Cancer Deaths among Radiation Workers 907 ig. 2 displays these time distributions. Calculations have been performed with A and B spread over 50 years as well as 30 years as in the figure and these ave distinguished as Ago, Aso, etc. This permits some estimate of the difference between leukaemia cdeaths—all of which may be expected to have occurred within 30 ycars—and other cancer deaths where the risk may spread over a longer period following exposure. 5 ~ @ o a. % <= % < S 3 S sa 4-00 5a on 0-07 0-06 0-05 0-04p 0-03 0-0? 0-0i0 5 {l c oe B — 10 a 15 a 20 at 25 30 Time following exposure lyears } Tig. 2. The three risk-time relations Aj, By, and C,,. Wo also consider Agg and Bago (i.c. the same form as Ay and Byg but spread over 50 years). Having established the three factors (1), (1) and (iit) we may now calculate the expected age-specific racdiation-induced deaths in our work foree. The munbers of workers in cach 10-year age group are shownin column B of tables 2 and 3 and the corresponding ‘non-radiation’ deaths in columns C, D and Is. Columns I? and @are calculated by summing a man’s risk of dying in any particular year arising from cach year of exposure up to that tine. Henee we caleulate the number of deaths in a group of the same age and exposure. This is repeated for all age and exposure groups and the results summarized in columns Pand G. Both columns Jf and G (Ag, and Az, vespeetively) have been evaluated with a risk cocfiiciont of LO-4 per rad so to find the expected number of radiation-induced leukacmia deaths we must seale column Tf appropriately fsce eqn 1). While the age-specific numbers of radiation-indueced deaths shown in tables 2 and 3 are applicable to a population of 100000 workers cach exposed to Lrad/ yeav with an associated risk of 10! per rad, we ean deduce the corresponding numbers for any other parameters from =mx i fe ALif = 105°x io-4 D=mP, = INL, RD{LO 4h (1) where m is the number of deaths shown in the table, 2, is the working population, & is the risk per rad and / is the average annual dose per worker. ‘This simple sealing is possible because of the negligible effect of radiation-induced deaths on the population distribution. The calculations described are for the steady state and ib would take 50 years for the exposure distribution to be reached. Since this is longer than the nuclear power industry has been in existence, it is of interest to consider the approach to the steady state situation. We have maintained the same work