Document: GRE-3-88-11

Test of the applicability of a simulation model for predicting radiocesium contamination in fish and water of two warm temperate cooling reservoirs.

LEWIS, G.P.* and B.E.TAYLOR

Savannah River Ecology Laboratory, Aiken, SC 29802 USA ¹

Abstract:
The VAMP model, developed using data from seven Chernobyl-contaminated lakes in Europe, was designed to predict radiocesium dynamics in a broad range of lake types in the cold temperate zone. We tested the model's ability to predict radiocesium dynamics in two warm temperate reservoirs (Par Pond and Pond B) on the Savannah River Site (SRS) in South Carolina. The reservoirs served as secondary cooling systems for a nuclear production reactor and received contaminated reactor effluents between 1954 and 1964. We modified the model to accommodate aquatic, rather than atmospheric, input of radiocesium to the reservoirs. We compared the model's predictions to measured radiocesium concentrations in predatory fish (largemouth bass) and water at intervals of one to three decades after radiocesium inputs ceased. The model predicted that radiocesium in fish and water would decline more rapidly than was actually documented (e.g., in Par Pond: 8-11 % per yr measured vs. 48-49% per yr predicted). The model assumed that burial of radiocesium into deeper sediments would diminish the amount of radiocesium that could be remobilized from sediments into the water column over time. Unlike radiocesium in the contaminated European lakes, which was buried gradually within sediments over time, radiocesium in the two SRS reservoirs remained in the upper few centimeters of sediment two to three decades after radiocesium inputs ceased. Further modification of the model so that radiocesium was not buried more deeply into the sediments only slightly decreased the rate at which concentrations in fish and water were predicted to decline (31% per yr rather than 48-49% per yr). Also, seasonal oscillations of dissolved radiocesium in the water column predicted by the model were much smaller than oscillations in the empirical data (e.g., 1.3-fold variation in modeled data vs. 3.3-fold variation in empirical data). We conclude that the VAMP model does not accurately predict radiocesium dynamics in warm temperate reservoirs, where intense biological activity in sediments may play a more important role in radiocesium remobilization than in cold temperate lakes.

Keywords: radiocesium, model, lakes, reservoirs

This abstract is being presented at: 10:30 AM in session:
Poster Session #9: Fish, Lakes, Streams and Wetlands.