By David E. Prudic, David A. Stonestrom, and Robert G. Striegl
Core water becomes increasingly depleted in D and 18O from the land surface to a depth of 30 meters, indicating that net evaporation of water is occurring near the land surface. Below a depth of 30 meters the stable-isotopic composition of core water becomes nearly constant and roughly equal to that of ground water. The stable isotopes plot on an evaporation trend. The source of the partly evaporated water could be either ground water or past precipitation having the same average isotopic composition as ground water but not modern precipitation, based on 18 months of record. Profiles of D and 18O in water vapor roughly parallel those in core water. The stable isotopes of core water appear to be in isotopic equilibrium with water vapor from UZB-2 when temperature-dependent fractionation is considered. The data are consistent with the hypothesis of evaporative discharge of ground water at the land surface.
The concentration of tritium in core water from depths less than 50 meters was higher than that of present-day atmospheric air, indicating that elevated tritium concentrations preceded drilling. The concentrations of tritium in core water from the deepest sample (85 meters) and in UZB-2 ground water (110 meters) were below detection. Thus, tritium in the unsaturated zone is not being introduced through ground water.
The shape of the tritium profile for core water was similar to the shape of the tritium profile for water vapor collected April 1994, except that concentrations were consistently lower in core water than in water vapor. Tritium concentrations in water vapor increased from April 1994 to May 1996. Similar to the stable isotopes, the highest tritium concentrations were measured at shallow depths. Concentrations of tritium in water vapor during core collection were estimated assuming isotopic equilibrium with core water. The computed concentrations for November 1992 and September 1993 form consistent temporal trends with subsequent tritium concentrations in water vapor collected April 1994, July 1995, and May 1996. Observations of a bimodal distribution of tritium, in which the highest concentrations are in a gravel layer at a depth of 1-2 meters, indicate lateral migration of tritium through the vicinity of UZB-2.
In September 1993, test hole UZB-2 was drilled to a depth of 114.6 m by the U.S. Geological Survey about 6 m northeast of UZB-1. This test hole was drilled to monitor changes in subsurface gas pressures in relation to atmospheric pressure and to determine the natural distribution of gases for estimating the depth of atmospheric air circulation in the unsaturated sediments (Prudic, 1996).
Core samples were collected at selected depths from both test holes. Sections of the core were analyzed for particle size, bulk density, porosity, water content, water potential, and chloride concentration. Bulk densities, water contents, and chloride concentrations are reported by Prudic (1994). Water potentials are reported by Andraski and Prudic (1997). Both test holes are part of a study to determine the factors affecting water and gas movement through unsaturated sediments at the desert site.
Gas was pumped from test hole UZB-2 in April 1994 and July 1995. Water vapor and carbon dioxide were extracted from the pumped gas. The water vapor was analyzed for the radioactive isotope tritium and the stable isotopes deuterium (D) and oxygen-18 (18O). Carbon dioxide was analyzed for the stable isotope carbon-13 (13C) and the radioactive isotope carbon-14 (14C). Tritium and 14C analyses of the gas samples are reported by Prudic and Striegl (1995). Interpretation of the distribution of tritium and 14C found in UZB-2 is summarized by Striegl and others (1996). Because tritium concentrations in UZB-2 gas samples were greater than expected, a method was developed in early 1996 to extract water from archived sections of cores. Extracted core water was analyzed for tritium, D, and 18O. In addition, water vapor was collected again from UZB-2 in May 1996 and analyzed for tritium, D, and 18O. Laboratory results for tritium in core water and water vapor were received in April 1996 and June 1996, respectively. Laboratory results for 18O were received in April 1996 and October 1996, and results for D for both sets of samples were received in October 1996. Laboratory results for D and 18O for one vapor sample and five water samples used in testing the cryodistillation apparatus were received in November 1996 and January 1997.
Funding for the tritium analyses of water from core samples was provided by the Idaho Operations Office, Secretary of Energy for Environmental Management, U.S. Department of Energy, under Interagency Agreement DE-AI07-94ID13282.
Test hole UZB-1 was drilled to a depth of 48 m and was dry. An initial hole was drilled to a depth of 24 m where the core sampler broke off and could not be retrieved. The hole was backfilled with a mixture of cuttings and bentonite and a second hole was drilled to a depth of 48 m about 2 m to the west. Because of their close proximity, both are referred to as UZB-1. Test hole UZB-2 was drilled to a depth of 114.6 m, about 5 m below the water table. UZB-2 is about 6 m northeast of the second UZB-1 hole. In UZB-2, casing was advanced to a depth of 91 m using the ODEX method. Below a depth of 91 m, UZB-2 was drilled using a rotary method with a tricone bit after the casing broke at a depth of 61 m and could not be advanced further. As in the ODEX method, air was used to remove cuttings. The hole was uncased below 91 m.
A solid tube sampler of hardened steel with an inside diameter of about 10 cm was used to collect core samples in brass liners with individual segment lengths of 7.5 and 15 cm. The sampler was driven into sediments at the bottom of the hole with an air hammer. Approximately 60 cm of core were collected with the sampler at each sampling depth. Core recovery was nearly 100 percent. Brass liners were removed from the sampler using a pneumatic ram that pushed the liners out of the sampler. The liners were immediately capped, taped, and labeled in the field. Samples collected from UZB-1 were sealed in ProtecCore tubular laminate. Samples collected from UZB-2 were capped, taped, then sealed in paraffin. The sample in the uppermost 7.5-cm brass liner of each core contained cuttings from the drill bit and was discarded.
Seven cores were collected from UZB-1 about every 3 m starting at 2.3 m below land surface. No samples were collected below 21.2 m because the solid tube sampler broke off from the down-hole air hammer at a depth of 24.4 m. No core samples were collected from the second UZB-1 hole. A new solid-tube sampler, with a greater wall thickness, was used for test hole UZB-2. Cores were collected from UZB-2 at 26.7 to 27.3 m, 35.9 to 36.5 m, 48.0 to 48.6 m, 60.2 to 60.8 m, 72.5 to 73.1 m, and 84.7 to 85.3 m below land surface. With two exceptions, the deepest 15-cm segment of each core was used for water extraction to minimize impacts from drilling. No core was analyzed from the interval 26.7 to 27.3 m because lubricating oil from the air hammer had coated the outside of the core barrel. Two 7.5-cm segments from the upper part of the 72.5 to 73.1-m core were combined because deeper cores had been used for other analyses.
Ten air ports were installed in UZB-2 (Prudic and Striegl, 1995). Each air port onsisted of a 30-cm-long stainless steel screen connected to 6mm diameter nylon ubing that extended to land surface. Each screen was embedded in a 60-cm interval of fine to medium gravel. A mixture of sand, silica flour, and powdered bentonite was placed above and below each gravel layer. Bentonite grout was pumped into the hole above the mixture of sand, silica flour, and powdered bentonite to seal the hole between screens. The uppermost 2.1 m was filled with cement grout. Water used for the bentonite and cement grouts was obtained from the site well (location shown in fig. 1B). This well was sampled in August 1989 by the U.S. Geological Survey (table 5) at which time the tritium concentration was less than 0.6 TU. The well was sampled on September 16, 1993 (the day after UZB-2 was completed), at which time the tritium concentration was 0.5±2.2 TU (Black and others, 1994, p. D-25). Additionally, the well is sampled quarterly by the site operator. Reported concentrations for samples collected in June, August, and October 1993 were less than the detection limit of 94 TU (Douglas Greffin, US Ecology, Inc., written commun., 1996).
An apparatus (fig. 2) was constructed to extract water directly from sediments in the brass core liners by cryodistillation. The brass core liners were sealed into brass end caps with o-rings. The caps were held in place by bolts (fig. 3A). A heating tape, wrapped around the core liner, was surrounded by thermal insulation (fig. 3B). A thermocouple probe at the center of the bottom end cap penetrated a few millimeters into the lower face of the core sample, measuring internal temperature at that point. An external thermocouple beneath the heating tape measured temperature at a point on the outside of the core liner. A ground glass fitting in the upper brass end cap connected the core assembly to a condenser. Circulating tap water cooled the condenser. The condenser connected to a freeze trap, which consisted of a drip point inserted into a 250-mL collection flask. The flask was cooled by a slurry of dry-ice shavings in isopropyl alcohol (fig. 3C). A vacuum pump removed air from the apparatus, facilitating the transfer of vapor from the core sample to the freeze trap.
The core samples were removed from their moisture-tight packaging and quickly fitted with brass end caps to minimize evaporation and isotopic exchange with atmospheric moisture. After the heater tape and insulation were secured around the liner and caps, the core assembly was installed on the condenser and a dry-ice slurry was placed around the collection flask. Four core assemblies were attached to individual condensers and freeze traps (fig. 4). These were connected in turn to a single, ballasted, vacuum pump. After the four flasks had cooled several minutes, a valve between each freeze trap and vacuum line was slowly opened. After the system reached an absolute pressure of about 500 Pa, the heater control system was activated.
The heater control system applied heat to each core sample when the
external temperature was below
110oC and the internal temperature was below 105oC. Water extraction ended when the internal temperatures of the core samples became nearly constant at about 100oC. This usually required 4-5 hours of heating.
Following water extraction, each core assembly and collection flask were removed from their condenser and dry-ice slurry, respectively. Rubber stoppers were inserted into the glass mouth of each core assembly and collection flask. The sealed flask was immediately placed under running water to melt accumulated ice and prevent the flask from breaking due to differential expansion of glass and ice. The outside of the flask was dried and the flask with stopper and accumulated ice weighed. After all ice had melted, a 14 mL aliquot of water was transferred by pipette from the flask to a glass bottle for analysis of D and 18O. The remaining water was transferred into 50-mL glass bottles for tritium analysis. The core assembly was disassembled and the sediments quickly removed from the brass liner. Sediments were weighed then dried overnight at about 104oC in a forced-convection oven for determination of water remaining in the sediments.
Water vapor was collected by pumping gas at a low rate (less than 10 mL/s) through a freeze trap immersed in a slurry of methyl alcohol and dry ice (Striegl, 1988). Condensed water vapor was collected as ice in the trap. Upon collection of a sufficient volume of water, the ice was allowed to thaw and the resulting liquid was poured into glass vials and sealed. About 10 mL of liquid water was collected every 24 hours. In July 1995, water vapor also was extracted from gas pumped from two hardened-steel probes driven into the ground along the security fence (probes A and B, fig. 1B). The probes were installed at depths of about 1.7 m below land surface. In May 1996, water vapor was extracted from probe B, from a shallow probe driven to depth of 1.68 m less than 1 m from UZB-2, and from a probe driven to a depth of 2.0 m about midway between UZB-2 and probe B (probe C in fig. 1B).
The efficiency of the cryodistillation method was tested by (1) comparing the quantity of water lost during cryodistillation with the weight of ice collected in the flask (trap efficiency); (2) determining the quantity of water that remained in the sample following cryodistillation (residual water); and (3) comparing the isotopic composition of the water added to the sediments with the isotopic composition of the water recovered from the sediments. Residual water is expressed as a percentage of the total weight of water. The total weight of water is defined as the weight lost during cryodistillation plus the weight lost during oven drying after cryodistillation. Trap efficiency and residual water were measured for each extraction.
Results of the tests are summarized in table 1. Trap efficiency for all samples was greater than 97 percent. Residual water was + 0.3 percent. These results indicate that virtually all of the water was extracted from the sediment, and that most of the extracted water was captured in the trap. Negative values of residual water mean samples gained weight during oven drying.
The D and 18O values of water extracted from sediments were about the same as those of added water, with no consistent trends in the differences. The largest difference was an increase in D of 3 permil for the high-tritium water extracted from the 72.5-m sample and a decrease in 18O of 0.4 permil for the low-tritium water extracted from the 48.2-m sample. These differences were less than the two standard-deviation (two sigma) precision of D and 18O determinations.
In the low-tritium experiments, concentrations in extracted water were greater by 3 to 5 TU than the concentration in the added water. In the high-tritium experiments, concentrations in extracted water were lower by 22 to 42 TU than the concentration in the added water. In all but one test (the repeated test on the 72.5-m sample), the observed differences exceeded the two sigma counting uncertainty.
Likely reasons for these differences include isotopic exchange with hydroxyl and other hydrogen-containing groups at mineral surfaces (Halevy, 1964; Gvirtzman and Magaritz, 1986) and isotopic exchange with tightly bound water at mineral surfaces. Additional (but less likely) reasons include fractionation effects due to incomplete volatilization or recovery of water during cryodistillation, contamination during sample processing, isotopic exchange with atmospheric water, contamination by atmospheric water, and cross contamination between samples.
Because of its exceedingly low abundance, tritium is more sensitive to contamination than is D or 18O. The largest observed difference, 42 TU, represents a difference of 42 tritium atoms per 1018 hydrogen atoms. A difference of 42 D atoms per 1018 hydrogen atoms (or 42 18O atoms per 1018 oxygen atoms) would produce a change of less than one billionth of a permil in D or 18O.
Although the reason for the differences in tritium concentrations between extracted water and water added to sediments is not known, the low-tritium experiments agreed within 3 to 5 TU. In contrast, the high-tritium experiments differed by 20 to 45 TU. Because of the higher tritium concentration in the added water, these experiments agreed within 4 to 8 percent. Concentrations of tritium in water extracted by cryodistillation from UZB-1 and UZB-2 sediments thus appear to give reasonable indications of tritium concentrations in the original pore water.
Figure 5 compares the gravimetric water content of UZB-1 and UZB-2 cores determined at the time of cryodistillation (March 1996) with that of adjacent cores determined shortly after sampling (January 1993 for UZB-1 cores and February 1994 for UZB-2 cores). Maximum differences in water content for adjacent cores were +0.023 gram water per gram oven-dry sediment at 8.9 m and -0.013 gram water per gram oven-dry sediment at 11.7 m. The close agreement between the two data sets indicates that water contents did not change appreciably during storage.
Tritium concentrations in table 2 have been adjusted to the appropriate sampling dates to account for decay between the times the core samples were collected and the times when the water extracted from them was analyzed for tritium. Tritium concentrations at time of sample collection (Cc) were computed using the following equation (Robert L. Michel, U.S. Geological Survey, oral commun., 1997):
where Ca is the tritium concentration at time of analysis, in tritium
is the decay constant for tritium, ln(2)/(12.43 years), and
and t is the time from collection to analysis, in years.
The highest tritium concentration, 473 TU, was measured in water from the shallowest core sample, at 2.8 m (fig. 6). Tritium concentration declined to 70 TU in the next core sample, at 6 m, then increased to 278 TU at 18 m before declining again to 4.3 TU at a depth of 72 m. Water from the deepest core sample, at 85 m, had a tritium concentration of 1.3 TU. This was less than the two-sigma counting uncertainty of 2.4 TU (table 2).
Tritium concentrations in water vapor extracted in May 1996 from UZB-2 air ports generally increased from previous samplings (fig. 6). Table 3 lists May 1996 tritium concentrations for UZB-2 along with May 1996 concentrations for samples collected from probes B and C (fig. 1B). The concentration of tritium in a sample collected May 10-17 from the 1.68-m probe at UZB-2 was 20,600 TU. This is more than 15 times larger than that of any other sample collected from UZB-2. These observations indicate the presence of a bimodal distribution of tritium with the highest concentration at a depth of 1.68 m (fig. 6).
The largest observed tritium concentration was from probe B at a depth of 1.8 m near the security fence (fig. 1B). Here the tritium concentration was about 29,000 TU on the basis of three samples collected May 10-17 (table 3). This value was nearly the same as the tritium concentration in a vapor sample collected during July 1995 at the same location (Prudic and Striegl, 1995, p. 4).
The profiles of the stable isotopes of water are simpler than those of tritium. The D and 18O values in core water generally become lower (more negative) with depth (table 2; fig. 7) and are nearly uniform below a depth of 30 m. An exception is the sample from 72.54-72.69 m. This sample was transferred from two 7.5-cm-long brass liners into one 15-cm-long liner before cryodistillation. The paraffin coating and plastic cap on one of the two 7.5-cm core samples was cracked. Evaporation during storage might explain the heavier isotopic composition compared to the other core samples from similar depths. All other core samples were cryodistilled in their original 15-cm long liners and had seals that were intact.
The D and 18O values for UZB-2 water vapor collected in April 1994, July 1995, and May 1996 are listed in table 4. Profiles of D and 18O in water vapor from UZB-2 roughly parallel corresponding profiles in core water but are shifted about -63 permil in D and -9 permil in 18O (fig. 7).
Table 4 also lists D and 18O values for water vapor collected from the shallow probes. These values were similar at all three shallow-probe locations (table 4). The D and 18O values are higher in water vapor from the shallow probes than in vapor from greater depths, indicating water that is isotopically heavier. The higher values indicate that water vapor originates from pore water that is more evaporated than water vapor sampled from deeper depths.
Most of the tritium observed in the profile cannot be explained by infiltration of precipitation at the site. Because of atmospheric nuclear testing, which began in 1952 and ended in 1963, tritium concentrations in precipitation exceeded 1,000 TU during the early 1960's (Yang and others, 1996, p. 25 and 53). Volume-averaged tritium concentrations in precipitation peaked at about 1,800 TU in 1963 and 1964 during which time precipitation at Beatty, Nev., totaled 19 cm (Nichols, 1987, p. 16). If all the precipitation from 1963 and 1964 entered the ground and moved piston-like, completely displacing pre-existing pore water, then at the prevailing volumetric water content of 0.07 cm3/cm3 this would produce a 2.7-m thick interval with a tritium concentration of 340 TU (after accounting for decay). Mixing of precipitation with older or younger water would reduce this concentration. Tritium in the shallowest core sample (473 TU; table 2) exceeded this maximum potential concentration from precipitation.
The relation of D and 18O in core water to that in ground water and precipitation is shown in figure 8. The D and 18O values of the 11 ground-water samples lie in narrow ranges spanning roughly twice the respective analytical precisions. Because analytical precision is only one component of total measurement error, the hypothesis of no difference among ground-water samples should not be rejected.
Water in the deepest core sample (85.13-85.28 m) had D and 18O values nearly identical to those in ground water (figs. 7, 8; tables 2, 5). Ground water from UZB-2 collected at the time of drilling (September 1993) had a D of -107 permil and a 18O of -13.7 permil. Ground-water samples collected in August 1989 from wells MR-3 and the site well (fig. 1B) had a D -108 and -107 permil, respectively, and a 18O of -14.0 permil for both samples (table 5).
Although the isotopic composition of precipitation was not analyzed at the site, 16 precipitation samples collected about 35 km to the southeast from February 1984 through July 1985 were analyzed for D and 18O (Milne and others, 1987; fig. 1A). The mean composition for the 18-month period, weighted by the volume of precipitation, was -77 permil for D and -10.3 permil for 18O. These values are similar to mean annual values reported for a station about 70 km east of the study site (Tyler and others, 1996). The altitude of the study site is 9 m lower than the precipitation station and about 113 m lower than the station reported by Tyler and others (1996).
Values to the right of the meteoric water line (fig. 8) usually indicate evaporation of meteoric water (Coplen, 1993, p. 235). Precipitation during the months of April through October is generally more enriched in D and 18O (that is, heavier, more evaporated) than precipitation during the months of November through March.
Core water at depths less than 30 m is more evaporated (heavier) than core water at depths greater than 30 m. Core water becomes increasingly enriched (heavier) in D and 18O towards land surface (table 2), consistent with evaporative discharge of water at the land surface (Fischer, 1992). The deepest core waters (greater than 30 m) have D values that are less than all but one of the precipitation samples collected during the 18-month period February 1984 through July 1985. If the isotopic composition of precipitation for the 18 month period is representative of modern precipitation, then the deepest core waters are not from modern precipitation. Because the stable isotopes plot on an apparent evaporation trend (fig. 8), the source of core water is ground water or past precipitation of a similar isotopic composition.
Unlike the stable isotopes of water vapor and liquid water, which
appear to be in equilibrium, tritium concentrations in core water are
consistently lower than those of water vapor from UZB-2 (fig. 6).
At equilibrium, water vapor in the unsaturated sediments should have
a lower tritium concentration than liquid water because of
fractionation. The expected tritium concentration in water vapor
(dotted line in fig. 6) was calculated from (Ferronsky and Polyakov,
where Cv is the tritium concentration in water vapor in equilibrium with liquid water, in tritium units;
Cl is the tritium concentration in liquid water, in tritium units; and
HTO is the fractionation factor for tritiated water.
The fractionation factor, for tritium, between liquid water and water vapor varies from 1.108 to 1.103 over the assumed temperature range corresponding to the depths at which core samples were collected (Ferronsky and Polyakov, 1982).
The difference in tritium concentration (that is, the apparent disequilibrium) between core water and subsequently sampled vapor thus could be due to generally increasing tritium concentrations throughout the profile (fig. 9). The appearance of relatively high tritium concentrations that are unaccompanied by corresponding changes in stable isotopes probably reflects the introduction of water having a much greater contrast in tritium than in the stable isotopes.
The highest tritium concentrations, by far, were observed at the shallowest observation depths (that is, less than or equal to 2 m in the shallow probes). The shallow probes were driven into a gravel layer that is present throughout the area. Moisture content in the gravel is about 0.02 gram of water per gram of soil (Fischer, 1992, p. 24). Lateral migration of tritium through the unsaturated sediments from the burial area is the most likely explanation of the tritium concentrations measured at UZB-2 (Striegl and others, 1996). As pointed out by Striegl and others (1996), however, migration of tritiated water is problematic. If water is in isotopic equilibrium with respect to tritium, and if the source of tritiated water is at least 100 m away, then the distribution of tritium at UZB-2 (including the shallow probe sample) cannot be explained by vapor transport alone (Striegl and others, 1996). Advective transport of tritium by liquid-water movement from the burial area should produce an increase in the water content at intervening points. Water content at the 30-m deep shaft facility (fig. 1B) has not changed measurably below a depth of 1 m since 1987 (Andraski and Prudic, 1997). An alternate mechanism for moving tritium from the buried wastes involves diffusive transport by hydrogen gas or other hydrogen-containing gas that can move with little retardation before oxidizing to water (Jalbert and Murphy, 1988).
The possibility that the observed increase in tritium concentration in water vapor might result from one or more artifacts of drilling and sampling is considered next. During drilling, about 25 m3 of air per second at 2.4 MPa was forced down the inside of the drill stem. Although much of this air returned to the surface through the annular space between the casing and the drill stem, some penetrated the sediments at the air hammer. Except for the core from 72.5- to 73.1-m depth, samples selected for water extraction were the deepest 15-cm section of each 0.6-m core and as such were the furthest removed from the point of air introduction during coring and drilling. Even so, the air-filled pore space may have been completely replaced by atmospheric air during sample collection. Furthermore, the advancing drill bit would act as a moving point source of high air pressure. Movement of gas away from this moving point source could smear any pre-existing peaks (that is, reduce tritium gradients with depth) and dilute any high tritium gas near the borehole, at least temporarily. Finally, the energetic hammering required to advance holes in the gravelly sediments may have increased the permeability of the sediments immediately adjacent to the borehole. All of these processes could result in atmospheric air entering pore spaces during drilling and core sampling.
Although some air undoubtedly penetrated the cored sediments during collection, the mass of water in the air-filled pore space is tens of thousands times lower than the mass of liquid water in the core samples, even for the relatively dry conditions at the site. Introduced air would be insufficient to significantly affect the tritium concentration in core samples, especially during a brief exposure. Present-day atmospheric air has a tritium concentration of about 15 TU. If introduced air had significant effects, then the core sample at 85 m would not have had the low tritium concentration observed, which was close to detection (table 2).
The small volumes of gas removed while sampling water vapor were insufficient to cause migration of tritiated vapor from the burial area to UZB-2. At most 4,300 L of gas was pumped from each port during a single sampling period (April 1994, June 1995, and May 1996). Assuming an air-filled porosity of 0.2 (volumetric water content of 0.1 m3/m3), the volume of unsaturated sediments containing the needed volume of gas was 22 m3. If the gas flowed to the air port equally from all directions (spherically), then the distance gas was pumped from the sediments extended 1.7 m from the air port. If gas flowed horizontally and was limited to the vertical thickness of the gravel pack around each air port (0.6 m), then the horizontal distance that gas was pumped from the sediments extended about 4 m from the air port. Both distances are considerably less than the 101 m to the security fence and the 220 m to the nearest burial trench inside the burial area. Moreover, the elevated tritium concentrations found in core water, given that cores were collected prior to pumping, indicates the introduction of tritium at UZB-2 before drilling.
Large changes in tritium concentrations between adjacent core samples and adjacent water-vapor samples in the upper 12 m (fig. 6) suggest that pumping gas from air ports did not introduce tritiated water vapor from the shallow gravel layer at 2 m down a hypothesized high-permeability zone created while installing UZB-2. However, residual effects of injected air could be significant. Pumping of UZB-2 air ports may be removing diluted air, and explain at least some of the continued increase in tritium concentration in water vapor.
If the observed increase of tritium in water vapor is caused by artifacts of drilling and sampling, then the observed tritium concentration in water vapor is not in equilibrium with the concentration in liquid water (fig. 6). Some liquid water may not be readily available for isotopic exchange because the water is adsorbed in tiny pores such as interlaminar spaces in clays. The ordered structure of water caused by hydrogen bonding extends outward from mineral surfaces into the pore space (Dragun, 1988). Exchange of tritiated water requires displacement of nontritiated water molecules, which may be slow because hydrogen bonds among water molecules must be broken for exchange to occur (John Matuszek, New York State Department of Health, written commun., 1996).
Loss of tritium to the atmosphere while core samples were stored prior to water extraction was probably minimal for most core samples. Evaporative loss of water from the core samples during storage would result in an increase, not a decrease, in the concentration of tritium in the remaining water. The relative humidity of the storage cabinet was sufficiently low that if the protective covering on the core samples was compromised, evaporation would have occurred. If evaporation during storage had been significant, then the predicted stable isotopic composition of water vapor from cores would not agree with the measured water vapor from UZB-2. As discussed on page 11 (section titled "Distribution of Tritium, Deuterium, and Oxygen-18 "), however, there probably was some evaporation from the core sample at 72.5 m.
Pore water was extracted from the archived core samples because water vapor collected from test hole UZB-2 in April 1994 and July 1995 had greater than expected tritium concentrations. An apparatus was constructed that extracted water directly from samples in their original brass core liners by cryodistillation. The extracted core water was analyzed for tritium, D, and 18O, and compared to water vapor collected from the air ports.
The D and 18O values in core water become more negative (more depleted, or lighter) with depth and are nearly uniform below a depth of 30 m. Core water at depths less than 12 m is more evaporated than core water at greater depths. The stable isotopes plot on an evaporation trend. The source of the partly evaporated water could be either ground water or past precipitation having the same average isotopic composition as ground water, but not modern precipitation, on the basis of 18 months of record.
Profiles of D and 18O values in water vapor are similar to that in the core water. When fractionation between liquid and vapor is considered, the measured values of water vapor are in equilibrium with that of core water. Values of D and 18O for water vapor from the shallow probe (1.7 m) at UZB-2 indicate that water vapor is heavier (originates from water that is more evaporated) than water vapor sampled from greater depths. Water vapor from two other shallow probes, at depths of 1.8 and 2.0 m, has a similar isotopic composition to the shallow probe at UZB-2. The data are consistent with the hypothesis of evaporative discharge of ground water at the land surface.
Tritium concentration in water from core samples collected at depths less than 50 m was higher than that of atmospheric air that was injected during drilling. Thus, the relatively high (less than 500 TU but more than 15 TU) tritium concentrations in water from core samples must have preceded drilling. The concentration in water from the deepest core sample (85 m) was below detection. Ground water from UZB-2 sampled at the time it was drilled also was below detection. Therefore, tritium at UZB-2 is coming laterally through the unsaturated zone as opposed to coming from ground water.
The shape of the tritium profile from core water is similar to that from water vapor collected in April 1994, except that tritium concentrations are consistently lower in core water than in water vapor. Tritium concentrations in vapor from UZB-2 increased from April 1994 to May 1996. Concentrations of tritium in water vapor that is assumed to be in equilibrium with liquid water from the 1992 and 1993 core samples form consistent temporal trends with measured concentrations in vapor samples collected in April 1994, July 1995, and May 1996. Increasing concentrations of tritium may explain the difference between the earlier core samples and the later vapor samples. The highest concentrations of tritium observed (up to 29,000 TU) were in water vapor from the shallow probes. These were installed in a laterally extensive gravel layer 1-2 m deep. Both core water and water vapor at UZB-1 and UZB-2 indicate a bimodal distribution of tritium with the highest concentrations at shallow depths.
Andraski, B.J., Prudic, D.E., and Nichols, W.D., 1995, Waste burial in arid environments--Application of information from a field laboratory in the Mojave Desert, southern Nevada: U.S. Geological Survey Fact Sheet FS-179-95, 4 p.
Black, S.C., Glines, W.M., and Townsend, Y.E., eds., 1994, U.S. Department of Energy Nevada Operation Office annual site environmental report - 1993, volume II - appendices: Las Vegas, Nev., Reynolds Electrical and Engineering Co., Inc., report DOE/NV/11432-123, v. II, 421 p.
Coplen, T.B., 1993, Uses of environmental isotopes, in Alley, W.M., ed., Regional ground-water quality: New York, Van Nostrand Reinhold, p. 227-254.
Coplen, T.B., Wildman, J.D., and Chen, J., 1991, Improvements in the gaseous hydrogen-water equilibration technique for hydrogen isotope ratio analysis: Analytical Chemistry, v. 63. p. 910-912.
Craig, Harmon, 1961, Isotopic variations in meteoric water: Science, v. 133, no. 3465, p. 1702-1703.
Dragun, James, 1988, The soil chemistry of hazardous materials: Silver Spring, Md., Hazardous Materials Control Research Institute, 458 p.
Epstein, S., and Mayeda, T., 1953, Variation of O18 content of waters from natural sources: Geochimica et Cosmochimica Acta, v. 4, no. 5, p. 213-224.
Evans, E.A., 1966, Tritium and its compounds: Princeton, N.J., Van Nostrand, 441 p.
Ferronsky, V.I., and Polyakov, V.A., 1982, Environmental isotopes in the hydrosphere: New York, John Wiley and Sons, 762 p.
Fischer, J.M., 1992, Sediment properties and water movement through shallow unsaturated alluvium at an arid site for disposal of low-level radioactive waste near Beatty, Nye County, Nevada: U.S. Geological Survey Water-Resources Investigations Report 92-4032, 48 p.
Friedman, Irving, and O'Neil, J.R., 1977, Compilation of stable isotope fractionation factors of geochemical interest (6th ed.): U.S. Geological Survey Professional Paper 440-KK, 12 p.
Fritz, P., and Fontes, J.Ch., eds., 1980, Handbook of environmental isotope geochemistry: New York, Elsevier, v. 1, 545 p.
Gonfiantini, R., 1978, Standards for stable isotope measurements in natural compounds: Nature, v. 271, p. 534-536.
Gvirtzman, Haim, and Magaritz, Mordeckai, 1986, Investigation of water movement in the unsaturated zone under an irrigated area using environmental tritium: Water Resources Research, v. 22, no. 5, p. 635-642.
Halevy, E., 1964, The exchangeability of hydroxyl groups in kaolinite: Geochimica et Cosmochimica Acta, v. 28, p. 1139-1145.
Hammermeister, D.P., Blout, D.O., and McDaniel, J.C., 1985, Drilling and coring methods that minimize the disturbance of cuttings, core, and rock formation in the unsaturated zone, Yucca Mountain, Nevada: Conference on Characterization and Monitoring of the Vadose (Unsaturated) Zone, National Water Well Association, Denver, Colo., November 1984, Proceedings, p. 507-541.
Jalbert, R.A., and Murphy, C.E., 1988, The Canadian experimental HT release of June 10, 1987, U.S. measurements: Los Alamos National Laboratory LA11368-MS, UC-702, 27 p.
Kendall, Carol, and Coplen, T.B., 1985, Multisample conversion of water to hydrogen by zinc for stable isotope determination: Analytical Chemistry, v. 57, no. 7, p. 1437-1440.
Milne, W.K., Benson, L.V., and McKinley, P.W., 1987, Isotope content and temperature of precipitation in southern Nevada, August 1983-August 1986: U.S. Geological Survey Open-File Report 87-463, 32 p.
Nichols, W.D., 1987, Geohydrology of the unsaturated zone at the burial site for low-level radioactive waste near Beatty, Nye County, Nevada: U.S. Geological Survey Water-Supply Paper 2312, 57 p.
Prudic, D.E., 1994, Estimates of percolation rates and ages of water in unsaturated sediments at two Mojave Desert sites, California-Nevada: U.S. Geological Survey Water-Resources Investigations Report 94-4160, 19 p.
-----1996, Water-vapor movement through unsaturated alluvium in Amargosa Desert near Beatty, Nevada--Current understanding and continuing studies, in Stevens, P.R., and Nicholson, T.J., eds., Conference on disposal of low-level radioactive waste, Reston, Va., May 1993: U.S. Geological Survey Water-Resources Investigations Report 95-4015, p. 157-165.
Prudic, D.E., and Striegl, R.G., 1995, Tritium and radioactive carbon (14C) analyses of gas collected from unsaturated sediments next to a low-level radioactive-waste burial site south of Beatty, Nevada, April 1994 and July 1995: U.S. Geological Survey Open-File Report 95-741, 7 p.
Striegl, R.G., 1988, Distribution of gases in the unsaturated zone at a low-level radioactive-waste disposal site near Sheffield, Illinois: U.S. Geological Survey Water-Resources Investigations Report 88-4025, 69 p.
Striegl, R.G., Prudic, D.E., Duval, J.S., Healy, R.W., Landa, E.R., Pollock, D.W., Thorstenson, D.C., and Weeks, E.P., 1996, Factors affecting tritium and 14Carbon distributions in the unsaturated zone near the low-level radioactive-waste burial site south of Beatty, Nevada: U.S. Geological Survey Open-File Report 96-110, 16 p.
Thatcher, L.L., Janzer, V.J., and Edwards, K.W., 1977, Methods for the determination of radioactive substances in water and fluvial sediments: U.S. Geological Survey Techniques of Water-Resources Investigations, book 5, chap. A5, 95 p.
Tyler, S.W., Chapman, J.B., Conrad, S.H., Hammermeister, D.P., Blout, D.O., Miller, J.J., Sully, M.J., and Ginanni, J.M., 1996, Soil-water flux in the southern Great Basin, United States: Temporal and spatial variations over the last 120,000 years: Water Resources Research, v. 32, no. 6, p. 1481-1499.
Yang, I.C., Rattray, G.W., and Yu, Pei, 1996, Interpretation of chemical and isotopic data from boreholes in the unsaturated zone at Yucca Mountain, Nevada: U.S. Geological Survey Water-Resources Investigations Report 96-4058, 58 p.