Investigation into optimal conditions for cross-flow filtration of high-level nuclear waste

Citation
Ra. Peterson et Jl. Gaddis, Investigation into optimal conditions for cross-flow filtration of high-level nuclear waste, SEP SCI TEC, 34(6-7), 1999, pp. 1507-1520
Citations number
4
Categorie Soggetti
Chemistry
Journal title
SEPARATION SCIENCE AND TECHNOLOGY
ISSN journal
01496395 → ACNP
Volume
34
Issue
6-7
Year of publication
1999
Pages
1507 - 1520
Database
ISI
SICI code
0149-6395(1999)34:6-7<1507:IIOCFC>2.0.ZU;2-#
Abstract
The Savannah River Site has 23 Type III high-level radioactive waste tanks, each with a storage capacity of 1.3 million gallons. These tanks contain n early 9 million gallons of precipitated salt. To immobilize the waste, the salt is dissolved through water addition, followed by precipitation of the radionuclides through the addition of sodium tetraphenylborate. This precip itate is then concentrated and washed to remove sodium through cross-flow f iltration. This waste pretreatment process started radioactive operation in late 1995. During the normal plant operation, the cross-flow filtration sy stem (consisting of two 216-square-foot filter elements) maintains a consta nt filtrate production rate. This objective is achieved by allowing the ope rating pressure to increase to maintain a constant filtrate production rate . A maximum pressure differential limit of 40 psig has been imposed on this system. When this maximum is approached, a high-energy backpulse of filtra te removes foulant from the surface of the filter, thereby restoring the fi lter flux. This laboratory work examined two key aspects of the anticipated facility o perating conditions: the efficacy of using pressure differential to control filtrate production rates and the risk posed to filter performance associa ted with pore plugging of the filter immediately following the backpulse. T ests used simulated tetraphenylborate precipitate and a bench-scale cross-f low filtration unit consisting of two parallel filter units each 4 feet in length. Tests used slurries containing between 1 and 10 wt% tetraphenylbora te to cover the anticipated range of operation. Data collected included bot h initial flux-decline measurements and steady-state filtrate production me asurements. Analysis of these data indicates, for the more dilute slurries, pressure was an effective tool in controlling filtrate flux. However, as t he slurry became more concentrated, the ability to manipulate filtrate flux by pressure greatly diminished. Analysis of the initial filtrate decline d ata using first- principle models indicates that the primary mechanism for decreasing filter flux involved development of a surface cake. Given the op erating constraints of the facility, these results provide guidance for fut ure filtration operation.