Preliminary Analysis of Species Partitioning in the DWPF Melter

Abstract

The work described in this report is preliminary in nature since its goal was to demonstrate the feasibility of estimating the off-gas entrainment rates from the Defense Waste Processing Facility (DWPF) melter based on a simple mass balance using measured feed and glass pour stream compositions and timeaveraged melter operating data over the duration of one canister-filling cycle. The only case considered in this study involved the SB6 pour stream sample taken while Canister #3472 was being filled over a 20hour period on 12/20/2010, approximately three months after the bubblers were installed. The analytical results for that pour stream sample provided the necessary glass composition data for the mass balance calculations. To estimate the "matching" feed composition, which is not necessarily the same as that of the Melter Feed Tank (MFT) batch being fed at the time of pour stream sampling, a mixing model was developed involving three preceding MFT batches as well as the one being fed at that time based on the assumption of perfect mixing in the glass pool but with an induction period to account for the process delays involved in the calcination/fusion step in the cold cap and the melter turnover. Some of the key findings of this study include:  The proposed concept of estimating off-gas entrainment rates from measured feed and glass pour stream compositions appears feasible at least for the major species and thus additional case studies are warranted.  The overall entrainment rate from the bubbled DWPF melter was calculated to be 2.4% of the calcined solids fed, which is ~2X the design basis entrainment rate for the non-bubbled melter.  The average entrainment rate of the four major non-volatile sludge components, including Al, Fe, Mn and U, was calculated to be 2.7% fed, while that of the frit components, excluding Na, was 0.7%. It means that the entrainment rate of sludge is higher than that of frit by a ratio of 4:1, which is in qualitative agreement with the results of two off-gas deposit samples taken earlier.  Specifically, the calculated entrainment rates of Fe and Si were 0.6-0.7%, while those of Al, B, Mn, Th and U were higher, ranging from 1.9 to 5.6% with Mn at the highest.  The calculated entrainment rate of Na was 7.5%, of which a significant fraction could have been entrained via volatilization in the form of borates, sulfates, and halides.  The calculated entrainment rates of semi-volatile Cs and Cd ranged from 16 to 18%, while that of sulfur was estimated to be higher than 50%.  The calculated DWPF entrainment rates were in general higher than the DM1200 rates, which may be attributed in part to the formic acid used as the reductant for the DWPF feeds compared to sugar used for the DM1200 feeds, as demonstrated earlier in terms of increased pressure spikes with formic acid-based feeds. The much higher entrainment rates of Na and semi-volatiles in the DWPF melter could be reflective of actual operating conditions such as extended idling. The proposed concept will be tested further against ad...

Preliminary analysis of species partitioning in the DWPF melter. Sludge batch 7A

“…One drawback is that the calcined/fused solids exiting the cold cap is assumed to blend into the melt pool instantly, which is not feasible considering the high viscosity of glass melt, which is three orders of magnitude or higher than that of water. To overcome this difficulty, a full glass residence time in the melt pool was subtracted in the calculation of melter turnover by the last batch, 5 which may be excessive under the forced-convection mixing induced by the glass bubblers. The mixing delay is further examined in this study.…”

“…The feasibility of expanding the range of data collected in Stage 2 was assessed in an earlier study by including the DWPF melter data taken in a radioactive production environment during the processing of Sludge Batch 6 (SB6). 5 The scope of the available DWPF data is broad but, as is the case with most manufacturing plants, data collection is geared toward production support through process control and troubleshooting. For example, the DWPF melter is equipped with comprehensive pressure, temperature, and flow control loops but has no off-gas sampling or monitoring capabilities, which means that off-gas entrainment rates cannot be measured directly.…”

“…Thus, it was proposed that steady state mass balance be set up around the time of glass pour stream (PS) sampling and off-gas entrainment rates be calculated as the difference between the calcined feed and glass pour rates as a representation for the entire sludge batch. 5 This proposed method was tested earlier using the SB6 PS sample taken while Canister #3472 was being filled over a 20-hour period on 12/20/2010, approximately three months after the glass bubblers were installed. 5 The analytical results for that PS sample provided the necessary glass composition for the mass balance calculations.…”

“…5 This proposed method was tested earlier using the SB6 PS sample taken while Canister #3472 was being filled over a 20-hour period on 12/20/2010, approximately three months after the glass bubblers were installed. 5 The analytical results for that PS sample provided the necessary glass composition for the mass balance calculations. To estimate the "matching" feed composition, which is not necessarily the same as that of the Melter Feed Tank (MFT) batch being fed at the time of PS sampling, a mixing model was developed by including three preceding MFT batches as well assuming perfect mixing in the melt pool.…”

“…Comparison of the resulting SB6 entrainment ratios with the DWPF design basis values and those measured during the DM1200 run showed that the proposed method of estimating off-gas carryover from a simple mass balance appears feasible at least for the major species (e.g., Si, Na, Fe and Al) and thus additional case studies were recommended. 5 Testing of the proposed method continued in this study using the SB7a PS sample data taken while Canister #3619 was being filled over a 21-hour period on 8/5/2011 with the glass bubblers in operation. Several changes were made to the approach used in the SB6 mass balance in an effort to more accurately interpret and apply available operating/process data, including dilution of each MFT batch by the addition of trickle water flow, blending of waste and frit according to the ratios reported in the Product Composition Control System (PCCS) spreadsheet and adjustment of measured slurry feed rates to the melter.…”