Home > Research > Publications & Outputs > Remediation of spent oxalic acid nuclear decont...

Associated organisational unit

Electronic data

  • 2018KetuskyPhD

    Final published version, 5.57 MB, PDF document

    Available under license: CC BY-NC-ND: Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License

Text available via DOI:

View graph of relations

Remediation of spent oxalic acid nuclear decontamination solutions using ozone

Research output: ThesisDoctoral Thesis

Published
  • Edward Ketusky
Close
Publication date2018
Number of pages304
QualificationPhD
Awarding Institution
Supervisors/Advisors
Publisher
  • Lancaster University
<mark>Original language</mark>English

Abstract

The Savannah River Site (SRS) has forty-three remaining very large underground tanks containing significant quantities of nuclear waste generated primarily from cold-war radiochemical separations. All of the tanks eventually must be closed. As part of decommissioning/closing the nuclear waste tanks, even residual quantities of the waste must be removed. Although most sludge can be removed mechanically, chemically cleaning (i.e. decontamination) with oxalic acid is used to aid in the removal of residual quantities.

Although oxalic acid works for cleaning the tanks, its downstream impacts are considered detrimental. To better understand and quantify the impacts, detailed models were developed to account for different potential processing strategies for handling the spent oxalic acid nuclear decontamination slurries. Although the results vary, the models show that regardless of the oxalate handling strategies: 1) significant washing to decrease sodium concentration/solids concentration in vitrification feed will be required, and 2) the creation of copious future additions of feed for salt processing will be unavoidable.

Using a Theory of Inventive Problem Solving (TRIZ) approach, a modified form of the Chemical Oxidation Reduction Decontamination (CORD) ultraviolet (UV) light was identified as being used with an analogous but already resolved problem that could be adapted to SRS HLW tank cleaning. A novel preliminary process called Enhanced Chemical Cleaning (ECC) was envisioned.

As part of maturing the technology, the literature review identified three possible oxalate decomposition mechanisms associated with ECC. They are: 
1)A heterogeneous non-Advanced Oxidation Process (AOP) mechanism where the target organic adsorbs onto the surface of a solid, often particulate, metal oxide at a so-called active site, followed by ozone attack on the sorbed organic;

2)A homogeneous non-AOP mechanism that operates under low pH acidic conditions and which involves complexation of the catalysing metal ion with the oxalate followed by ozone attack on the complex; and,

3)A homogeneous AOP mechanism that operates at a high basic pH and which involves metal ions catalysing the formation of hydroxyl radicals from ozone, with the said hydroxyl radicals then driving the oxalate decomposition.

Process testing was conducted using slurries made from simulants designed to be chemically similar to real High-Level Waste (HLW) sludge types. Testing using slurries made from real HLW sludge was also performed, but because of safety limitations associated with handling HLW, only a much smaller scale test apparatus could be used. With the much smaller scale test apparatus, the purpose of the real HLW based testing was confirmatory. Each of the simulant decomposition test slurry was created using an Fe-rich or an Al/Mn-rich sludge simulant using either 1 wt% or 2.5 wt% oxalic acid. The real HLW based slurries were formed using 2 wt% oxalic acid.

As part of the main postulate of this effort, both the simulant decomposition test slurries and real HLW based slurries demonstrate that UV light increased the decomposition rate. Even without UV, by adding only ozone, the oxalate decomposition was completed on an industrially relevant timescale.
Also using simulant based testing, the overall oxalate decomposition exhibited four distinct stages related to the metal catalysts:

Stage One – At short ozonation times, ozone decomposes Fe oxalates and solubilise Fe from ozone action on the metal oxide constituents of the sludge.

Stage Two – At intermediate ozonation times, as a result of the loss of the solution capacity to complex (and so solubilise) Fe, Mn, and Ni ions due to O3 driven oxalate decomposition, as well as the pH increase that accompanies that decomposition, Fe begins to precipitate. Oxalate decomposition is still primarily catalysed by Fe ions during this stage. 
Stage Three – At intermediate ozonation times, Fe precipitation is near complete, and oxalate decomposition is now driven by ozone and Mn catalysis – Mn playing a major role in determining the final time to process endpoint of 1.1 × 10-3 M oxalate in solution. 
Stage Four – At long ozonation times, the process endpoint with Mn precipitation now near completion with Ni being the dominant metal ion in solution.
Constructed plots compare the pH and remaining oxalate concentration, both as a function of time, suggesting some relationship. Regression analysis of the negative log of the oxalate concentration shows the R2 values are all greater than 0.80, confirming correlation. Thus, pH can be used as a field measure for confirming when oxalate decomposition is complete.

As a principal hypothesis of this effort, using simulant based testing, both the scavenging effects of “all-ready present” nitrite (a soluble component of the sludge simulants) and oxalate mineralisation-derived carbonate are advantageously used in lieu of introducing hydroxyl radical probe compounds to the process. Specifically, differing nitrite concentrations between slurries showing no impact on the decomposition rates, and the build-up of carbonate not inhibiting the decomposition process strongly suggest that the decomposition is not the result of radicals. Instead, the oxalate decomposition is likely the result of a direct reaction of ozone with metal complexed oxalate (i.e. mechanism 2 discussed above).