Bomb-Pulse Chlorine-36 at the Proposed Yucca Mountain Repository Horizon: An Investigation of Previous Conflicting Results and Collection of New Data

Abstract

Previous studies Los Alamos National Laboratory (LANL) found elevated ratios of chlorine-36 to total chloride (36Cl/Cl) in samples of rock collected from the Exploratory Studies Facility (ESF) at Yucca Mountain (YM). The data were interpreted as an indication that fluids containing “bomb-pulse” 36Cl reached the repository horizon in the ~50 years since the peak period of above-ground nuclear testing. Due to the significance of 36Cl data to conceptual models of unsaturated zone flow, the United States Geological Survey (USGS) implemented a study to validate the LANL findings. The USGS drilled new boreholes at select locations across zones where bomb-pulse ratios had previously been identified. The drill cores were analyzed at Lawrence Livermore National Laboratory (LLNL). Because consensus was not reached between the USGS/LLNL and LANL on several fundamental points including the presence or absence of bomb-pulse 36Cl, an evaluation by the University of Nevada, Las Vegas (UNLV), was initiated. The overall objectives of the UNLV study were to investigate the source of the validation study’s conflicting results, and to obtain additional data on bomb-pulse isotopes at the repository horizon. UNLV engaged in discussions with previous investigators, reviewed reports, and analyzed archived samples. UNLV also collected new samples of rock from the ESF, soil profiles from the surface of YM, and samples of seep water from inside the ESF. Samples were analyzed for 36Cl/Cl ratios, and 99Tc and 129I in select samples. A column experiment was conducted mimicking the passage of bomb-pulse 36Cl through YM tuff. The work faced several obstacles including an extended shutdown of the tunnel. Only one sample yielded a background corrected 36Cl/Cl ratio that was higher than the accepted bomb-pulse threshold (1250 x 10-15). Specimen 01034214 obtained from the Drill Hole Wash fault (19+33) had a ratio of 1590 ± 80 (1σ) x10-15, whereas the other separate sample from this fault zone yielded 1160 ± 50 (1σ) x 10-15. Three samples collected from Alcove 6 averaged 490 ± 100 (1σ) x10-15; a sample from Sundance Fault resulted in a ratio of 920 ± 60 (1σ) x10-15, and a sample from the Bow Ridge Fault produced 530 ± 20 (1σ) x10-15. The results are significant because: 1) they tend to be lower than LANL data for comparable samples, albeit in agreement with the range of data produced in the area, and 2) they show that a bomb-pulse 36Cl/Cl ratio was measured in rock collected at the repository horizon level by a second and independent group of investigators (UNLV). Because of time UNLV was not able to replicate the results, and these few data points are insufficient to draw major and definitive conclusions. Leachates of soil samples collected from the surface above the ESF yielded several ratios with bomb-pulse 36Cl, particularly for samples encompassing the wetting front. Soil samples collected above the south ramp, where there was limited soil coverage due to a large amount of rock outcrop, had relatively large ratios ranging from 2170 ± 110 (1σ) x10-15 to 5670 ± 350 (1σ) x10-15. Soil samples from profiles from above the north ramp ranged from 820 ± 70 (1σ) x10-15 to 2390 ± 160 (1σ) x10-15, which compare favorably with previous measurements near the site. Water seepage into the ESF south ramp and 36Cl standards made from NIST material were also analyzed. The standards were produced to have nominal 36Cl/Cl ratios (10-15) of 500, 2,500 and 10,000 and the results showed good agreement with the calculated ratios. The seepage samples ranged between 680 ± 40 (1σ) x10-15 to 1110 ± 40 (1σ) x10-15, consistent with that found for modern meteoric water, with a small bomb-pulse component. Bomb-pulse 36Cl may not have been incorporated in this fast-path water because the surface above the infiltration zone consists mostly of outcrop and the flow pathways have probably mostly been leached. 99Tc was measured in five of nine leaches of ESF rock but poor analytical recoveries and lack of data overlap with 36Cl limit interpretations of these data. The detection capability of the ICP-MS was insufficient for measuring 129I without preconcentration, and detection by AMS may be preferable. Experiments were conducted using bromide instead of chloride as a carrier, which is advantageous because it may eliminate the need for blank subtraction. ESF samples prepared using bromide had 36Cl/Cl ratios (x10-15) with acceptable levels of uncertainty (1σ): 720 ± 30 and 1250 ± 90 for Bow Ridge and Drill Hole Wash Faults, respectively. The result for the latter is noteworthy because it constitutes a second detection of a 36Cl/Cl bomb-pulse ratio at the site, albeit from a non-Q measurement. The source of the conflicting results between USGS/LLNL and LANL could not be determined. There was no evidence that the different AMS facilities were a source of the discrepancy between the results.