External Dose Estimates for Future Bikini Atoll Inhabitants (UCRL-51879 REV.1)

Page 5

EXTERNAL DOSE ESTIMATES FOR FUTURE BIKINI ATOLL INHABITANTS To evaluate the potential radiation doses that may be received by the returning Bikinians, we surveyed the residual radioactivity on Bikini and Eneu Islands in June of 1975. An integral part of the survey included measurements of gamma-ray exposure rates which are used to estimate external gamma-ray doses. The survey showed that on Bikini Island the rates are highly Variable: values near the shores* are generally of the order of 10 to 20 HR/h, while those within the interior average about 40 pR/h with a range of roughly 30 to 100 pR/h. Eneu Island, how– ever, is characterized by more or less uniformly distributed gamma radiation levels of less than 10 DR/h over the entire island. These data, in conjunction with population statistics and expected life styles, allowed us to estimate the potential external gamma-ray doses associated with proposed housing locations along the lagoon road and within the interior portions of Bikini Island as well as along the lagoon side of Eneu Island. As expected, living on Eneu Island results in the lowest doses: 0.12 rem during the first year and 2.!3 rem during 30 years. The highest values , 0.28 rem during the first year and 5.9 rem over 30 years, may potentially be received by inhabitants living witbin the interior of Bikini IsJ.and. Other options under consideration proBikini Atoll was one of the coral. reef surrounding a lagoon U.S. nuclear weapons testing sites with major and minor axes having in the Pacific. It i.s situated dimensions of 35 and 27 km, respccin the nortl]ern part of Micronesia tively (1’ig. ~). The total. land in the Central. Pacific Ocean Wea .is about 6 km , and the .~ancl about 3600 km southwest of llonol.ulu. Ileight generaLLy averages 3 to 5 m The atoll consists of a number above mean .sca level.. The islands OC small islands on an el.1.iptical. vary In size from small. sanclbars of

Page 6

a few hundred square meters to Eneu are the most likely islands to A total of 23 nuclear tests took place during the testing period. Most of the tests were conducted on barges anchored in the lagoon or on the reef. All. islands were subjected to varying degrees of close–in fallout. Generally, the prevailing winds transported the radioactive clebris clouds toward the southwest. One exception, however, occurred during the Bravo event when of Bikini Atoll. une:{pected changes in the wind directions caused the cloud to travel towarcl the east over Bikini Island. Most of tl~e radioactive contamination on Bikini Islancl is due to this event. This recen~ survey was designed to evaluate the potential external gamma doses associated with proposed housing locations on Bikini and llneu Islands, and to evaluate the potential doses received through the major terrestrial food crops on the atoll.. In this report we only assess the ex~ernal gamma doses.

Page 7

Techniques Used to Measure Gamma-Ray Exposure Rates Because the external dose is primarily due to gamma-emitting radionuclides , with only minor contributions from alpha and beta emitters , we had to obtain the best possible description of the geo– graphical. variability of the gamma-ray exposure rates on Bikini and Eneu Islands. Any technique for measuring gamma exposure rates has its own set of limitations (e.g., nonlinear energy response, portability of equipment, and extent of geographical coverage). We therefore used four different techniques to obtain the detailed geographical coverage and accuracy scintillation detectors, a commercially available pressurized ion chamber, and two types of thermoluminescent dosimeThe portable scintillation meter readout. The detectors were calibrated in microroentgens per source on the primary calibration range of the National Environmental Research Center, Las Vegas, Nevada. Calibration was repeated on selected instruments following the survey. The detectors measured the ex– posure rates at 1 m above the ground at about 2500 locations on a 30–m rectangular grid over the entire surface of Bikini Island, and at about 200 locations on a 120-m grid on Eneu Island. Since the response of the detectors was energy-dependent and they were calibrated with a point source, they were expected to overrespond to the gamma flux on the atoll because the flux is depth distributed and has a higher scatter component — and, therefore, a lower energy — than the point source. The detectors could be carried easily, which allowed us to make measurements at many locations on a uniform grid of the islands. They are virtually insensitive to cosmic radiation. The response of the detector was compared with that of the pressurized ion chamber over the entire range of observed exposure rates. The ion chamber consists of a stainless steel sphere filled with The current produced by the radiationinduced ionization within the chamber is measured by a sensitive electrometer with digital readout. The detector was calibrated by the manufacturer and verified by several ERDA laboratories. It exhibits a relatively flat energy response over the gamma-ray energies of

Page 8

radiation field. Therefore, its response is often used as a reference to which other measurements may be compared. The chamber walls are sufficiently thick to render the detector insensitive to the beta radiation present in fallout fields. The instrument is, however, sensitive to cosmic Further gamma exposure rate comparisons were made by means of dosimeters (TLD’S) placed at 80 locations. The LiF chip displays an essentially flat energy response and excellent thermal stability. The response of LiF is within approximately 1% of being air equivalent for a typical environ– mental radiation field. The CaF2: Dy TLD’s have an enhanced energy response at low energies, and were used to cletect possible low– energy radiation fields by comparison with the Lij? readings. The LiF and CaF2 chips were matched to 5% and ~t% respectively within each batch. The TLI)’s were annealed on the atoll. immecliatel.ybefore being placecl on the two islands. Two Lawrence personnel. badges con~aining tlmee packets were attached to trees by nylon straps or placecl on wooden stakes at a height of 1 m above the ground. ‘rhe locations were carefully chosen to obtain exposures over the full. range of gamma exposure. rates observed by the portable instrument period, the dosimeters were retrieved and handcarried (by air) in a lead container to Livermore for readout. We studied calibration and signal fading by exposing separate sets of TLD’s to a 137 Cs point source before and after the exposure period. A fixture was constructed for field use which aided in obtaining l Using a NBS calibrated Radoconk l Comparing the response of a set calibrated Co source to that obtained from the calibration ‘rhe calibration is known within f 3% at one standard deviation. We stored a set of control TLD’s in a lead pig 011 a “clean” island i~~ the Marshalls cluring the *Reference to a company or product name does not imply approval or recommendation of the product by the University of California or the U.S. Energy Research & Development Administration to the exclusion of others that may be suitable,

Page 9

exposure period for background determination. The background exposure was essentially all contributed by cosmic radiation during the 3–month exposure period and during the aircraft flight to LLL. Additional TLD’s were stored on the periphery of the lead pig to identify possible inadvertent exposures. The average background exposure for the two types of TLD’s was subtracted from al]. field measurements so that the results represent only the terrestrial radiation exposure rates. We found that sunlight had a negligible effect on this packaging arrangement. The correspondence between the results obtained with the NaI scinti].later and the pressurized ion chamber is presented in Fig. 2. The ion chamber readin~s have been contribution at that latitude. The tor overresponded because of its nonlinear energy characteristics. occurs at a range switching point were measured on both low and high range, and those resu].ts are shown in solid circles. On the range of O to 30 DR/h, a correspondence range, rhe correspondence, though Ion chamber response —~R/h 2. Comparison of responses of the Nal scintill.ator and the pressurized ion chamber. linear , deviates more markedly from approximately 21% relative to the LiF. This 1s consistent with similar studies made at Enewetak Atoll and with environmental monitoring perresponse varies with energy and this ratio (1.21) corresponds to an average gamma energy of about 500 lceV. This is reasonable based on and the predominance of 137cs activities distributed in the soil. To assess Lllebeta contribution to the LilF exposure rates, various

Page 10

thicknesses of aluminum absorbers Calibration of the pressurized were placed over an array of dosimeters 226Ra ion chamber against a point at three sites on Bikini Island. A feather analysis of the beta attenuation curves gave a maximum beta energy between 1.5 and 2.2 hleV. Given the known predominance of 90 Sr-gOY beta activities in the this energy range is consistent The analysis also revealed that the average beta contribution to the a rather significant contribution. Therefore, it was necessary to reduce the LiJ? results by this amount to The comparison between the ion ray exposure rates is presented in Fig. 3. A linear regression of the two data sets gives agreement of about 13% between the two methods. One also finds that the correlation of points in Fig. 3 is no~ as good as that in Fig. 2. This clifference is most likely CIUC to the beta con– tribution to the Lit? results, which may vary throughout the i.slancls, causing spread in the daL.a.” Departure from the 1:1. relationship in Fig. 3 may be clue to an overcorrection of the ‘1’LDdata for beta response or to insufficient consi.cleration of the ion chamber clata for energy dependence. source, the method used with the instrument in this study, leads to about a 3% overestimate in the tal fields in this country. If a similar correction were made to these data, the agreement of the two TLD) would be within 10%. This is considered to be satisfactory agreement between the two reference techniques used in this work. Hence, on the basis of these results, the NaI scintillation readings were normalized to the output of the pressurized ion chamber. Fig. 3. Comparison of responses of the pressurized ion chamber bution due to cosmic radiation

Page 11

Results of Gamma-Ray Exposure Rate Measurements The geographical variability of the gamma-ray exposure rates for Bikini and Eneu Islands is shown in Figs. 4 and 5. ‘rhe contribution due to cosmic radiation has been sub– tracted. On Bikini Island the individual measurements from which the contour levels were derived are listed in the Appendix. Note the complex patterns displayed throughout the island. This complexity may be due, in part, to the inhomogeneity in the original fallout pattern produced by the Bravo event, but it certainly reflects the extensive earth moving activities performed over the entire island as part of the agricultural. rehabilitation program. The exposure rates near the shores are typically of the order of 1.0 to 20 llR/h, whi.].ethe wide range of roughly 30 to 1.00 ).IR/h. may be visualized as having a general background of about 30 to 40 ]~R/h with numerous irregularly shapecl areas posed i.n a random fashion over this visual.lzed by viewing the three graphical displays of the exposure vertical coordinate is a measure of the gamma exposure rate. ThLIS , the values near the shores appear as relatively I.OW flat areas. Note especially the low flat area situated on the ocean side near the The gamma exposure rates measured on Eneu Island (Fig. 5) show that the island is characterized by low (less than, 10 ]JR/h) and more or less uniIormly distributed gamma radiation levels over the entire These total gamma-ray exposure rates are the basis for the external mine. the annual. dose and dose commitment , it was also necessary to determine the fractional contribution made by the predominant buted i.n the soil.. Based on our the data of Bennett and Beck 3 Survey, we expected chat the primary contribution Lo the gamma exposure rates would be due to soil. ‘l’racequantities of other gamma emitters such as >

Page 14

rates measured on Bikini Island as viewecl from (a) Lhe ocean sidti and (b and c) the lagoon side. The vertical coordinate is a measure of the exposure rates. Elevated exposure rates appear as “peaks” wh~le the lesser values show up as relatively low flat areas. Note the lower va~ues along the shores and the higher values within the the ocean side near the center of ~he is].and.

Page 17

Eu, and Am were expected to contribute at most a few percent to the total exposure rates. This was confirmed by Ge(Li) gamma spectral analyses of several hundred soil samples collected on both islands during the June 1975 survey. The detailed results of tj~e soil survey will be published in a subsequent report. In summary, the soil survey included the random collection of two types of soil samples on each island: surface and profile. Each surface sample consisted of two 1.5–cm–deep cores. Profile samples were obtained from the sidewall of a trench dug for the purpose. On Bikini Island the exhibited by the 1.5–cm-deep core samples were 41 pCi/g ancl 0.74 pCi/g respectively; whi].e on Eneu Island, the corresponding values were 2.5 the profile samples showed a wide range of activity distributions AS a function of depth on the two i.sl.ands. Even though generalizations are difficult to make, the activities on Bikini Island usually decreased with depth in the first few centimeters with a relaxation length of about 5 cm (the depth at which the activity is e , or 37% of the surface activity) . On Eneu Island, the activities were relatively low and uniform throughout the full range of depths sampled. Using these data in conjunction with the data of Beck et al., 4 we estimated the average Co contribution to the total gamma-ray exposure rates over the two islands to be 94% and wel-e assumed to be valid over the remaininz j.slands of the atoll. External Dose Estimation In addition to tile gamm:~-ray exposure rates, wc need to consider the expectecl living patterns of the future inhabitants in order to evaluate the extern:~.1close problem. Of course, ma~ly uncertainties are inherent in the prediction of future living patterns. llowever, the following cases, shown in Table 1, have been chosen as a conditions tl~atwould cover the range of doses that could be received by tion. These were basecl upon our with personnel. from the Trust Territory of the Pacific Islands.

Page 18

Table 1. Assumed living patterns, 1. No use of Bikini Island for the present as a housing production area. Use of Eneu Island for hou~ing and tion. Unrestricted use of fish throughout the atoll. 2 Limited use of Bikini. Island with residence in houses already constructed. No additional house construction on Bikini Island for the present. Use of coconuts grown on Bikini Island. Other food crops grown on Eneu Island only. Unrestricted use of fish from all parts of the atoll. Use of Bikini Island lens water for 3 L,imited use of Bikini Island with the following remedial actions taken: (a) placing 5 cm of clean …