Results of the Quarterly Tritium Survey of Four Mile Creek and its Seeplines in the F - and H-Areas of SRS:June 1993 (u)

The En vir on me n tal S c i e n ce s Section (ESS ) e s ta b li s hed a qu arterl y m o n i toring p r o gram o f the F o ur M i le C r ee k see p line down gradient from the F- and H- A rea seepage basins. The program surveys and tracks changes in tritium, specific conductivity, and pH for the seepline water. Measurements from the fifth quarterly survey (June 1993) showed lower tritium and conductivity measurements and higher pH values (pH 5 - 6) than measurements from previous studies. The results of the quarterly surveys suggest that infiltration of rainfall may be diluting and flushing the contaminants from the groundwater system. More measurements are needed to confirm these trends.

activities ranged from 30 to 639 pCi/ml on the east side of surveyed the Four Mile Creek seepline down gradient the drainage and from 91 to 57,500 pCi/ml on the west from the F-and H-Area seepage basins for tritium, specific side. The tritium concentration measured in the stream of conductivity, and pH. The survey was the fifth of eight the natural drainage was 15,300 pCi/ml. Results suggest quarterly surveys for monitoring the movement of contam-that the tritium outcrop area has been encapsulated by the inants from the basins since closure. Surface-water sam-sampling locations established on the west side of the pies were collected from 60 locations along the seepline; drainage. Conductivity and pH measurements taken on Haselow et al. (1990) sampled 44 of the locations, 16were both sides of the drainage were similar and were within on the seepline south of 643-E, which is a decommis-the range of normal values for the area. The low conducsioned area in the Solid Waste Disposal Facility. Samples tivity values measured along the drainage way suggest that were collected from three locations along Four Mile the tritium plume outcropping in the area emanates from Creek.
643-E since wastes introduced into 643-E contained low levels of salt ions compared to the waste in the F-and H-Results indicate that rainfall may be diluting the contami-Area basins. nant plume that leached from the seepage basins and flushing the contaminant plume out of the shallow groundwater The results suggest that contaminant concentrations are in the wetland systems below the basins. ESS found tri-decreasing at the F-and H-Area seeplines and are consistium activities in the F-and H-Area seeplines significantly tent with results from previous quarters (Dixon and Roglower than the activities measured by Haselow et al. ers, 1992;Dixon and Rogers 1993a, b, and c). Areas of (1990). Conductivity measurements exhibited the same stressed vegetation are showing signs of recovery as contrends as tritium activities in both areas, indicating that salt taminant concentrations decrease and pH increases. Aluions are also being removed from the wetland system, minum concentrations measured along the seepline in Concentrations of hydrogen ions also decreased in both 1989(Haselow et al., 1990 were high enough to be potenareas, which indicates that conditions are changing from tially toxic to plants, increases in pH (to about neutral) very strongly acidic to slightly acidic (pH 5 -6), which is will reduce the solubility of aluminum and thereby reduce normal for these wetlands, this _ a possible source of plant toxicity. Concentrations of aluminum, as well as other metals, measured along the The seepline south of 643-E, along a tributary of Four seepline in July 1992 were substantially lower than 1989 Mile Creek, is influenced by tritium from the burial concentrations reflecting the increase in pH (Dixon and grounds. The tributary is a natural drainage that was used Introduction locations on the H-Area seepline south of 643-E were established and have been incorporated into the quarterly Seepage basins in the F and H Area of SRS received low-sampling plan. level radioactive-waste effluent from the chemical separation processes in the general separation areas. The basins Methods retained the effluent to delay its release to the environment. The waste efltuent consisted principally of sodium ESS conducted the fifth round of sampling in the quarterly hydroxide, nitric acid, low levels of various radionuclides, tritium survey in June 1993. Sampling locations were and some metals (Killian et al., 1985a and1985b). Dis-selected that, according to 1989 data, exhibited high and charges of Iritiated water to the seepage basins accounted low values for the three variables of concern. Attempts for most of the radioactivity (Fenimore and Horton, 1972).
were made to establish even ground coverage along both seeplines. ESS collected 60 samples along the seeplines in The Sevannah River Laboratory (now the Savannah River tion Department(HPD) collected andmonitoredsoil samconductivity, and tritium values along the seeplines and pies from several locations along both seeplines for determinedthatcontaminants leaching from the F-andH-gamma radioactivity. HPD did not detect measurable Icy-Area seepage basins were impacting the areas, els of gamma radiationand protective clothing was chosen based on the findings. During sampling operations in the SRS stoppeddischarges to the seepage basins in 1988 and seepline areas, protective clothing consisted of rubber cappedand sealed the basins in 1990 to reduce the release boots and disposable rubbergloves. These measures were of the, contaminants. Scientists hypothesized that after intendedto preventdermalcontact with seepline water. eliminating the contaminant source, annual rainfall amounts and natural groundwaterflow would dilute and Seepline sampling locations were markedand labeled preflushthe remainingcontaminant plume out of the shallow viously with PVC stakes. Water samples were collected groundwaterover time. After the contaminant plume in the within a 3-ft radiusof the PVC stake by boring a hole into shallow groundwateris diluted and flushed out, the wet-the soil with a 3.5 in. soil auger until water was reached. land systems below the basins will recover.
Polyethylene sample containers (25 ml) were dipped into the water until full and then capped to collect waterfor tri-To investigate this hypothesis, a quarterlysampling pro-tium analysis. The outside of each container was then gram was established. ESS sampled 44 of the seepline rinsed with deionized waterandsealed in a small polyethlocations sampled by Haselow et al., (1990) for tritium, ylene bag to minimize the possibility of contamination. pH, and specific conductivity. The Haselow et al. (1990) The small bags were then placed in a large polyethylene results established the baseline to which the results from bag, which was sealed for analysis. The Environmental this study were compared. It was chosen as the baseline Monitoring Section (EMS) performed tritium analysis. since it was the only data availablethat was collected dur-EMS counted aliquots of 5 ml for 20 minutes, which ing the time frame that basin discharges were discontin-yielded a lower detection limit of 1.3 pCi/ml (WSRCued. The I-Iaselow et al. (1990) data should be 3Q1-4, 1992). representativeof conditions duringnormal operation.This sampling program is intended to complement semiannual ESS measured specific conductivity and pH in situ with sampling of the seepline for selected Resource Conserva-conductivity and pH electrodes (WSRC-L14.1, 1992a and tion Recovery Act, Appendix IX, characterization, which 1992b). The electrodes were rinsed with deionized water began in July 1992. A draft report summarizing results aftereach sampling and all sampling equipmentwas thorfrom the first sampling event of the semiannual sampling oughly rinsed with waterat the end of each day. programhas been completed (Dixon and Rogers, 1993d).

Results and Observations
Generally,the locations with the greatest decrease in tritium activity are closest to the seepage basins. In contrast, Concentrationsof variablesmeasured at seepline sampling those further away show little change or even an increase locations fluctuate throughout the year.Climate, especially in tritium activities. These changes indicate that the bulk rainfall, influences measuredconcentrations. Low rainfall of the tritium plume in the wedand system has passed the for a few monthsprior to sampling could concentrate con-locations closest to the seepage basins and is exiting the stituentsand high rainfall amountscould dilute the constit-wetland system at the points closest to Four Mile Creek. uents in the shallow groundwaterat the seepline intercept.
This supports the general hypothesis that annual rainfall Rainfall measured at SRS at the weather station in F Area amounts would dilute the remainderof the contaminant from January through June 1993 was 79.83 cm. From plume emanating from the seepage basins. Due to the 1960 to 1991, the average rainfall in F-Area from the complex hydrologic system present in the basin areas, it is months of January to May was 66.01 cm. This indicates inappropriateto draw final conclusions at this point in the that for the fifth quarterly tritium survey, average rainfall study about why tritium activities are exhibiting these amounts in F and H Areas were above normal. However, trends. Final conclusions should not be drawn until after the surplus in rainfalloccurred primarilyin the first three further sampling is completed and seepline data is cornmonths of the year.This should be consideredwhen inter-paredto groundwaterdata. pretingthe surveyresults.

H Area
'

Figures 3 through 8 show comparisons of 1989, March and
April 1993, and June 1993 tritium, conductivity, and Tritium values in the H-Area seepline ranged from 85 to pH measurements; data for the first four surveys can be 16400 pCi/ml (see Table 2 and Figure4). Of the 22 samfound in Dixon and Rogers (1992, 1993a, 1993b, and piing locations, four had tritium activities elevated above 1993c). Figures 9 through 11 show the data for the loca-background that exceeded the 1989 measurements by tions on the seepline south of 643-E in H Area.These sam-more than 10%.No values exceeded the maximum value piing locations were assigned the identification prefix in 1989 of 24,000 pCi/ml. Figures 5 and 6 show that most FHB. Figures 12 through 14 show the data for the FMC locations either decreased or remained relatively locations, unchanged.These results are generally consistent with the results fromprevioussampling events.

Tritium Measurements
To investigate the hypothesis that June 1993 tritium activities are significantly less than in 1989, ESS conducted a F Area Wilcoxon signed-rank test. This test was chosen because it Tritium values in the F-Area seepline ranged from 13 to allows comparisons of paired data without assumptions of 13200 pCi/ml (see Table 1 and Figure 3). Of the 22 sam-normality (Daniel, 1978). The results showed that the piing locations in F Area, only one had a tritium activity June 1993 concentrations were significantly less elevated above background that exceeded the 1989 mea-(P---0.005)than the 1989concentrations. surements by more than 10%. No values exceeded the maximum value recorded in 1989 of 14,000 pCi/ml. The ESS found the tritium activities in H-Area sampling locagraphs in Figure 3 show that the tritium activity at most tions nearest to the seepage basin exhibited the greatest sampling locations either decreased significantly or decrease. The activity at locations further away had a less remained relatively unchanged. These results are generally significant decrease or increase. These changes indicate consistent with the results from previous sampling events, that the bulk of the tritium plume in the wetland system has passed the locations closest to the seepage basins and To investigate the hypothesis that the June 1993 tritium is exiting the wetland system at the points closest to the activities are significantly less than in 1989, ESS con-creek. Again, due to the complex hydrologic system ducted a Wilcoxon signed-rank test. The Wilcoxon signed-present in the basin areas, it is inappropriate to draw final rank test uses the sign and the magnitude of the rank of the conclusions at this point in the study about why tritium differences between pairs of measurements to compare activities are exhibiting these trends. Final conclusions nonparameu'ic data (Daniel, 1978). This test was chosen should not be drawn until after further sampling is cornbecause it allows comparisons of paired data without plete and seepline data is compared to groundwater data. assumptions of normality. The results showed that the June 1993 concentrations were significantly less (P=0.02) than the 1989 concentrations. 93X2425Jmk Conductivity Measurements recoveryof wetlandspotentiallystressed indirectlyby low pH. Aluminum concentrations measured along the F Area seepline in 1989 (Haselow et al., 1990) were high enough to be potentially toxic to plants. Increases in pH (to about Conductivity measurements in the F-Ar_',_seeplineranged neutral) will reduce the solubility of aluminum and from 36 to 1433 _tS/cm(see Table 1aria Figure 5). Due to thereby reduce this as a possible source of plant toxicity. the extreme variability of conductivity measurements, Concentrations of aluminum, as well as other metals, mea-ESS only considered differences of 100 _tS/cmor more to sured along the seepline in July 1992 were substantially be significant. Of the 22 sampling locations in theF-Area lower than 1989 concentrations reflecting the increase in seepline, only two had measurements of more than pH (Dixon and Rogers, 1993d). Field observations of the 1001aS/cmabove the 1989 measurements. A comparison stressed areas revealed that vegetation in most is of the graphs in Figures 3 and 5 suggests that conductivity recovering. follows the same general trends as the tritium activities. Using a Spearman rank correlation test for nonparametric Four Mile Creek data, it was found that the probability that tritium and conductivity exhibited independent trends was P<0.001. The Figures 9 through 11 show the tritium, conductivity, and rank correlation coefficient was found to be rs = 0.85, sug-pH values for the Four Mile Creek sampling locations and Table 3 provides the data used in the figures. Tritium activgesting that the two parameters are exhibiting the same ities at these locations ranged from 61 to 563 pCi/ml sugtrends. This similarity was expected because tritium traces the movement of the contaminant plume from the basins gesting that seepage from the basins is impacting the (Haselow et al., 1990). These results suggest that the salt stream. Conductivity measurements ranged from 101 to ions are also exiting the wetland system in F Area. 121 I.tS/cmand pH ranged from 5.9 to 7.2. Conductivity and pH values are at near normal levels.

H Area Solid Waste Disposal Facility Conductivitymeasurements in theH-Areaseeplineranged (643-E Old Burial Ground) Seepline
from 39 to 443 I_S/cm (see Table 2 and Figure 6). Of the 22 sampling locations, only two had measurements of Results more than 100 laS/cm above the 1989 measurements.
The graphs in Figures 12 through 14 show tritium, con-Observation of Figures 4 and 6 suggest that conductivity ductivity, and pH values for the seepline and stream samand tritium are following the same general trends. The piing locations south of 643-E, which is part of the Solid Spearman rank correlation test for nonparametric data was used to investigate the correlation of H-Area tritium activ-Waste Disposal Facility. Table 4 provides the data used in ities and conductivity values. Again, it was found that the the figures. This seepline is along the natural drainage that was used to discharge effluent from F-Area separations probability that the two parameters exhibited independent prior to the construction of the engineered effluent canal. trends was P<0.001. The rank correlation coefficient (rs = 0.70) for H Area was slightly lower than in F Area, but still Prior to the forth quarterly survey (March/April 1993) several new sampling locations were established on the west suggested a strong correlation, side of the drainage. These locationswere in addition to those on the east side of the drainage. The new sampling pH Measurements locations were established in an effort to encapsulate the tritium outcrop area along the drainage. F-Area pH values ranged from 3.8 to 6.7 with an average value of 5.1 (Table 1 and Figure 7). H-Area pH values ranged from 5.0 to 6.7 with an average of 6.0 (Table 2 and Tritium activities for the locations on the east side of the Figure 8). The pH for the entire seepline (F and H Areas drainage ranged from 30 to 639 pCi/ml. Activities on the combined) averaged 5.5. The average increased 0.6 units west side of the drainage ranged from 91 to 57,500 pCi/ over the 4.9 average in 1989 (Haselow et al., 1990). While ml. The tritium activity at the stream location in the drainthe average pH remains in the acidic range, the fourfold age (FHB012)was 15,300 pCi/ml. decrease in hydrogen ion concentration further indicates that the contaminant plume is exiting the system. The Conductivity measurements for locations on both sides of decrease in hydrogen ion concentration has ranged from the drainage were near background and ranged from 30 to four to fivefold compared to 1989 measurements in each 339 I.tS/cm. No correlation (rs = 0.14) was found to exist of the past four surveys. The increase in pH will affect the between conductivity and tritium for these locations using solubility of metals in the soil, which will enhance the the Spearman rank correlation test. Values of pH ranged from 4.8 to 6.6 with an average of 5.3.