Review Of Rheology Models For Hanford Waste Blending

Koopman, D. C. ; Stone, M.

关键词: Rheology;    WTP;    Hanford;   

DOI  :  10.2172/1095107
RP-ID  :  SRNL-STI--2013-00423
PID  :  OSTI ID: 1095107
Others  :  TRN: US1400084

学科分类：核能源与工程

美国|英语

来源: SciTech Connect

【 摘 要 】

The area of rheological property prediction was identified as a technology need in the Hanford Tank Waste - waste feed acceptance initiative area during a series of technical meetings among the national laboratories, Department of Energy-Office of River Protection, and Hanford site contractors. Meacham et al. delivered a technical report in June 2012, RPP-RPT-51652 ''One System Evaluation of Waste Transferred to the Waste Treatment Plant'' that included estimating of single shell tank waste Bingham plastic rheological model constants along with a discussion of the issues inherent in predicting the rheological properties of blended wastes. This report was selected as the basis for moving forward during the technical meetings. The report does not provide an equation for predicting rheological properties of blended waste slurries. The attached technical report gives an independent review of the provided Hanford rheological data, Hanford rheological models for single tank wastes, and Hanford rheology after blending provided in the Meacham report. The attached report also compares Hanford to SRS waste rheology and discusses some SRS rheological model equations for single tank wastes, as well as discussing SRS experience with the blending of waste sludges with aqueous material, other waste sludges, and frit slurries. Some observations of note: Savannah River Site (SRS) waste samples from slurried tanks typically have yield stress >1 Pa at 10 wt.% undissolved solids (UDS), while core samples largely have little or no yield stress at 10 wt.% UDS. This could be due to how the waste has been processed, stored, retrieved, and sampled or simply in the differences in the speciation of the wastes. The equations described in Meacham's report are not recommended for extrapolation to wt.% UDS beyond the available data for several reasons; weak technical basis, insufficient data, and large data scatter. When limited data are available, for example two to three points, the equations are not necessarily satisfactory (justified) for interpolations, due to the number of unknown variables equal the number of known data points, resulting in a coefficient of determination of one. SRS has had some success predicting the rheology of waste blends for similar waste types using rheological properties of the individual wastes and empirical blending viscosity equations. Both the Kendall-Monroe and Olney-Carlson equations were used. High accuracy was not obtained, but predictions were reasonable compared to measured flow curves. Blending SRS processed waste with frit slurry (much larger particles and the source of SRS glass formers) is a different sort of problem than that of two similar slurries of precipitated waste particles. A different approach to rheology prediction has had some success describing the incorporation of large frit particles into waste than the one used for blending two wastes. In this case, the Guth-Simha equation was used. If Hanford waste is found to have significant particles in the >100 μm diameter range, then it might be necessary to handle those particles differently from broadly distributed waste particles that are primarily <30 μm in diameter. The following are recommendations for the Hanford tank farms: Investigate the impact of large-scale mixing operations on yield stress for one or more Hanford tanks to see if Hanford waste rheological properties change to become more like SRS waste during both tank retrieval and tank qualification operations; Determine rheological properties of mobilized waste slurries by direct measurement rather than by prediction; Collect and characterize samples during the waste feed qualification process for each campaign; o From single source tanks that feed the qualification tanks; o Blends from the qualification tanks; Predictive rheological models must be used with caution, due to the lack of data to support such models and the utilization of the results that come from these models in making process decisions (e.g. the lack of actual operation experience). As experience is gained, the use of blending models that have been validated with real waste may become useful to predict future blends; Obtain more data measurements to check the validity of unknown coefficients for a given blending equation.