The behavior, quantity, and location of undissolved gas in Tank 241-SY-101

Abstract

Mitigation of episodic flammable gas releases from Hanford Waste Tank 241-SY-101 was accomplished in July 1993 with the installation of a mixer pump that prevents gas retention. But it has not been possible until recently to measure the effects of mixing on the waste or how much gas remains and where it is located. Direct measurements of the void fraction and rheology of the mixed waste by the void fraction instrument provide estimates of the location, quantity, and behavior of undissolved gas in the tank. This report documents the compilation and integration of the information that enables this understanding. * and ball rheometer along with previous data iii V possible to measure properties in situ at specific points by using the ball rheometer and the VFI. When integrated, this information provides a framework from which the quantity and distribution of retained (undissolved) gas in Tank 241-SY-101 can be estimated. The results of this analysis, summarized in Table 4.4 of the report, indicate that the tank currently contains approximately 5,700 ft3 of gas in situ (3.8% average void), or about 7,800 ft3 when expanded to atmospheric pressure. In addition to the "best" estimate derived from in-tank temperature and property measurements, two other estimation methods were considered in this report. The first of these correlates atmospheric pressure and waste surface level to determine the waste compressibility. This compressibility, along with an assumed gas elevation, yields an estimate of the gas volume that is consistent with the best estimate described above, provided that the gas elevation is defmed properly. The second method, developed in support of the current mixer pump safety assessment, uses an extrapolation of retained gas estimates from historical data prior to pump installation to determine a gas-free waste level. The current gas inventory is interpolated between the gas-free state and that just prior to a large gas release event using waste surface level measurements. As proposed, this method gives significantly larger volumes than the best estimate, but it can be made consistent by correcting some of its assumptions. Direct measurements, similar to those of the VFI, are difficult, expensive, and available. from only a few locations in the tank. But indirect measurements, such as the response of the tank waste level to barometric pressure changes, provide little information about the distribution of the gas and, in fact, require some knowledge of the distribution of the gas in order to provide an estimate of retained gas quantity. To help address this problem, a set of dynamic bubble/particle models has been developed that provide insight into the state, amount, and distribution of retained gas in Tank 241-SY-101 as it responds to the mixer pump and barometric pressure variations. The models predict average amounts of retained gas in various layers of the tank-crust, convective layer, and loosely-settled layer that are compared with the VFI measurements to calibrate the model. While the models have no...

In situ rheology and gas volume in Hanford double-shell waste tanks

“…The void required to float the crust in SY-101 was found to be 25 & 8% (Brewster et al 1995) consideriqg the estimated porosity and density of the crust material. This void estimate is assumed to apply to the crust in other tanks as well.…”

“…They have performed as designed in all tests and had no significant problems. The initial testing program in SY-101 and SY-103 is documented in Stewart et al (1995), Brewster et al (1995), and Shepard et al (1995).…”

“…The uncertiinty in the temperature read by a thermocouple, whether on an MLT or thermocouple tree, is estimated at k 13°C (3°F) (Brewster et al 1995). However, experience with SY-101 data (Stewart et al 1994 has shown that MIT thermocouples are easily able to register temperature changes of 0.06"C (0.1"F).…”

“…Precision in sensing level changes is far more important.than absolute accuracy, however. A study of SY-101 data (Brewster et al 1995) concluded that the Enraf correlated most closely with gas releases and was the most precise level instrument. It records level in 0.025-cm (0.01-in.)…”

“…If the surface tension pressure were significanq the VFI would under-predict the void fraction; however, calculations show that the pressure added by surface tension is significant only for bubbles less than one micron in diameter. Studies 'of gas release signatures and photomicrographs of core samples from SY-101 (Brewster et al 1995) indicate that the volume-average bubble size is on the order of a few hundred microns. Tanks that are not forcibly mixed (all except SY-101) would be expected to have larger.…”