Historical Doses from Tritiated Water and Tritiated Hydrogen Gas Released to the Atmosphere from Lawrence Livermore National Laboratory (LLNL). Part 6. Summary

Throughout fifty-three years of operations, an estimated 792,000 Ci (29,300 TBq) of tritium have been released to the atmosphere at the Livermore site of Lawrence Livermore National Laboratory (LLNL); about 75% was tritium gas (HT) primarily from the accidental releases of 1965 and 1970. Routine emissions contributed slightly more than 100,000 Ci (3,700 TBq) HT and about 75,000 Ci (2,800 TBq) tritiated water vapor (HTO) to the total. A Tritium Dose Reconstruction was undertaken to estimate both the annual doses to the public for each year of LLNL operations and the doses from the few accidental releases. Some of the dose calculations were new, and the others could be compared with those calculated by LLNL. Annual doses (means and 95% confidence intervals) to the potentially most exposed member of the public were calculated for all years using the same model and the same assumptions. Predicted tritium concentrations in air were compared with observed mean annual concentrations at one location from 1973 onwards. Doses predicted from annual emissions were compared with those reported in the past by LLNL. The highest annual mean dose predicted from routine emissions was 34 {micro}Sv (3.4 mrem) in 1957; its upper confidence limit, based on very conservative assumptions about the speciation of the release, was 370 {micro}Sv (37 mrem). The upper confidence limits for most annual doses were well below the current regulatory limit of 100 {micro}Sv (10 mrem) for dose to the public from release to the atmosphere; the few doses that exceeded this were well below the regulatory limits of the time. Lacking the hourly meteorological data needed to calculate doses from historical accidental releases, ingestion/inhalation dose ratios were derived from a time-dependent accident consequence model that accounts for the complex behavior of tritium in the environment. Ratios were modified to account for only those foods growing at the time of the releases. The highest dose from an accidental release was calculated for a release of about 1,500 Ci HTO that occurred in October 1954. The likely dose for this release was probably less than 360 {micro}Sv (36 mrem), but, because of many unknowns (e.g., release-specific meteorological and accidental conditions) and conservative assumptions, the uncertainty was very high. As a result, the upper confidence limit on the predictions, considered a dose that could not have been exceeded, was estimated to be 2 mSv (200 mrem). The next highest dose, from the 1970 accidental release of about 290,000 Ci (10,700 TBq) HT when wind speed and wind direction were known, was one-third as great. Doses from LLNL accidental releases were well below regulatory reporting limits. All doses, from both routine and accidental releases, were far below the level (3.6 mSv [360 mrem] per year) at which adverse health effects have been documented in the literature