Analytical Program - 1975 Bikini Radiological Survey (UCRL-51879 Part 2)

Page 1

ANALYTICAL PROGRAM — 1975 BIKINI Keith O. Ilamby, Austin L. Prlnclle, and Ilarris B. levy

Page 3

Initial Processing of Field Samples ................... Vegetation and Animal Samples ................... Measurement of 241Am Concentrate

Page 5

/tical program for con samples of soil, vegetation, and of measurements, and colilpaL_iSOrlS animal tissue collected during the of gamma spectrometr~ l.’l(-llwet .June 1.975 fielcl survey of Bikini chemistry cleterrnina~ions of Am. phases of this program are discussed examine differences in Pu:Am in chronological. order: inici.al ratios and Pu–isotope ratios as processing 0[ samples, gamma spec- a function of the type of sample trometry, and wet chemistry . and the location t~here samples were In June 1975 a field survey was conducted on the islands of Bikini ancl Eneu within the Bikini Atoll. During this survey, several hundrecl samples were collected to assess the radiological status of the islands tation by the Bikini people. Instrumental to the radiological. assessment was a thorough and com– prehensive program for the analysis of collected samples. Since many facets of the Bikini program were similar to those employed for Enewetak, we used the excellent discussion of the Enewetak analytical program by lIoff et al? as a source document in the preparation of this report. .4 listing of the samples submitted for analysis is presen~ed More than 950 samples were collected from Bikini and Eneu All samples were processed prior to selection for gamma spectrometry andlor wet chemistry. Of the total samples processed, 624 were counted by gamma spectrometry at LLL on the Ge(Li) detector systems of the Biomedical and Radiochemistry were performed by the McClellan Central. Laboratory (MCL) on 588 of Atomic Energy Commission, Nevada Operations Office, NVO–~40, VOI . 1>

Page 6

Table 1. Analytical program for Bikini samples. Sample type Total collected Gamma c.ountin: Idet chemistry the samples analyzed by gamma spectrometry. All radionuclide concentrations, whether determined by gamma spectrometry or wet chemisCry, were reported to a reference time of 1 Jan 1975 (001.000 Z, 75). All initial processing was primarily of clrying, hornogenizi.ng, and packaging Ehe samples. SoiL and ve~etation samples were dried by heating in ordinary ovens. Ten samples of pig and chicken tissue collected on Bikini were lyophilized. by MCL involved the dissolution of a sample aliquot, chemical separation of the desired elements, and radiation measurement of the elemental. samples. In no case was an entire sample consumed in a single dissolution. All vegetation and animal tissue samples submitted for wet chemistry had been analyzed previously by gamma spectrometry. Separate aliquots from each large soil sample were submitted for wet chemistry and for gamma spectroscopy. Met chemistry was re– quired for ccr~ain nuclides that could not be measured by gamma counting; the majority of these nucl.ides were either alpha or beta emitters. Discussions of the individual quality contro]. programs arc included Reproducil>ility of measurements was examined by stati.sticall.y comparing ratios of the individual measurements of a given isotope. ‘Me mean value and standard deviation of the ratios were then calculated. The siznlfidance of a mean value differing from unity, i..e., indication of possible bias, was testecl by calculating the standard error, SC of the mean II (1.ogarithrnic which i.sbaseci upon the 95 percent

Page 7

confidence level and is obtainecl from mic mean exceeded t“s — the observed s~andard tables, and comparing the bias was said to be significant value of E.stil~ith p. If tl~e logarith- with a 95 percent level of confidence. Initial Processing of Fielci Samples Soil samples, by car the largest category, were treated similarly to those samples obtained during the ‘ 1.972 Enewetak Survey. 1. The treatment consisted of drying, pulverizing, blending, screening, packaging, and preliminary gamma assay, ‘rhree separate aliquots were produced from each soil sample: an aluminum “tuna can” containing 300 to 350 g and two vials containing 50 g each. carefully surveyed for possible filter samples and swipe samples taken from the floors were analyzecl content . There was no detectable contamination. The area was considered suitable for initial processing of soils. This monitoring program continued throughout opera– tion of the facility. Drying ovens designed and l>uilt for the Enewetak Survey were used for strutted of asbestos boarcl with steel blew warm air into each unit, which was equipped with a fan in the vent pipe. I’inal.drying was accomplished in a large commercial drying oven Samples were ground in a l-gal paint can using eight l–in. steel balls. The cover of each can was taped securely; then the entire can was covered with a galvanized-steel jacket tl~at was held in place by two large rubber “0” rings to prevent mil].ing. A maximum of 48 samples could be milled overnight to pzovicle 15 to 24 h of grinding. Packaging, weighing, ancl labeling of samples were perforrnc:clby hand. All work with finely divided soil was performed in fume hoods. Before each samp].e was packaged, clean paper was laid out on the hood bench. Care was taken to prevent cross–con~arnination The following is a detailecl chronological clescription of the operations: l The samples were first unpackaged from the shipping con~alner anti loggecl. The appearance of each

Page 8

organic matter, color, presence of large chunks, etc.). o The samples were transferred to a disposable aluminum cake pan and covered with aluminum foil. Holes were punched in the top of the foil to permit evaporation. 0 The samples were transferred to preliminary drying ovens that were designed to handle about 200 samples. These ovens were set at a temperature of approximately 70”c and operated contiunously. The average residence time per To assure complete dryness, the samples were placed in a second sample residence time averaged The samples were transferred to a l–gal paint can and dry weights were determined. Sample weights varied from 1.00 g to 2 kg. The samples were millecl with eight sample residence time in the ball. mill was between 15 to 24 hr. The soil was screened through 2–mm produce a uniform, homogeneous sample for analysis. The finely ground soil was prepared spectrometry and wet analysis by placing it in EWO different contail>ers. The placed in tightly sealed “tuna num. The large can was 3.9 Ci]]high, 8.3 cm in diameter, with a crosssectional area of 53.8 cm and a volume of 210 cm . The small can was 3.3 cm high, 6.0 cm in diameter, with a cross–sectional area of and a volume of 95 cm . Soil–sample weights in these cans ranged from 1.00 to 375 g. TwO samples for wet–chemistry analysis, each weighing approximately 50 g, were placed in vials. One of the vials was committed to chemical analysis , and the other held as a were assayed for gross gamma counts data gul.ded the scheduling of more precise measurements with Ge(I,i) Flore Lhan 810 samples were processed in the soil-preparation facility between 4 September 1975 and 10 by an average working

Page 9

VEGETATION AND ANIMAL SAMPLES Both plant and animal samples were received frozen with dry ice. The plant samples were spread in stainless steel pans and dried at approximately 80°C for at least 24 h in a forced draft oven until they reached constant weight. The dried plant materials were ground in a Wiley mill with a z-mm screen, pressed into the aluminum “tuna cans” with a Carver press at about 14,000 psi, and sealed. Two sizes of cans were used, one containing insufficient in volume to fill a small can were packaged in plastic vials . Sample weights were logged for calculation of specific activities. Coconut meat, :because of its high oil content, was not ground bllt was broken into small chips and pressed into the aluminum cans. Coconut milk was nixed with formaldehyde and canned. Litter samples were sifted through a 3 l/2mesh screen (5.613–mm openings) before being pressed into cans. The anima.L tissue samples were sliced thinly and freeze dried. Skin and bone were removed from muscle tissue. Freeze–dried tissues were cut inro small pieces and pressed for plant materials. Aliquots for wet chemistry were packed into 30 mm snap top plastic vials. ‘ All gamma measurements of Bikini facility used four Ge(Li) cliodes soil, animal, and vegetation samples ranging from 7 to 19 cm were made by the I{adiochemistry and Biomedical Divisions of LLL. A total. of 624 samples were analyzed, 282 by I<adiocllemistrY and 342 by t:he Biomedical facility. Radiochmnistry used several Ge(Li)–diode detector systems with diodes that were 50 cm or more in volume. The. Biomedical. the more active samples were analyzed for 300 to 400 min. All. gamma spectra were transferred to magnetic tape for analysis on a cI)C-7600 computer using the GAMANAL code. A detailed description of measurement

Page 10

equipment, calibration procedures, and CAFMN.4Lare given in the Enewetak Radiological Survey report. 1 Most of the Bikini samples were packaged in 3.3- or 3,9-cm-high aluminum cans with nominal volumes of 95 and 210 cm , respectively. Isotopic activities are reported as “disintegrations per minute per gram (dpm/g). Eleven nuclicles have been observed in Bikini samples: 241An1 Ifien these radionuclides were not cletected, upper limits were calculated by defining the upper limit photopeak area to be twice the square root of the number of counts observed in the continuum normally occupiect by the photopealc. Thirty-one samples were submitted for comparative measurement to both the Biomeclica] and Radiochemistry facilities. The results arc presented in Table 2. Testing the ratio for bias indicated that there was no significant difference in the results from the two facilities and that on the average, for a series of samples, both would obtain the same result. Statistical variaLion does, of course, exist in the measurement of any individual sample, but for close assessment the average value of many samples is the important factor. in adclition to the interfacility comparison, a series of samples ori– ginally measured in the Biomedical facility was resubmitted to the facility at a later date for comparison of the analytical results. The data are presented in Table 3. Again there is no indication of any statis– tical bias in the data. These comparisons reconfirm the observed duri.n~ the course of the analytical work for the .1972 Enewetak GENE17AL amount of elemental carrier or Wet-chernistry analyses were fication or the desired element, required to quantify a number of gravimetri.c or tracer yi.elcling,and nuclides that could not be determined quantification by an appropriate by gamma spectrometry. Briefly, this techniclue. Table 4 presen~s a list of procedure involves the dissolution of the measured nuclides, their half– a sample in the presence of a known lives , principal radiation, ancl

Page 11

Table 2. Concentrations of 137 Cs in selected vegetation Biomedical vs Radiochemistry gamma detection. and soil samples.

Page 12

Table 3. Concentrations of 137 Cs in selected soil samples. Duplicate counting for gamma detection. Master log Ffeasurement 1 method of measurement. Most of the analyses were for Sr and cent of the samples scheduled for wet chemistry were analyzed for The primary purpose of the 241h1 analyses was for comparison with the gamma-spectrometry results. In the case of some vegetation samples, these analyses fulfilled a secondary role of extending the sensitivity for 241Am detection to lower levels. ‘1’heremaining nuclides in Table 4 were measured in only a small fraction of the samples to provide an indication of their existing levels. Analyses for stable iron, calcium, and strontium were performed on a limited number of samples. Table 5 summarizes the wet-chemistry analyses performed by FIcL. Samples provided for wet chemistry were 50-g aliquots of finely divided coral, 20– to tation, and 50 plus-g aliquots of lyophilized animal tissue. sample aliquot. A separate aliquot

Page 13

Table 4. Nuclides measured in wet-chemistry analyses. Nuclide Half-life Principal radiation Type Of detection 63Ni 92 y !3particle Liquid scintillation counter. daughter proportional counter. 87 y 6 particle Liquid scintillation counter. Fe, Ca, Sr Stable None Mass spectrometry, alpha pulse

Page 14

Table 5. Summary of wet-chemistry analyses. Sample type Number of samples Nuclides analyzed a Total includes 13 samples where duplicate soil samples were analyzed. b Total includes 4 samples where duplicate soil samples were analyzed . was used to determine the and 151 Sm concentrations. For coralline soil and animal Lissue, these aliquots were nominally 5 g. A smaller aliquot of approximately 3 g was taken from vegetation samples. In all cases, samples were ashed at 950°C for 8 h as the initial step in the chemical dissolution. The mination of Sr and Pu from coralline soils, vegetation, and animal tissue is outlined in Table 6. Iron-55 was isolated by passing the working solution, containing iron carrier, through a Dowex l-X8 and elution from a Dowex I-X8 column (Cl- form), and final mounting by aliquot of each sample was ashed and dissolved for elemental analysis of iron by atomic absorption required “to provide corrections to

Page 15

Table 6. Chemistry scheme for determination of 90Sr and Pu in coralline soils, vegetation, and animal tissue. Dissolution Fire coral, vegetation, or animal tissue at 950”C for 8 h. Add HNO ~, boil to oxidize Pu, convert to Cl-. Precipitate Y(OH)3 by adding NH40H; wash precipitate with Extract twice with 10% HDEHPC in Coluene. Precipitate Y oxalate by adding saturate~ oxalic acid, Filter precipitate, dry, fire to Y203 at 900°C, 2 h. Weigh, count 90Y betas. Pu Purification To column eluant add 5M NH20HQHCI, LaC13 carrier, saturated Load on Dowex l-x8 column. Transfer to plating cell; electroplate (for Pu determination by a-pulse-height analysis). b. Transfer to mass spectrometry for filament loading (for Pu determination by mass spectrometric analysis). a 242Pu was used as a tracer for pulse-height analysis (a-PHA) and for mass.-..,. Spectrometry . Note that ‘3BPu could be determined only in that were assayed via &_PHA. b The addition of HI is necessary to insure equilibration of with the plutonium in the working solution,

Page 16

the chemical yields for iron originally present in the samples. Standard chemical procedures were used for the isolation and purificaGravimetric measurement of the recovered nickel and samarium carriers provided yields for the and 151 Sm samples. The addition of 243 Am tracer was required to determine the yield of the Am samples. Nickel was purified by numerous precipitations as nickel dimethylglyoxime. The rare earth, samarium, was separated from americium on a Dowex 50 cationexchange column by gradient e].ution Strontium-90 was determined by beta measurement of the chemically separated from radiochemical contaminants were identified and eliminated through least-squares analysis of the data. These procedures are rather common for determination of the long-lived separately quantified via mass spectrometric measurement techniques. Observation of the characteristic mass-to-charge ratio for each isotope provided the means of separation and measurement. To determine the atom concentrations of in each sample, Pu mass tracer was added during the chemical dissolution. Specific activities were calculated from the measured atom concentrations and appropriate decay constants . Since Pu could not be determined mass spectrometrically, alpha pulse height analyses of Putraced plutonium samples was required. Alpha pulse-height analysis was also essential for the quantification of 241h Chemical yields for the americium samples were determined from the 243Am tracer. Quantification of 238 Pu and 241Am was accomplished by the ratio of the characteristic alpha peak areas to those of the systems were used to measure the All samples …