Uranium Conversion and Enrichment Technologies: Sources of Atmospheric Fluoride

Abstract

AbstractTo illustrate the proportional hydrogen fluoride (HF) residuals from two nuclear fuel cycle technologies, uranium hexafluoride (UF6) conversion (feed) and enrichment, air, and vegetation monitoring data were obtained from the Paducah (Kentucky) Gaseous Diffusion Plant (PGDP) from 1972 to 1977, during which the feed manufacturing plant and the diffusion cascade were operating. In the spring of 1977 the feed plant was placed on standby such that effluents containing fluoride came only from the gaseous diffusion operation. Monitoring data obtained without emissions from feed manufacturing (1978) were also obtained. Gaseous fluoride was collected in alkaline solution (NaOH) scrubbers until 1974, when they were replaced by chemically treated (K2CO3) filter papers. Air samples were collected weekly (7‐day averaging time) from 10 stations around the facility. Results indicate the absence of routine releases with observed fluoride concentrations randomly distributed. Of the total number of observations during the 6‐year period, 90% of the concentrations were < 1 µg/m3 (7‐day average). Ambient concentrations consisted of infrequent peaks superimposed upon a predominant distribution of lower concentrations. The scatter plot, by station, reflects distributions downwind and upwind from the source and identifies the area of greatest impact. Vegetation monitoring data were obtained for the same period (1972–1978), collected every 2 months from 18 locations. The vegetation data were summarized by scatter plots which reflect air quality data, depict ranges in concentration among the monitoring network, and identify the stations with significant fluoride accumulation.Ground‐level air (HF) concentrations in the environs of the PGDP were estimated using two dispersion models (Atmospheric Transport Model and Climatological Dispersion Model) and compared to the observed data. The Atmospheric Transport Model estimates were less than the observed concentrations, because the model estimates considered only a single source (gaseous diffusion), whereas the monitoring data incorporated emissions from feed manufacturing and gaseous diffusion operations.