Integrated AMP-PAN, TRUEX, and SREX Flowsheet Test to Remove Cesium, Surrogate Actinide Elements, and Strontium from INEEL Tank Waste Using Sorbent Columns and Centrifugal Contactors [report]

R.S. Herbst, J.D. Law, T.A. Todd, D.J. Wood, T.G. Garn, E.L. Wade
2000 unpublished
Three unit operations for the removal of selected fission products, actinides, and RCRA metals (mercury and lead) have been successfully integrated and tested for extended run times with simulated INEEL acidic tank waste. The unit operations were ion exchange for Cs removal, followed by TRUEX solvent extraction for Eu (actinide surrogate), Hg, and Re (Tc surrogate) removal, and subsequent SREX solvent extraction for Sr and Pb removal. Approximately 45 L of simulated INTEC tank waste was first
more » ... ocessed through three ion exchange columns in series for selective Cs removal. The columns were packed with a composite ammonium molybdophosphate-polyacrylonitrile (AMP-PAN) sorbent. The ion exchange system was operated continuously for ~34 hours at 22 bed volumes per hour in the first two columns. The first two columns were each sized at a bed volume of 60 cm 3 and were operated to 100% breakthrough. The experimental breakthrough data were in excellent agreement with modeling predictions based on data obtained with much smaller (1.5 cm 3 ) columns. The third column (220 cm 3 ) was used for polishing and Cs removal after breakthrough of the up-stream columns. The Cs removal was >99.83% in the ion exchange system without interference from other species. Most of the effluent from the ion exchange (IX) system was immediately processed through a TRUEX solvent extraction flowsheet to remove europium (americium surrogate), mercury and rhenium (technetium surrogate) from the simulated waste. The TRUEX flowsheet test was performed utilizing 23 stages of 3.3 cm centrifugal contactors. The TRUEX test was operated a total of 71.3 hours and processed ~41 L of the IX effluent using 1.5 L of TRUEX solvent with constant solvent recycle. The TRUEX solvent was recycled through the flowsheet an estimated 17 times. Greater than 99.999% of the Eu, 96.3% of the Hg, and 56% of the Re were extracted from the simulated feed and recovered in the strip and wash streams. Minor operational problems were encountered in the solvent wash section, where Hg precipitated as HgO; the problem did not require shutdown of the system. Flooding was never observed during the experiment. Over the course of the test, there was no detectable build-up of any components in the TRUEX solvent. The raffinate from the TRUEX test was stored and subsequently processed several weeks later through a SREX solvent extraction flowsheet to remove strontium, lead, and Re (Tc surrogate) from the simulated waste. The SREX flowsheet test was performed using the same centrifugal contactors used in the TRUEX test after reconfiguration and the addition of 3 stages. Approximately 51 L of TRUEX raffinate was processed through the system during 77.9 hours of continuous operation with 1.5 L of SREX solvent and continuous solvent recycle. The SREX solvent was recycled through the system an estimated 45 times without measurable build-up of any components in the solvent. Approximately 99.9% of the Sr, >99.89% of the Pb, and >96.4% of the Re were extracted from the aqueous feed to the SREX flowsheet and recovered in the strip and wash sections. Operational problems such as flooding and precipitation were not encountered during the SREX test. Approximately 41 L of simulated tank waste (based on the volume processed through the TRUEX flowsheet) was processed through the integrated flowsheet and resulted in 175 L of liquid high activity waste (HAW) and 219.6 L of liquid low activity waste (LAW). The HAW fraction would be evaporated, dried and subsequently vitrified for final disposal. Based on current baseline assumptions, including a maximum phosphate loading of 2.5 wt. % in the HAW glass, the flowsheet tested would result in the production 0.195 kg of glass per L of tank waste processed. The LAW fraction would be solidified (via evaporation and denitration) and subsequently grouted. The current baseline assumptions for grouting the LAW stream indicate 0.37 kg of grout would be produced per L of tank waste treated. Under these assumptions, treating the current inventory of ~5E+6 L (5,000 m 3 ) of tank waste would result in 375 m 3 of HAW glass and 1,135 m 3 of LAW Class A performance grout. The HAW glass volume could be significantly decreased by suitable TRUEX flowsheet modifications. iii analytical support for the copious number of samples generated during the testing. Without the dedication of these individuals, this work would not have been possible. Troy Tranter modeled the ion exchange column performance and calculated dynamic capacities. Troy's efforts were essential for a successful outcome and are greatly appreciated. Charles Barnes provided numerous flowsheet calculations to estimate volumes of HAW and LAW associated with the integrated testing. iv
doi:10.2172/774307 fatcat:lm27bzkmfrftrc3fzqiuykidbm