Introduction

    The water wells in the immediate vicinity of the WIPP site provide water primarily for livestock, industrial usage by oil and gas production operations, and monitoring studies conducted by various groups. From April to August 1998, water samples were collected for the CEMRC's environmental chemistry studies from six sources in the vicinity of the WIPP: Loving, Otis, Carlsbad, Private Well #2, WIPP-Double Eagle, and Hobbs (Table 11).

    Aquifers in the region surrounding the WIPP include Dewey Lake, Culebra-Magenta, Ogalalla, Dockum, Pecos River alluvium and Capitan Reef. The main Carlsbad water supply is the Sheep Draw well field whose primary source is the Capitan Reef. The Hobbs and WIPP-Double Eagle water supplies are drawn from the Ogalalla aquifer, while the Loving/Malaga and Otis supply wells draw from deposits that are hydraulically linked to the flow of the Pecos River. The source for the sampling site designated as Private Well #2 is a private well seven miles southwest of the WIPP; this water is drawn from the Culebra aquifer.

    The analyses of water samples reported herein continues the baseline evaluation that began in 1997, for which preliminary results were reported in the CEMRC 1997 Report. As in all of the other WIPP-EM studies, the first priority of this work is to establish baseline concentrations for substances of environmental concern as a result of operations at the WIPP facility. A secondary objective for the non-radiological studies reported here was to validate the analytical methods through an inter-laboratory calibration.

Methods

    Nine L of water were collected at each source for analyses of various non-radiological constituents and water quality parameters. All samples were collected following purging of the sources for approximately 5 min or at least 50 L.

    Samples for analyses of SDWA constituents were sent to the Soil Water and Air Testing (SWAT) laboratory at NMSU. All samples for the SWAT laboratory were collected according to EPA protocols for the collection, handling and preservation of drinking water samples.

    For the inter-laboratory comparison, separate aliquots of all drinking water samples collected in 1998 were analyzed by CEMRC and the SWAT laboratory. The analyses performed by CEMRC were done by using ion chromatography, inductively-coupled emission spectrometry, and atomic absorption spectrometry as summarized in Appendix K. A de-ionized water blank also was analyzed by both groups as part of this comparative study. This inter-laboratory comparison provides a means of evaluating the developmental efforts at CEMRC over the past year.

Results and Discussion

    Analyses were performed by the SWAT laboratory for several general categories of substances, including metals, volatile organic compounds, semi-volatile organic compounds, and general secondary water quality parameters. The analytes chosen for study were those regulated under the Safe Drinking Water Act (SDWA), plus selected compounds and elements identified as possible constituents of wastes to be deposited in the WIPP. For constituents regulated under SDWA, primary and secondary maximum contaminant levels (MCLs) are referred to as "reference levels" to provide readers with a basis for comparison. However, the results presented are not appropriate for use as evidence of compliance or non-compliance with any regulatory requirements. Instead, these results are intended to provide only a general indication of the chemical composition of the drinking water sources prior to the opening of the WIPP.

    The overwhelming majority of inorganic analytes were below detectable levels in the drinking water samples, and only a few of the organic analytes were above their detection limits. Analyses of the 1997 samples indicated the presence of bromoform in samples from Hobbs, Otis, and a private tap five miles northwest of the WIPP (designated Private #1). Bromoform was not detected in the 1998 Otis sample. Private #1 was not sampled in 1998, but the water system for Private #1 is the same as for the 1998 WIPP-Double Eagle sample, which showed no bromoform present. Interestingly, bromoform was present in the 1998 Hobbs sample at about the same concentration as in the 1997 sample (9 mg L^-1 in 1998 vs. 10 mg L^-1 in 1997). In 1997, dibromochloromethane was detected in a single sample from Hobbs (collected 16 July 1997), where the concentration was 1.1 mg L^-1. This finding was confirmed by the analysis of the 1998 Hobbs sample, which had exactly the same concentration. Despite their presence, the levels observed for bromoform and dibromochloromethane were below reference levels.

    To evaluate trends in concentrations over time, the data for three pairs of drinking water samples collected in 1997 and 1998 (i.e., Loving, Otis, and Hobbs) were compared and contrasted. Those pairs of samples were collected at about the same time of year, thus eliminating the possible effects of seasonal trends. The concentrations of almost all analytes in the sample pairs were in close agreement. Of the 80 pairs of values compared, there were only a few cases in which the differences between years were a factor of 5 or greater. For the Hobbs samples, the Pb concentration was a factor of 10 lower in the August 1998 sample compared with the July 1997 sample (0.8 vs. 8.7 mg L^-1 ). In contrast, the Tl concentration was more than four-fold higher in the 1998 sample compared to the 1997 sample (0.26 vs. 0.06 mg L^-1). Ammonium (as nitrogen) concentrations also differed substantially in the 1997 and 1998 Hobbs samples, with concentrations of 0.02 and 0.2 mg L^-1, respectively.

    At Otis, the K concentration in 1997 was 4.5 mg  L^-1, but in the 1998 sample, K was less than the detection limit of 0.4  L^-1, representing more than a factor of 10 change. It is also noteworthy that Tl was detected in the 1998 Otis drinking water sample but not in the 1997 sample. Less remarkable was the detection of Kjeldahl nitrogen and total P in the 1998 Otis sample. There were no particularly noteworthy differences in the Loving sample for 1997 compared with 1998.

    As in 1997, several inorganic non-radiological substances exceeded reference levels (secondary maximum contaminant levels) in the 1998 samples from the Otis and Private #2 sources. Specifically, these were chloride (by autoanalyzer), sulfate, and total dissolved solids. The cited reference levels for these analytes are non-enforceable guidelines (secondary MCLs) under the SDWA. The fact that the 1998 and 1997 data showed essentially identical patterns for these analytes indicates that the high concentrations were not spurious results but rather a true indication of elevated levels. Compared with the other sites, the Otis and Private #2 drinking water samples also (1) had relatively high concentrations of Ca, Mg, Na, and Cl^-; (2) were relatively hard, and (3) exhibited high electrical conductivity. All of these factors are consistent with a high mineral content in the Otis and Private #2 drinking water sources.

    Some differences between the CEMRC and SWAT laboratory Ca and K concentrations were observed in the inter-laboratory comparison. The average relative percent difference (RPD) for the Ca concentrations was 24%, with the CEMRC concentrations higher than the SWAT values in five of six cases. The RPD for K was 70%, with the CEMRC concentrations lower than SWAT for all six samples. It is interesting and noteworthy that in all cases, the CEMRC data were closer to the 1997 SWAT lab data than the 1998 SWAT lab data. Why this should be the case is unclear, but as demonstrated in the QA section, the CEMRC instrumental methods for the analysis of Ca and K (as well as Na and Mg) have been at least partially verified through an intra-laboratory calibration exercise in which the ICP-ES and IC data were compared. As discussed elsewhere in Appendix K, the results of this intra-laboratory comparison showed that the instrumental analyses by the CEMRC are likely to be quite accurate for Ca, K, and other alkali and alkaline earths.

    With respect to heavy metals, the Cu concentration in the 1998 Hobbs sample appeared to be higher than at the other sites, and this same pattern was also noted in 1997. In contrast to the 1997 results, however, the Pb concentration at Hobbs was not higher than at the other sites but instead roughly comparable; there was a ten-fold drop in the Pb concentration for the 1998 Hobbs sample relative to 1997. While the Cu and Pb concentrations were well below reference levels in all samples, some differences did exist between the CEMRC and SWAT data for these and several other heavy metals and metalloids as discussed below. As was the case in 1997, the highest concentrations of Ni, Se, and Tl were found in drinking water from Private Well #2, but as before, those concentrations were well below reference levels. Again, it is important to emphasize that these results are not appropriate for use in assessments of regulatory compliance.

    The inter-laboratory comparison showed consistent results, if not close agreement, for 14 of 20 analytes (relative percent differences of less than 30%). For Cu, the between-laboratory differences are skewed by the Hobbs 1998 sample, where CEMRC reported a concentration of 2.0 compared with 15 mg L^-1 reported by SWAT. As noted above, the SWAT data indicated that Cu was substantially higher at Hobbs than at the other sites. On the other hand, if the lower CEMRC value for Hobbs is compared with the other data, the Cu concentrations would be similar at all sites, ranging from 1.8 to 4.1 mg L^-1. Patterns in the between-laboratory differences in the Cu data are not clear cut. Although the CEMRC data for Hobbs are more in line with the other sites, the Cu concentrations determined by SWAT for Hobbs were high in both years, even though the reported concentration was substantially lower in 1998 than in 1997 (15 versus 56.1 mg L^-1).

    For both Ni and Pb, the CEMRC reported "less than" values that were lower than the concentrations reported by SWAT for the Hobbs water sample and the blank. The first point to keep in mind with respect to these results is that the concentrations of both elements in the drinking water samples were orders-of-magnitude below the respective MCLs. The second point that bears mention is that a preservative was placed in the SWAT sample containers before sampling while none was used for the CEMRC samples. Therefore, one possible explanation for the observed differences is that small but detectable amounts of these metals were introduced into the SWAT samples with the preservative. However, an equally valid explanation is that these metals were lost to the walls of collection containers because CEMRC did not acidify samples immediately after collection. Follow-up studies on preservatives will be conducted in the upcoming year.

    For Hg and Sb, CEMRC reported concentrations lower than the "less than" values reported by SWAT. These results are not inconsistent, and they are the result of the very low detection limits made possible by the flow injection system used for the analysis of these elements at CEMRC.

    Differences in the Se data are less easily explained, although contamination from the preservative used by SWAT or sorption to the walls of the CEMRC sample containers are still possible explanations. For Se, the CEMRC concentration for the Hobbs sample was < 0.2 mg L^-1 compared with 10 mg L^-1 for SWAT. In all other samples, the CEMRC reported below detection limits (< 0.2 mg L^-1) while SWAT reported concentrations ranging from 1.2 to 10.2 mg L^-1. Results for arsenic in one of the six samples, (the WIPP-Double Eagle sample) are similarly discrepant. The flow injection analysis accessory system used for the As and Se analysis at CEMRC eliminates most matrix interferences, and therefore the sensitivity and precision for these elements should be quite good. In comparison, the ICP-mass spectrometric methods used by the SWAT laboratory are more subject to matrix interferences for these elements. Follow-up studies of the As and Se analytical methods will be undertaken in the coming year.

    The data for fluoride, chloride, nitrate (as N) and sulfate agreed well between laboratories, with RPDs generally below 25%. Of these anions, the fluoride concentrations showed the greatest discrepancies. The configuration of the ion chromatograph at CEMRC is such that the quantitation of F^- is difficult. As F^- is usually not of great concern, this difference between laboratories is not considered a major problem.

    In summary, the analysis of drinking water samples in 1998 showed remarkable consistency with the results of the previous year’s analyses based on SWAT data. Few organic contaminants were detected and those inorganic substances that were quantified were, with a few exceptions, below reference levels. Finally, with the exception of the Se analyses and some questions regarding sample preservation, the CEMRC data appear to be well validated based on the inter-laboratory comparisons with SWAT.