TABLE 3-3. VARIATION IN THE 2414m CONVERSION FACTOR* WITH DIFFERENT VALUES FOR SOIL DENSITY AND THE MASS ATTENUATION COEFFICIENT Mass Attenuation Coefficient (1/p), (em2/g) 0.309 (-20) 1.22 (-20) 1.36 (-1c) 1.5 (mean) 1.64 1.78 (+10) (+20) 8.61 8.49 8.38 8.66 8.29 8.57 8.22 8.56 8.95 8.86 8.79 9.24 9.52 9.15 9.43 9.08 9.36 0.321 0.333 (-1c) (mean) 8.89 9.18 8.77 9.06 0.345 0.357 (+1o) (+20) 9.47 9.75 9.35 9.63 *(s °jel Np Soil Density p (g/em 3) (pCi/g)/eps) with detector height of 7.4 m. attenuation coefficient leads to a #6.5 percent change in the conversion factor. Since the soil density and the in situ soil mass attenuation coefficient, in general, both vary from location to location, it is more appropriate to examine their combined effect on the conversion factor. As seen in Table 3-3, the maximum effect occurs with a low soil density combined with a high mass attenuation coefficient or a high density combined with a low mass attenuation coefficient. For the appropriate 2c limits this case would lead to a +9 percent change in the conversion factor. In reality, however, low density areas were generally found to be those areas having higher organic and/or soil moisture content, which would lead to a lower mass attenuation coefficient. Similarly, high density areas generally had a higher mass attenuation coefficient. For this combination the appropriate 20 limits lead to a +5 percent change in the conversion factor. This is more typical of the actual range of uncertainty in the data due to observed variations in the wet soil density and in situ soil composition. Depth Distribution One of the most critical factors in relating an in situ measurement to radionuclide concentration in the ground is a knowledge of the source distribution with depth. This is especially true when attempting to determine the total activity per unit area. For the Enewetak 1am eonversion factor, depth distribution data were obtained from profile measurements made during the 1972 reconnaissance survey (NVO-140). A total of 108 profile measurements were made on 20 islands from Alice to Wilma. The data for each profile, most taken to a depth of 30 em, were fit to an exponential distribution, as given in Equation (2), and a value computed for the relaxation length. Of the 108 profiles, 11 had a relaxation length between 3 and 5 cm, 45 had a relaxation length between 9 and 10 em, 15 had a relaxation length between 10 and 20 em, and the remaining 37 were best ’ represented by a uniform distribution. The last group included those distributions which were slowly decreasing with depth, Slowly increasing with depth, or oscillating up and down with depth. Based on these data, the actual conversion factor was computed from a weighted average of the values obtained for relaxation lengths of 4. em, 7.5 em, 15 em, and 1000 em (i.e., a uniform distribution). Figure 3-4 shows the variation in the 241 Am conversion factor for average concentration in the top z em, with z varying between 0 and 10 em, for several different depth distributions. As can be seen, the conversion factor ean vary significantly with variation in the depth distribution. This variation, however, is minimized when determining the average concentration in the top 2-3 em. In particular, for the 3 em average specified in the Enewetak cleanup criteria, the conversion factor varies from a value of 8.63 pCi/g per eps for a relaxation length of 3 cm to a value of 9.00 pCi/g per cps for a uniform distribution, compared to a value of 8.95 pCi/g per eps obtained from the weighted average. Thus, even for the extreme case of the measured depth distributions, there is only a 4 percent error in the conversion factor. For 90 percent of the distributions measured, the uncertainty in the conversion factor due to variations in the depth distribution is on the order of +1 percent. For this reason, no effort was made to obtain additional depth profiles during the cleanup project. 94