904 J. A. Reissland et al. In section 2 we describe the model we have used to represent a work force of raciation workers and the way in which we have estimated their chances of dying from an induced cancer. Section 3 discusses the time necessary to achieve sufficient data to be able to draw conclusions with some specified level of confidence. We also consider the magnitude of risk that is detectable in a given time. Results of calculations are presented and discussedin section 4 and we make comments on possible conclusions in the final section. 2. The model There are three aspects to be settled before any calculations can be carried out: (i) the age distribution of the work force and the rate at which workers leave, (ii) the natural incidence of deaths from causes which may also be induced by radiation, and (iii) the risk of death due to exposure to radiation and howthis risk is distributed in time. 2.1. Lhe work force Our calculations refer to a work force of 100000 distributed in age as shown in fig, 1. This distribution is based on the actual distribution at an established nuclear energy site ancl does not differ greatly from any typical British industry. We have assumed for simplicity that the annual percentage Jeaving radiation work other than by death or retirement is the same for cvery age group. Stability of the distribution is maintained by introducing new workers (with no previous industrial radiation exposure) to replace those leaving any group by death, resignation or retirement. ‘The total number of ex-radiation workers is dependent on the leaving rate. Irom the records of the National Radiclogical Protection Board (NRPB) and from informai discussions with employers, we have arrived ab a figure of between 8 and 10%for the annual percentage of workers who cease radiation work other than by death or retirement. We have considered 5 and 10%which, understeady state conditions, lead respectively to 171000 and 330000 living ex-workers. We have assumed also that those leaving radiation work do not return to ib within the latent period of risk following their last exposure. uw aT @ ©568 2000} c 2 e Be Oo ~~ ot Qs@ ‘SO OG ZO pone 2800 st 2500 o 210001 a 1 —s 93000 —4 _—j— 13000 1000; F r f — f OS 30 BL 38-33-96 LO a LE 8? 5660 Gh Age {years} Wig. 1. he age distribution of the work foree used throughout this paper.