Review of Neal's Dump Groundwater Monitoring Data

The US EPA is currently conducting a CERCLA 5-year remedy review the Neal's Dump site, Owen County, Indiana. As part of this review, I have reviewed the recent groundwater monitoring data for site. This memorandum will review those data and provide recommendations for future monitoring.

Background

The Neal's Dump site (Figure 1) is located in NW1/4, NW1/4 Section 12, T9N, R4W, Owen County, Indiana. The site was operated as a dump for industrial waste materials, including capacitors containing PCBs, from 1967 to 1972. The site was remediated in the fall of 1998 by removing the waste materials and PCB contaminated soil.

The site is located on a high terrace, or upland area, adjacent to the East Fork of White River. The site elevation is about 590 feet above mean sea level (amsl). The adjacent floodplain elevation of the East Fork White River is about 540 feet amsl. Soil materials at the site consist of 70 to 90 feet of complexly stratified silts, sands, clays and mixed diamicts. Bedrock is limestone assigned to the Ste. Genevieve Limestone of middle Mississippian age.

Surficial soil materials at the site have been interpreted to be wind-blown silts and sands of Pleistocene age. These materials range from 10 to about 25 feet in thickness. The wind-blown silts and sands are underlain by a silty clay unit that is 17 to 25 feet in thickness and that has been interpreted to be a glacial diamict of pre-Wisconsinan age. Beneath the diamict is a glacial sand and gravel unit that is generally 10 to 20 feet in thickness. The top of the unit occurs at about 540 to 550 foot elevation, or 40 to 50 feet beneath the ground surface. Beneath the sand is a 0 to 5 foot thick silt unit that overlies a clay unit. The clay unit has been interpreted to be of lacustTine origin. The clay overlies the limestone bedrock at depths of 70 to 80 feet.

Monitoring Wells

Currently there are four ground water monitoring wells (Figure 1, MW-1D, -2D, -3D and -4D). These wells ring the perimeter of the former dump area. Each is about 50 feet deep and monitors the glacial sand and gravel unit. Water levels measured in these wells indicate that the potentiometric surface in the sand and gravel unit occurs at about 547 to 551 foot elevation, or slightly above the floodplain elevation. The ground water gradient in the glacial sand and gravel unit is very slight, and water level measurements from the wells often show varying ground water flow directions. Generally, well MW-2D, on the east side of the site, has the highest water level, and MW- 1D, on the west side of the site, has the lowest water level. These data suggest a slight westerly gradient to the entrenched White River valley.

Residential Wells

There are several residential wells in the site area; no public water supply is available. The two closest wells at the site are the well at the White residence about 50 feet east of the site and a well about 100 feet southwest of the site that serves the Eads and Hattin residences (Figure 1, Eads Well). The log for the White well (Indiana Department of Natural Resources, Division of Water Reference Number 224132 indicates that the well is 140 feet deep and is cased to a depth of 131.5 feet. The well produces water from the limestone bedrock. No log is available for the Eads well.

There are several wells located at residences on the White River floodplain west of the site. These wells are reported to be completed in sand and gravel alluvium, but some of these wells may be completed in bedrock. No logs are available for any of these wells.

PCB Sample Results

A extensive residential well sampling event was conducted in July 1991. A total of 13 wells were sampled, including the Eads and White wells. No PCBs were detected in any of the well samples.

The White and Eads residential wells have been sampled for PCBs since July 1991. The wells have been sampled on at least a semiannual basis since 1994. No PCBs have ever been detected in the Eads residential well. PCBs were detected in the White well at 0.11 ug/1 on June 12, 1996. This is regarded as a questionable detection based on QA/QC. Confirmation samples from this well were collected by Viacom on July 2 and July 19, 1996, and no PCBs were detected in either sample.

The four existing monitoring wells completed in the glacial sand and gravel at the site have been sampled 39 times since May 1987 (Table 1). This sampling has been on a quarterly or semiannual basis. PCBs have been detected at low concentrations in all of these wells. PCBs have been detected above the US EPA maximum contaminate level (MCL) of 0.5 ug/1 on five occasions. Three of these exceedences have been at MW-4D. The highest PCB value at this well has been 2.2 ug/1. One exceedence each has occurred at MW-2D and MW-3D. No exceedences have occurred at MW-ID, the presumed downgradient most monitoring well. No exceedences of the MCL have occurred since November 2000 at any well. This encompasses eight quarterly or semiannual sampling events.

The PCB detections at monitoring wells have tended to occur in multiple wells at the same time. Examples of this were in May 1991, November 1993, November 1994, and November 2000. These detections are separated by multiple events where detections do not occur in any wells. In some cases, PCB detections are associated with higher than normal TSS levels in the sample, but this is not always the case.

Trend Analysis

Trend analyses of PCB time-series data were conducted for all monitoring wells. The purpose of these analyses were to determine if there were any statistically significant trends in the PCB data. The Mann-Kendall non-parametric test for trend was employed. Calculations were performed using WQStat, a widely used statistical analysis software for environmental data. Non-detect values were assigned a value equal to one-half of the detection limit or 0.05 ug/1. Time series plots, values for the Mann-Kendall statistic, and critical values for various levels of significance are shown in the attached figures.

No significant trend was noted at any of the monitoring wells. In all cases the value of the Mann Kendall statistic is lower than the critical value associated with the 0.2 (two-sided) significance level. The null hypothesis employed for the test (no trend) cannot be rejected at the 0.10 significance level. There a slight decreasing trend at MW-3D and MW-4D, as indicated by the negative values for the Mann-Kendall statistic, but these trends are also not significant at the 0.10 significance level.

The trend test indicates that there is no significant upward trend in the PCB concentration at any monitoring well. A slight, but insignificant, downward trend is noted at MW-3D and MW-4D. Clearly no monitoring well at the site is showing an upward trend that would indicate the continued monitoring is warranted.

Summary and Recommendations

The current status of ground water monitoring at the Neal's Dump site is summarized in the following.

1. The principal water bearing unit currently monitored at the Neal's Dump site is a glacial sand and gravel unit at about 50 foot depth and at 540 to 550 foot amsl elevation. Four existing monitoring wells are completed in this unit, and these surround the site.
2. Several residential wells occur in the immediate vicinity of the landfill. These are believed to be completed in a deeper limestone bedrock aquifer.
3. Detections of PCBs in monitoring wells since 1987 have been sporadic and at low levels. A total of five detections in three wells have exceeded the US EPA MCL value. There are not statistically significant increasing trends at any of these wells.
4. PCBs have been detected only one time in a residential well. This detection is well below the US EPA MCL value, and is regarded as a questionable detection based on QAIQC. PCBs were not reported from this well based on two subsequent confirmation samples.

Based upon this data review, it is recommended that monitoring may be safely discontinued at the site.