A spare mixing pump from the Hanford Grout Program was installed in Hanford double-shell waste Tank 241-SY-101 on July 3, 1993, after being modified to take advantage of waste stratification. It was anticipated that pump mixing would prevent large episodic flammable gas releases that had been occurring about every 100-150 days. A cautious initial test plan, called Phase A, was run to find how the pump and tank would behave in response to very brief and gentle pump operation. No large gas releases were triggered, and the pump performed well except for two incidents of nozzle plugging. On October 21, 1993, the next test series, Phase B, began, and the pump was applied more aggressively to mix the tank contents and mitigate uncontrolled gas releases. Orienting the pump in new directions , released large volumes of gas and reduced the waste level to a near-record low. Results of the entire period from pump installation to the end of Phase B on December 17, 1993, are presented in detail in this document. Though long-term effects require furtherevaluation, we conclude from these data that v the jet mixer pump is an effective means of controlling flammable gas release and that it has met the success criteria for mitigation in this tank. iii SZrMMARY A mixer pump was found to be effective in controlling and possibly eliminating large episodic flammable gas releases from Hanford Tank 241-SY-I01. A gas release event (GRE) is apparently initiated when a major portion of the gas-bearing sludge layer accumulates sufficient gas to become buoyant, pull flee from the surrounding material, rise through the surface crust, and release the trapped gas into the dome space. Mixer pump operation is intended to keep enough of the gas-generating material in suspension so that gas is released continuously instead of periodically in large, potentially dangerous GREs.. A spare mixing pump from the HartfordGrout Program was modified and installed in the tank on July 3, 1993, seven days after a typical GRE that met the safety criteria for pump installation. The modifications were made to take advantage of the density stratification in the waste, providing vertical v buoyant motion as well as high jet velocity to promote mixing. The initial pump operations in Phase A testing were extremely gentle, beginning with a series of daily pump 'bumps' intended to keep the pump nozzles clear. Because nozzle plugging did occur, bump speed and duration were increased, eventually arriving at the accepted five-minute period at 1000 rpm on July 26. There has been no nozzle plugging since then. Bumping was initially performed twice a day through mid-August and once a day through the start of Phase B testing October 17. By the end of Phase B, thrice-weekly bumping during nontesting periods became the rule. ACKNOWLEDGMENTS The authors wish to thank Jeanne Lecher for supplying plots of various data under high pressure and short notice, to our editor Sheila Bennett for working extra hours to put all the pieces together, and to Kathy Hildebrant for helping with th...

The Full-Scale Mixer Pump Test Program was performed in Hanford Tank 241-SY-101 from February 4 to April 13, 1994, to confirm the long-term operational strategy for flammable gas mitigation and to demonstrate that mixing can control the gas release and waste level. Since its installation on July 3, 1993, the current pump, operating only a few hours per week, has proved capable of mixing the waste sufficiently to release gas continuously instead of in large episodic events. The results of Full-Scale Testing demonstrated that the pump can control gas release and waste level for long-term mitigation, and the four test sequences formed the basis for the long-term operating schedule. The last test sequence, jet penetration tests, showed that the current pump jet creates flow near the tank wall and that it can excavate portions of the bottom sludge layer if run at maximum power. Pump mixing has altered the 'normal' configuration of the waste; most of the original nonconvective sludge has been mixed wi_ the supernatant liquid into a mobile convective slurry that has since been maintained by gentle pump operation and does not readily return to sludge.

No abstract

No abstract

PrefaceThis report was prepared to satisfy the DOE-RL milestone T22-97-006, "Complete RGS Data Analysis for 4 of 4 Flammable Gas SSTs." This task supported the deployment of the Retained Gas Sampler (RGS) system in Flammable Gas Watch List Tanks.The emphasis of this report is on presenting the measurements resulting from retained gas sampling of Tanks U-103, S-106, BY-101, and BY-109, not on interpreting the effects of the findings on understanding tank behavior or on the safety issues. The retained gas sampling information is a direct measurement of the amount and composition of gas retained in these tanks. This information will be combined with information from other sources to develop a better understanding of the mechanisms for gas generation, retention, and release, which will lead to resolution of the Flammable Gas Safety Issue and provide a more detailed understanding of tank behavior to support retrieval operations. iii Executive SummaryThis report providesethe results obtained for the single-shell tanks (SSTs) sampled with Retained Gas Sampler (RGS) during 1997: Tanks 241-U-103,241-S-106,241-BY-101, and 241-BY-109. (Hereafter these tanks will be referred to without the "241-,I' following standard he practice.) The RGS is a modified version of the core sampler used at Hanford. It is aesigned specifically to be used in concert with the gas extraction equipment in the hot cell to capture and extrude a gas-containing waste sample in a hermetically sealed system. The retained gases are then extracted and stored in small gas canisters; The composition of the gases contained in the canisters is measured by mass spectroscopy. The total gas 'volume in the sample is obtahed from analyzing the extraction process, as discussed in detail throughout this report. The RGS has been confiied as a viable approach for measuring the composition of retained gases in SSTs at Hanford.The retained gas inventories calculated from the local measurements of gas volume fraction made by the RGS can differ significantly from the total gas inventories estimated by the barometric pressure effect (BPE) or surface level rise (SLR) methods. These discrepancies occur together with irregular waste layers and other strong indications of lateral inhomogeneity in the waste. Because the RGS samples are localized they capture little of this variation. Therefore, the BPE or SLR methods, which are related to the overall gas in the tank, must be used to supplement RGS measurements in estimating the gas inventories in high-wastevariability tanks.Nitrogen, an inert gas, was found to be a major component of the retained gas (more than 20 ~01%) in each of the tanks that were sampled.In two of the tanks sampled, the ratios of hydrogen to nitrous oxide were observed to be significantly higher than 1; i.e., these tanks are fuel-rich. However, U-103 samples contained as much as 52% nitrous oxide and as little as 14% hydrogen.Segment 2 of U-103 contained 44 vol% of gas under in-situ conditions. This measurement was confimed by a large gas volume shown in th...