T. DO. Pflaum -2- The following table summarizes some selected fatality rates and risks. Activity Fatality Rate* Risk Al) industries (1976) 14 1.4 x 107 Construction (1976) 57 5.7 x 1074 At work (1980) State of Nevada DOE & Contractors (1978-82 average) 4.9 4.9 x 10 5.6 5.6 x 10 _5 _5 NTS (1965-1981 average) 27 2.7.x 107° Enewetak cleanup 70 7.0 x 1074 | *Per 100,000 worker-years. Because of the great variability in the data, and the requirement to interpolate and extrapolate, it is essential that a careful uncertainty analysis be made by EPA. This analysis is necessary to ensure confidence that the risk of cleanup does not exceed theprisk from leaving the contamination undisturbed; which may be as low as 10” 2. or lower. Imaginary Versus Real Deaths! The models used to assess the health effects (i.e. radiation-induced cancer fatalities) on the Enewetak people during the planning phase estimated < 3 health effects (cancer deaths) over 30 years with no cleanup and no restrictions on island or food usage. An analysis of the total radiation dose to the returning people of Enewetak after the cleanup leads to the conclusion that there might be an additiona} 0.026 deaths in 30 years from cancer caused by radiation. This is compared to the two persons who died in course of the three-year cleanup. The uncertainty which is inherent in cancer-risk estimates is graphically illustrated in Table V-4, page 147 in the 1980 BEIR report in which the expected number from continuous exposure of one rad per year to a population of 1,000,000 ranges from zero to 568, The risk estimates of cancer deaths as required by the proposed EPA standard (maximizing risk estimates) give hypothetical, or imaginary deaths as compared to the real deaths which do occur in construction projects. The fact is that no increase in cancer rate has been, nor can be, identified at the dose levels comparable to background radiation levels.