Twitter LinkedIn


Radiological Sampling, Modeling and Analysis Methods Presented at REEW 2017
Posted on July 26th, 2017

RSCS recently presented at the EPRI Radiological Effluents and Environmental Workshop held in Indianapolis, IN from June 26-29, 2017.  Eric Darois, Matt Darois and Jonathan Holden-Remick presented on topics related to sampling, modeling and analyzing data on radiological discharges from nuclear utilities.

A common problem for all nuclear utilities is the generation of tritiated water during routine plant operations.  The risks involved in disposing this water was covered by Jonathan Holden-Remick in “Comparative Risk Study of Transporting Tritiated Water vs. Discharging at Decommissioning NPPs”.  This presentation explores the various options available to decommissioning plants, which are limited by NRC and EPA regulations and can be further restricted by local state regulations.  An example of the risk informed decision on how to dispose of tritiated water at the Vermont Yankee Plant was provided.  This case study compares the risk of discharge in accordance with the ODCM limits committed to by the plant to the risk of an accident while transporting this water to the US Ecology site in Utah.  While both risks to the public are comparatively low, the study concludes that the discharge of tritiated water provides considerably less risk than transport on US roadways to the disposal facility.

A method for determining the origins of radionuclides potentially discharged by a plant and collected in sewage samples that tested positive for activity was provided by Eric Darois in “Evaluating the Potential for Low Levels of Radioactivity in Sewage using Monte Carlo Analysis Methods, A Case Study”.  This presentation covered an evaluation using a comprehensive Monte-Carlo analysis of the potential input sources of Sr-90 into sewage system that resulted in a positive result. The mathematical model of sewage dilution flow rates based on metered sewage processing data was correlated with onsite worker population data.  Empirical statistical distributions were established based on the data collected and were propagated through the mathematical model using a Monte-Carlo-based tool, ModelRisk. Variables considered include the potential dietary excretion rate of Sr-90, station related sources of radioactivity and potential sources of Sr-90 from specific cancer treatment modalities using Y-90 where Sr-90 is a potential contaminant.  From this analysis, it was concluded an isolated detection of only Sr-90 is unlikely to be from plant related radioactivity, dietary sources or sources involved in medical cancer treatment. Rather, we concluded that the detection was most likely from cross contamination at a commercial laboratory.

Several utilities discharge cooling tower water directly into waterways.  The presentation “Hydrology Study of Harris Lake”, co-authored by Matt Darois and Heather Baxter of Harris Nuclear Plant, provided the methodology and assumptions used to determine if there were groundwater implications of the plant’s discharge of tritiated water.  Given the potential for seepage of lake water into the aquifer proximal to the Harris Lake dam, RSCS developed a first order kinetics equation to model the mass balance of tritium entering and leaving the lake from all sources.  This model is capable or reasonably predicting lake concentrations, when accounting for all inputs/outputs.

 Finally, statistical methods used to determine the activity in samples were provided by Eric Darois in the presentation “MDA, MDC, LLD, Ld, Lc, Uncertainty, a priori, a posteriori, error rates, MARSSIM, MARLAP: What does it all mean to me?”.  This presentation discussed the statistical concepts associated with data analysis and reporting for radiological analysis such as MDC, MDA, LLD, Critical Level (Lc), Detection Level (Ld), reported uncertainty for both a priori and a posteriori reporting. The survey methodology contained in MARSSIM (Multi-Agency Radiological Site Survey and Investigation Manual) intended for principally supporting characterization and final surveys of sites containing residual radioactivity was presented in regard to RETS REMP programs.  The processes described in MARSSIM specify the methods used to perform surveys of surface soils and building surfaces, the use of data quality objectives and data quality assessments and the application of non-parametric statistical tests to evaluate the data.  In addition, The MARLAP (Multi-Agency Laboratory Analytical Protocols) manual, a companion document that may be more directly applicable to RETS REMP programs, was discussed as it applies the principles of the MARSSIM manual for more general directed sampling programs.  This document also applies the data quality objectives and data quality assessment process used in MARSSIM. 

All presentations from the EPRI Radiological Effluents and Environmental Workshop are available for download from the conference website here.

Read All News