar additions were required near mid points of Tests W and Z to reduce excessive foaming that severely limited feed processing rates. The sugar additions were very effective in recovering manageable processing conditions, albeit over the relatively short remainder of the test duration. Tests W and Z employed the highest melt viscosities but not by a particularly wide margin. Other tests, which did not exhibit such foaming Issues, employed higher concentrations of manganese or iron or both. These results highlight the need for the development of protocols for the a priori determination of which HLW feeds will require sugar additions and the appropriate amounts of sugar to be added in order to control foaming (and maintain throughput) without over-reduction of the melt (which could lead to molten metal formation). In total, over 8,800 kg of feed was processed to produce over 3200 kg of glass. Steady-state processing rates were achieved, and no secondary sulfate phases were observed during any of the tests. Analysis was performed on samples of the glass product taken throughout the tests to verify composition and properties. Sampling and analysis was also performed on melter exhaust to determine the effect of the feed and glass changes on melter emissions