readings were selected for soil sampling. Six samples were collected from different piles with each sample comprised of soil from the top, middle and bottom of the pile. The results from these samples showed a range of TRU activities from 101 pCi/g to 304 pCi/g indicating that it was possible a lot of the original top soil remained in the dirt and brushpiles. Subsurface Characterization Because Pearl] had one GZ, subsurface sampling was conducted in December 1977 and January 1978 to search for any pockets of contamination around Inca GZ and also the anomalous area around 5-S-3. Figures 7-86 and 7-87 show the locations and highest TRU result for each location for these two areas. Two iterations of sampling took place with the first being auger samples and the second being sidewall samples. (See Section 6.9.) Neither area showed any TRU activity greater than 160 pCi/g averaged over 0.0625 hectare below 20 cm. samples were collected after debris removal sampling. As previously mentioned, additional subsurface Figure 7-88 shows the results and locations for this Cleanup Activities In Mareh 1979, it was decided to clean Pearl to below 80 pCi/g based on the data collected after debris removal. IMP measurements were taken on some 25 m grid nodes to better define the boundaries for areas where TRU activity exceeded 80 pCi/g. The fine grid data were measured only around the original 50 m boundaries and not over the entire area because additional data in the interior would not change the 0.5 hectare average. (Originally 0.25 hectare estimates were made but the TRU eriterion for an agricultural island is 80 pCi/g over 0.5 ha. Refer to Section 2.2.) The ratio of 6.91 determined from soil samples collected after debris removal was used on the fine grid data. Figure 7-89 shows the 7.75 hectare area where TRU was estimated to exceed 80 pCi/g averaged over 0.5 hectare based on all the data. The areas requiring cleanup were excavated and all the soil stockpiled on the west end of Pearl for later removal to Yvonne. This was done so that the IMP could measure the areas wherethe soil had been removed and also in "no-lift" areas that were downwind or otherwise could be affected by soil removal. The IMP results indicated that three more small areas required a lift in order for the surface TRU to be below 80 pCi/g averaged over 0.5 ha. Two of the areas were on the fringes of the initial removal boundaries, therefore these removals were first lifts. The other lift was in an area where soil removal had already occurred. This was the only second lift necessary on Pearl. After the removal of the stockpile and the three additional areas, these areas were remeasured by the IMP. In addition, twelve locations were soil sampled for ratio determination after cleanup. Two ratios were estimated for Pear: 6.81 + 0.30 for cleanup areas and 4.35 + 0.50 for noncleanup areas. The highest 0.5 hectare average TRU after surface soil removal was 61 pCi/g (based on original data). The estimated amount of surface soil removed was 11,096 cubie meters (14,513 cubic yards) and the estimated curies of TRU activity removed was 1.64. Fission Product Sampling In support of the dose assessment, fission products sampling (Section 6.11) was done for the eastern part of Pearl (noncleanup area) in March 1979. The remainder of the island was sampled in May and June 1979 after surface soil removal was complete. This sampling was conducted on the 50 m grid already established with 29sr analysis done on 17 of 72 sampling locations. Using the nearest located TRU ratio based on the post-cleanup data rather than a mean value and the 241 Am gamma data from this additional sampling, some suspect pockets of subsurface contamination were revealed. Four locations showed a TRU value above 160 pCi/g at some depth. Because the initial subsurface sampling was 5 em cuts at 20 em intervals and the fission products sampling was at different increments, these four areas were not discovered in the initial subsurface investigations. The first step in investigating these spots was to examine the validity of the ratio used in computing the TRU activity. The ratios did not change significantly so the areas werestill suspect. The next step was to collect soil samples as described in Tech Note 18. Figures 7-90 through 7-93 show the results and sampling locations for the four areas on Pearl after sampling. As shown by Figures 7-90 through 7-92, no other elevated subsurface TRU activity was found for three of the areas and no 309