• There is a high degree of hydraulic connection at 795 to 800 feet amsl.
• At 795 feet amsl, there is high bulk conductivity and low storativity.

2.2.3 Seismic Tomography Investigation - August 2001

Based on the results of the short-term pumping tests, Viacom postulated the existence of a PCB-carrying conduit in the vicinity of MW-4I. To locate the conduits, seismic tomography investigations were conducted between MW-4I and MW-17 and between MW-4I and MW-18 in August 2001. Prior to conducting these investigations, a previously installed bentonite plug was removed from MW4I, opening the well to the 795 to 800 feet amsl elevation. The seismic tomography indicated two anomalies in the vicinity of MW-4I.

Viacom installed MW-20 and MW-21 at the locations of both anomalies identified in the seismic tomography investigation (see Figure A-1). MW-20 is located about 5.5 feet south of MW-4I and was drilled to 793 feet amsl. A number of small fractures or voids were encountered in the borehole. Water was first encountered at about 802 feet amsl. No water samples were collected during drilling.

MW-21 is located about 9 feet north of MW-4I and was drilled to 793 feet amsl. A number of small fractures or voids were encountered in the borehole, including a 1O-inch-tall open void at about 800 feet amsl. Water was first encountered at about 805 feet amsl. Upon completion of drilling, the well was producing an estimated 10 to 20 gpm of water. No water samples were collected during drilling.

Figure A-8 is a geologic cross section along the eastern perimeter of the landfill compiled from Viacom field logs. The cross section incorporates the newly installed phreatic zone wells along the east side of the landfill. As shown in the cross section, the majority of larger voids observed along the east side of the landfill are to the north in the MW-4 wells, MW-20, and MW-21. Figure A-9 is a geologic cross section along the southwest edge of the landfill. This figure includes the new MW-15 and piezometer PZ-F. Significant epikarst voids were encountered at LF-6.

Based on the results of the short-term pumping tests, Viacom selected wells MW-16 and MW-21 to conduct long-term pumping tests (see Sections 2.2.5 and 2.2.6).

2.2.4 Multiple Tracer Dye Injections - October and November 2001

The first phase of the dye injection tests involved three tracer dye injections in monitoring wells and piezometer LF-6. Attachment B provides a copy of Viacom's work plan for the dye tracer test. Tracer dyes were injected on October 31. Various wells and spring monitoring points were sampled for periods of 2 days to several weeks. Tracer dye background concentrations were determined at several monitoring points prior to the test. The goals of the tracer tests were to determine (l) if any of the wells surrounding the landfill were on or near a major conduit that would integrate all three dyes into one flow path toward the ICS; (2) travel times from various areas of the landfill to the ICS; and (3) if the Quarry B spring, located downstream of the Illinois Central Swallow Hole, would receive dye while flow was stopped to the Illinois Central Swallow Hole (see Figure A-10).

Three different fluorescent dyes were injected on October 31, 2001, at the locations below (see Figure A-l).

• Eosine (Acid Red 87) was injected southwest of the site at NN-700 at 10:15 a.m. NN-700 is open to the 795 to 800 feet amsl zone of solutional development.
• Phloxine B (Acid Red 92) was injected at SP-1 at 11:15 a.m. This well is also open to the 795 to 800 feet amsl zone.
• Fluorescein(Acid Yellow 73) was injected at LF-6 (8-inch-diameter well) at l l :50 a.m. Dye was injected into the Groundwater recovery gallery that was placed during the RA in 2000. The bottom of the gallery is located at about 840 feet amsl.

2.2.4.1 Background Tracer Dye Analysis

Table B-4 shows the sample designations, locations, and date and time of each sample collected. "Scale" refers to SSFP instrument settings (Excitation [Ex]/Emission [Em]) slit widths (nanometers [nm]) used in the SSFP analysis (more width means more light to the detector and greater sensitivity). "Lambda" refers to the wavelength (nary) of any observed fluorescent emission peaks. Tracer dyes are identified by the wavelength of the fluorescent emission and may be quantitated by the height of the fluorescent peak (see "Response" column of Table B-4). Equivalent dye concentrations (in micrograms per liter [ug/L]) are shown in the second to the last column and are based on a comparison to SSFP calibration curves. The identity of the tracer dye is shown in the last column.

Definitive fluorescent peaks at about 512 rim are regarded as fluorescein (Acid Yellow 73). Eosine (Acid Red 87) fluoresces at about 536 nm. The phloxine B emission peak is at about 556 nm. Rhodamine WT (Acid Red 388), which was not used in this investigation, fluoresces at about 576 nm. Recent research indicates that this dye may undergo chemical degradation reactions and alter to a compound with a fluorescent peak at about 522 nm (Davies 2001); therefore, the presence of Rhodamine WT used during previous investigations was identified during the background investigation

Sources of background fluorescence include both natural sources and anthropogenic sources such as dyes used in shampoos, radiator coolant, hydraulic fluids, colored paper, and felt tip pens. However, the largest single source of background fluorescence observed during this investigation is residual dye from previous groundwater tracing dye studies performed around Lemon Lane Landfill. Previous injections in the vicinity of the landfill have included the following dyes:

• Fluorescein at MW-10(1989)
• Rhodamine WT at MW-lO,MW-lS, and MW-7 (1990)
• Fluorescein and Rhodamine WT injected into swallow holes around Lemon Lane Landfill (1992)
• Fluorescein near MW-7 (1996)
• Rhodamine WT in a swallow hole north of the landfill (1996)

ICS had background fluorescence in the fluorescein and Rhodamine WT range in all samples collected. Rhodamine WT dye was identified from previous site investigations, but concentrations were not quantified or reported by Viacom (2002c) because this dye was not used during the background investigation. Fluorescein was also detected in two background samples at the Quarry B spring and one background sample at the swallow hole seep.

Minor fluorescence in the Fluorescein range was apparent in MW-15, NN-625, O0-300 and 00-370. These results were all less than 1 ug/L. A much higher apparent Fluorescein concentration was noted at MW-10, apparently from the 1989 tracer injection.

Several monitoring wells also had high Rhodamine WT concentrations. Based on SSFP response, the highest concentration was at MW-10 and other wells near MW-lO(MW4S, MW~I, MW-17,MW-20, and MW-21). The apparent Rhodamine WT degradation product, which fluoresces at about 522 nm, was also noted in several of these wells.

A small eosine peak was noted in one background sample at MW-6 and in a background sample collected at MW-21 immediately prior to the dye injection on October 31. Eosine was not apparent in any other background sample.

2.2.4.2 Tracer Dye Injection and Monitoring

The tracer dye test was conducted to determine if there is a common underflow connection from the Lemon Lane Landfill conduit system to both ICS and the Quarry B spring. Both springs contain PCBs. Each tracer dye injection consisted of 300 grams of dye mixed in 1 gallon of water. For the NN-700 and SP-1 locations, the dye was pumped into the well using discharge tubing installed down the well casing to 795 feet amsl. The dye mixture was poured down the well casing at LF-6 (the 8-inch-diameter well). After injections, llO gallons of water was injected into the well to flush the dye from the well bore hole.

Monitoring for the tracer dyes was conducted at several wells along the east and south sides of Lemon Lane Landfill and at four springs, the ICS, and the Quarry B and Slaughterhouse springs, and a seep in the ICS branch upstream of the Illinois Central Swallow Hole (see Figure A-10). Viacom collected samples along the east and north sides of the landfill from MW-2, MW-4S, MW~I, MW-10, and MW-15 through MW-21. EPA collected samples open to the 795 to 800 feet amsl zone south of the landfill from MW-6, NN-12, NN-300, NN-625, 00-125, 00-300, 00-370, 00-387, and 00-587. Based on initial sampling results, several other wells south of the landfill were sampled on November 1, 2001, including MS-1, NN-12A, NN-300A, 00-125A, 00-300A, and 00-587A. Several samples were collected from each well. In addition, single samples were collected from KK-112, II-87, II-112, MW-B2, and MW-ll. These wells are located about a quarter mile southeast of the landfill (see Figure A-1). All monitoring well samples were collected using dedicated bailers during this phase of the testing.

Samples for dye tracer analysis were collected at ICS on a 1-, 2-, or 4-hour basis. Under low-flow conditions, water discharging at ICS flows south a short distance and sinks at the upstream-most Swallow Hole seep (see Figure A-10). This flow is known to reappear a short distance downstream at the Quarry B spring. Tracer dyes and PCBs emergent at ICS follow this pathway and also emerge at the Quarry B spring. Because the spring flow to the swallow hole is continuous, it has not been possible to definitively establish by tracer dye test if a separate direct underflow conduit to the Quarry B spring exists. Such a conduit would allow PCBs to bypass the ICS treatment plant and emerge downstream at the Quarry B spring.

During the tracer test, the ICS treatment plant was shut down and all water from ICS was pumped to storage to stop the flow of water and tracer dye into the swallow hole seep. Flow also enters the swallow hole from several seeps in the ICS branch upstream from the swallow hole seep and downstream from the ICS treatment plant intake (see Figure A-10). To stop this flow as well, EPA installed pumps upstream of the swallow hole and pumped all water entering the swallow hole to treatment plant storage. Therefore, all discharge into the swallow hole seep was stopped for the initial portion of the tracer dye test and any dye emerging at Quarry B spring would indicate that the spring is supplied, at least in part, by an underflow conduit system directly from the Lemon Lane Landfill. Conversely, the absence of dye would conclusively demonstrate that no underflow connection exists and that the PCBs detected at the Quarry B spring could be attributed solely to the discharge of PCB- contaminated water into the swallow hole seep or to residual contamination remaining in the section of karst conduit between the swallow hole and the spring. Samples for dye analysis were collected at the pump location upstream of the swallow hole seep and at Quarry B spring for a period of about 36 hours after dye injection.

Periodic samples were also collected from the Slaughterhouse spring during the test. Water samples collected from all wells and springs were submitted to the Indiana University Department of Geological Sciences for SSFP analysis as conducted for the background samples.

2.2.4.3 Multiple Tracer Dye Injection Results

Tracer dye test results are discussed below by dye type.

Fluorescein tracer dye was injected into the epikarst at 840 feet amsl at piezometer LF-6 (8-inch-diameter well) in the southeastern portion of the landfill (Viacom 2002c). The tracer dye was injected about 25 feet above the phreatic zone. The epikarst in this area is known to contain high concentrations of PCBs. Free product PCB oil was recovered from LF-6 when the well was installed. This area of the landfill may be a significant source of the PCBs released at ICS; therefore, tracing the flow path of water from this area may be useful in understanding the nature of PCB releases from the landfill.

Fluorescein dye injected at LF-6 was unambiguously detected at several monitoring wells at the Lemon Lane Landfill and at ICS. Table B-5 summarizes positive fluorescein results in chronological order. Fluorescein was first detected in a single sample at MW-4s and later detected at a cluster of wells in the Valhalla Cemetery (00-370, NN-300, NN-300A, 00-300A, and 00-300). Fluorescein was first detected at ICS (see Figure A-11) at 9:00 a.m. on November 1, 2001. Travel time to the spring based on first arrival was 21 hours. Figure A-12 provides an interpretation of the fluorescein dye flow path.

MW-4S is the only well on the east side of the landfill that received an apparently large pulse of fluorescein. Tracer dye from LF-6 may have appeared first at MW-4S, within 6 hours of injection, but this detection is problematic. Residual concentrations of fluorescein dye from a 1989 injection were noted at MW-10 in this area during the background monitoring (see Table B4). Second, only a single sample from MW-4s was found to contain fluorescein; all other locations that were positive for fluorescein yielded more than one sample that was positive for the dye. Viacom is continuing to investigate the possible connection between LF-6 and MW-4S.

The detections of fluorescein dye in the cluster of wells in the Valhalla Cemetery are unequivocal. The tracer dye was first detected at 00-370 within 7 hours after injection, and within 10 hours, concentrations were high enough that the tracer dye was visually apparent. The tracer dye appeared in well NN-300 at this same time. Sampling of the "A" series wells at the Valhalla Cemetery began the morning of November 1, and the first samples from 00-300A and NN-300A also contained fluorescein. Tracer dye was also detected in well 00-300 on the morning of November 2, 2001.

The high concentrations of dye detected in the cluster of Valhalla Cemetery wells and the fact that these samples were obtained without any well purging suggests that the wells are close to a conduit that drains water from the epikarst zone at LF-6. Results further suggest that the movement of water in the vadose zone near LF-6 is southwest and downward, consistent with the researched flow models (Palmer 1986 and 1999).

The wells that received dye are screened in two different stratigraphic (or conduit) zones. Wells NN-300A and 00-300A are shallow wells that do not extend to the 795- to 800-foot amsl horizon. Wells 00-370, NN-300, and 00-300 do extend to 795 to 800 feet amsl. The data suggest that the tracer dye had descended through the vadose zone to an elevation only slightly above the 795- to 800-foot amsl conduit zone at a point near these wells, or alternatively, followed two different flow paths to this area. Viacom continues to investigate interconnections between LF-6 and wells in this area.

A distinct breakthrough curve for Fluorescein was observed at ICS (see Figure A-12). Samples were collected at the Quarry B spring on an hourly basis from the time of dye injection until 8:00 p.m. on November 1, 2001. Subsequent samples were collected at 2-hour intervals beginning at 8:00 a.m. on November 2, 2001, or 44 hours after dye injection. Fluorescein was detected in only 1 of 42 samples collected at the Quarry B spring. This detection (0.26 1lg/L) occurred at 1:00 a.m. on November 1, well before the arrival of the dye at ICS. The detected level was well below the maximum background level and is therefore not regarded as a positive trace result. Fluorescein was not detected in any of the 13 samples collected from the swallow hole seep location.

This information suggests that no significant underflow condition exists at Quarry B or the swallow hole seep. Quarry B is supplied primarily by water from ICS sinking at swallow hole seeps and some additional drainage basin, perhaps several acres in size, that supports base flow to the spring when there is no flow to the swallow hole. Treating Groundwater emerging at ICS would minimize the amount of water to be treated because ICS is the farthest known upstream location where the PCBs within this flow system can be captured and treated.

Fluorescein was not detected at the Slaughterhouse spring. This spring was sampled for almost 2 days after injection.

Viacom (2002c) estimates Fluorescein tracer dye recovery at ICS to be 185.6 percent. Usually, much less dye is recovered than was actually injected due to adsorption, dead spaces, and dispersion losses along the flow path. This situation indicates a significant error in either spring flow rate measurements or dye concentration analysis at the spring. Dye concentration analysis errors are most likely (Viacom 2002c). Dye concentration data are likely biased high, and should be considered semiquantitative.

where

The average spring flow rate during the period from the fluorescein injection to the first arrival of the dye at ICS was about 185 gpm (see Figure A-11). Based on the above relationship, the predicted PCB travel time to ICS is about 14 hours. However, the first arrival of fluorescein dye at ICS occurred 21 hours after injection. It is not known if the PCB travel time correlation is applicable to nonstorm conditions. If it is directly applicable, the observed travel time is significantly longer than expected and may imply that mobilization of PCBs directly from LF-6 is not a significant contributor to the PCB peak observed at ICS during storm events. Viacom continues to investigate PCB travel time, and storm event tracer dye injection is planned at LF-6.

Eosine from the NN-700 tracer dye injection on October 31, 2001, was detected at several wells in Valhalla Cemetery (NN-625, 00-587, 00-370, NN-300 and 00-387), and possibly at several wells on the east side of the landfill (MW-15, MW-16, and MW-18, through MW-21) (see Table B-6). All of these detections are at much lower concentrations than those observed for the fluorescein dye injection, despite the fact that equal quantities of each tracer dye were injected. Figure A-13 provides an interpretation of the tracer dye flow path.

Eosine was definitely detected at 00-370 about 340 feet east of the injection site (4.2 hours after injection) and at NN-300 about 390 feet east of the injection site (22 hours after injection). The dye probably persisted at both of these locations for a time but was masked by overwhelming concentrations of fluorescein dye that appeared soon after the initial eosine detections (see Table B-5). Because the eosine fluorescent peak (536 nm) is close to the fluorescein (512 nm) peak and because fluorescein is a much more fluorescent dye than eosine, it is difficult to detect eosine in the presence of a large concentration of fluorescein.

Later detections of eosine, represented by strong detections in multiple samples, are evident at 00-387 and 00-587, but the dye did not appear at these locations until about 12 hours after injection. The last two samples collected at NN-625 approximately 121 and 166 hours after injection showed definite eosine detections.

Eosine detections at the several wells on the east side of the landfill were generally at very low levels, and this raises some question about the significance of eosine data. Eosine was not detected in any preinjection background samples; therefore the results may indeed be valid detections. Viacom (2002c) reports detections of eosine at MW-20 and MW-21, but both of these wells had considerable background fluorescence at about 525 nm attributable to degraded Rhodamine WT, which interferes with the eosine fluorescence peak. This raises questions about the validity of these cosine detections.

Eosine did not appear in any of the Valhalla Cemetery "A" series wells, demonstrating that the tracer dye was migrating solely through the 795- to 800-foot amsl conduit zone and not at the higher 815 feet amsl elevation zone that these wells are screened in.

The eosine data generally indicate that the tracer dye flowed east from NN-700 (see Figure A-13) and detections as far east as NN-300 appear definitive. Viacom (2002c) interprets low-level eosine detections in wells on the east side of the landfill as indicating a northward flow to the vicinity of MW-20 and MW-21. One problematic aspect of this interpretation is inconsistency between the fluorescein and cosine sampling results. Both dyes were detected early during the test at wells 00-370 and NN-300, suggesting close proximity between the fluorescein and cosine dye flow paths. However, fluorescein, the more analytically sensitive of the two dyes, was not detected at any of the east side wells where eosine was reported. A convergent flow path to the MW-21 area is not supported by the fluorescein data because only a single questionable detection of fluorescein occurred in any of the east side wells (see Table B-5).

Eosine was not detected at any of the monitored springs. There are two likely explanations for this: (1) eosine arrival at ICS was later than that of fluorescein and masked by the much larger fluorescein concentration (the more analytically sensitive dye), or (2) the eosine was diluted beyond SSFP detection limits (Viacom 2002c).

Phloxine B injected at SP-1 was not detected unambiguously in any well or spring. Possible reasons for this are summarized below (Viacom 2002c):

• An inadequate quantity of phloxine B was injected to overcome sorptive properties.
• Similar to cosine, phloxine B dye may have arrived at ICS at the same time as the fluorescein from LF-6, and the phloxine B may have been masked by the high concentrations of fluorescein dye.
• The dye could have gone to an unmonitored location or arrived at a monitored location after monitoring ceased.

PCB sampling was conducted at ICS and the Quarry B spring during the October 31, 2001, tracer dye test. The purpose of the PCB sampling was to determine if PCB levels increased at ICS because of the injection of water at LF-6 or decreased at the Quarry B spring because of cessation of discharge into the ICS swallow hole. PCB results are presented in Viacom (2002c). Most of the ICS PCB analytical data were rejected as biased low due because of quality assurance and quality control concerns. A total of 12 samples for PCB analysis were collected at the Quarry B spring during the test period. PCB concentrations ranged from 1.3 to 3.6 ppb and showed no definite temporal trend.

2.2.5 Long-Term MW-21 Pumping Test - November 2001

The long-term pumping test was conducted under low-flow conditions. The goals of the long-term test were to (1) determine the pumping rate that would affect the ICS, (2) evaluate the impact of long- term pumping on PCB concentrations at ICS, and (3) evaluate the impact of long-term pumping on PCB concentrations at each pumped well. The test was conducted in conjunction with natural gradient tracer dye tests (see Section 2.2.4).

The pumping test at MW-21 began at 10:00 a.m. on November 7, 2001, and ended on November 9, 2001. Attachment C presents a copy of Viacom's work plan for the long-term pumping tests at MW-21 and MW-16. The general purpose of this test was to determine if pumping a well along the east side of the landfill, believed to be close to a conduit that drains the Lemon Lane Landfill, could affect the PCB content of ICS discharge. Data collection, MW-21 pumping test results, and conclusions drawn based on pumping test results are discussed below.

2.2.5.1 Data Collection

The pumping rate was set at 17 gpm and maintained approximately at this rate for the duration of the test. The pump was shut off on November 9, 2001, at 2:00 p.m. after 52 hours of pumping. The pumped water was contained, treated on site in a mobile treatment trailer, and eventually discharged to Sargent's Pond.

Selected wells in the vicinity of the landfill were monitored for groundwater elevations before, during, and after pumping. IDEM continuously monitored flow and conductivity at the ICS treatment plant. MW-21 was sampled for PCBs and tracer dyes. Residual levels of dyes from the October 31, 2001, injections were still possibly present in the conduit system, and the arrival times of the dyes at the pumping well could provide insight about the closeness of the well to the dye travel paths. Additionally, well LF-6 was flushed with 300 gallons of water on November 8, 2001, at 3:30 p.m. to remove residual dye from the well bore and vicinity. Samples were collected periodically from ICS during the duration of the test and analyzed for tracer dyes and PCBs

2.2.5.2 MW-21 Pumping Test Results

Table B-7 briefly summarizes MW-21 pumping test information. Figure A-14 shows time-series plots of several data sets collected during the MW-21 pumping test. These data sets include ICS flow and PCB and fluorescein tracer dye concentrations at MW-21 and ICS.

Upon initiation of pumping at MW-21, ICS flow began to drop (see Figure A-14). The apparent cyclic variations in spring flow observed are related to the cycling of pumps in the SRS building at the ICS treatment plant and do not indicate actual variations in flow. However, by comparing flow values from the same point in the pumping cycles, an accurate assessment of the variation in spring flow may be obtained. The effect of the MW-21 pumping is most apparent at the time pumping ceased on November 9, 2001, at 2:00 p.m. Over a period of a few hours, the spring discharge increased about 15 gpm , an amount nearly equal to the MW-21 pump discharge. No significant rainfall occurred during the MW-21 pumping test; therefore, the pumping at MW-21 clearly affected ICS flow. The rapid response of both the spring flow and water levels in several monitoring wells indicates that the pumping well is very well connected to the conduit zone(s) that supply ICS.

Water levels were monitored continuously in MOO-6, MW-8S, MW-15, MW-17, MS-1, NN-625, and SP-1. Drawdowns at the end of the test ranged from 0.70 foot (MW-8S) to 0.98 foot (MW-17). Maximum drawdowns observed were slightly greater and occurred about 7 hours before the completion of the test.

Drawdowns observed in the monitoring wells during the MW-21 pumping test may be used to evaluate certain aquifer conditions. Figure A-15 is a semilogarithmic plot of drawdown versus time for the MW-21 pumping test. In this figure, the slopes of the plotted lines are roughly parallel, suggesting relatively homogeneous bulk aquifer transmissivity.

The PCB concentrations at MW-21 remained consistent at 5 to 7 ppb during the test (see Figure A-l 4). These levels are well below those observed at adjacent monitoring well MW-4I during two short-term pumping tests in Summer 2001 (see Table B-3). The data first suggest that MW-21 is not on or hydraulically well-connected to a major conduit that carries a large quantity of PCBs to ICS. The 5 to 7 ppb PCB concentrations observed during the MW-21 pumping test compares with typical ICS PCB levels of 10 to 20 ppb under the flow conditions observed during the test (Viacom 2002c). These data are consistent with the PCB response at ICS during MW-21 pumping (see Figure A-14). The PCB concentration at ICS was about 12 ppb prior to pumping, but within 4 to 6 hours after pumping began, the levels increased to above 20 ppb, remained above 15 ppb, and did not return to the 12-ppb prepumping level until about 8 hours after the pumping was terminated (see Figure A-14). Relatively diluted water was removed from the MW-21 pumping well, resulting in a more concentrated PCB discharge at ICS.

According to Viacom (2002c), the PCB results at ICS must be used with caution because the sample bottles remained in the autosampler for some time at an elevated temperature and were not cooled during that time. Therefore, the conclusion that PCBs levels decreased at the ICS after the pump was turned off is uncertain.

Viacom continues to investigate why MW4I produced high PCB concentrations in Summer 2001 while adjacent well MW-21 produced relatively low PCB concentrations during the November pumping test. Viacom (2002c) notes that MW4I was modified by removing some bentonite at the bottom of the well between the summer pumping tests and the November 7, 2001, test. This change could have modified the zone that supplies most of the water to the well upon pumping and thereby changed the PCB concentration in water produced from the well.

Samples for tracer dye analysis were collected at both MW-21 and ICS during the pumping test. Dye was sampled for at the pumping well to determine if any residual dyes previously injected in the aquifer on October 31, 2001, would be drawn to the pumping well. In addition to the test discussed above on pages 16 through 19, an additional fluorescein tracer dye injection occurred at 3:30 p.m. on November 8, 2001, as 300 gallons of water was flushed into LF-6.

Data contained in Table 8 of Viacom (2002c) suggest that before the pump was turned on, MW-21 contained significant concentrations of the Rhodamine WT (573 nm) fluorescent peak as well as a second peak at about 525 nm, which was interpreted to be a degraded Rhodamine WT by-product. These background concentrations remained significant throughout the pumping period and are most probably attributable to the previous 1990 injection of Rhodamine WT at MW-10.

Viacom (2002c) reports the presence of phloxine B dye in samples collected after more than 30 minutes after pumping was initiated. However, this interpretation is extremely ambiguous because of the large fluorescent background of these samples.

Viacom (2002c) also reports the presence of fluorescein dye in samples from MW-21 beginning 30 minutes after pumping was initiated. A definitive fluorescein peak is indicated on several SSFP tracers, and these results are much more definitive than the phloxine B results; however, the source of the fluorescein in the initial sample results is unclear. Because fluorescein was detected in the background sample from MW-10, it is possible that the initial detections ofthe dye at MW-21 are related to this source. A large fluorescein peak occurred at 4:00 p.m. on November 9, 2001, that was most likely related to the flushing of LF-6 on November 8, 2001, at 3:30 p.m. (see Figure A-14). This interpretation would suggest that under pumping conditions, MW-21 may capture some vadose zone infiltration from LF-6.

Samples were also collected from ICS during the MW-21 pumping test (see Figure A-14) for dye analysis. Fluorescein was detected in all of these samples as a result of the tracer in injection at LF-6 on October 31, 2001, and dye concentrations were found to decline slightly during the pumping period. Neither cosine or phloxine B were detected. A distinct fluorescein breakthrough curve occurred beginning at 8:00 a.m. on November 10, 2001 (see Figure A-14). This breakthrough curve began 40.5 hours after the LF-6 fluorescein flush on November 8 and is probably directly attributable to it. Average ICS flow during this time was 63.5 gpm . The equation presented in Section 2.2.4.3 suggests a travel time of 34 hours at this flow rate. This is slightly shorter than the actual travel time observed during the test.

2.2.5.3 MW-21 Pumping Test Conclusions

However, an apparent eosine peak was detected in a background sample from this well at a far greater concentration than any of the post-injection samples (see Table By). The positive eosine interpretation is also questionable because of interfering fluorescence at about 525 nm. This 525-nm fluorescence most likely is due to the degradation of Rhodamine WT tracer dye previously injected near this location.

The eosine detection is also problematic because there is an inconsistency between the fluorescein and cosine well sampling results. As stated above, both dyes were detected early during the test at wells 00-370 and NN-300, suggesting close proximity of the wells to both the fluorescein and eosine flow paths. However, fluorescein, the more analytically sensitive of the two dyes, was not detected at any of the east side wells where eosine was reported, including MW-21.

MW-2l did not contain fluorescein or phloxine B at detectable levels during the natural gradient portion of the test but did receive these two dyes within 30 minutes of the start of the pumping of the well on November 7. This indicates that at least some of these dyes flowed near the well. However, it should be noted that the source of the fluorescein tracer may have been the previous dye injection at MW-10. A substantial background concentration of fluorescein occurred in this well (see Table B-4), which is only a few feet away from MW-21. In addition, it is uncertain that the dye detected was phloxine B because of the persistent fluorescent background at about 525 to 574 nm in the samples collected.

MW-21 received a distinct pulse of fluorescein after LF-6 was flushed on November 8, 2001. This indicates that at least some of the fluorescein flushed from LF-6 was captured by the pumping well. However, ICS received a much larger pulse of dye from the November 8 flush, indicating that MW-21 was not effective in capturing the largest portion of the tracer dye. These data suggest that there may be two flow paths for water flushed from the LF-6 area and that MW-21 is near only a minor flow path. This interpretation appears to be correct.

Viacom (2002c) concludes that MW-21 is not an efficient pumping location to capture materials flushed from the epikarst at LF-6. Pumping at MW-21 did not lower the PCB concentrations at ICS after 52 hours. In fact, PCB levels appear to have increased shortly after pumping began. Viacom (2002c) suggests that the conduit network carrying the PCBs from the LF-6 area may be at an elevation high enough to prevent capture by wells pumping at 795 to 800 feet amsl.

2.2.6 Long-Term MW-16 Pumping Test - November 2001

The long-term pumping test was conducted under low-flow conditions. The goals of the long-term test were to (1) determine the pumping rate that would affect the ICS, (2) evaluate the impact of long- term pumping on PCB concentrations at ICS, and (3) evaluate the impact of long-term pumping on PCB concentrations at each pumped well. The test was conducted in conjunction with natural gradient tracer dye tests (see Section 2.2.4).

The pumping test at MW-16 began at 12:00 p.m. on November 13, 2001 and ended on November 19, 2001. Attachment C presents a copy of Viacom's work plan for the long-term pumping tests at MW-16 and MW-21. The general purpose of this test was to determine if this MW-16 would provide a better capture point for PCBs in groundwater from the LF-6 area. The test was conducted for almost 6 days to allow time to develop as large a capture zone as possible. The MW-16 pumping rate was set at about 17 gpm. The pump was shut off on November 19, 2001, at 8:30 a.m. after a total of 140 hours of pumping. A total of approximately 143,000 gallons of water was pumped from the well. The water was temporarily stored in the on-site retention basin and then treated in the mobile treatment trailer to remove PCBs prior to discharge at Sargent's Pond. Data collection, MW-16 pumping test results, and conclusions drawn based on pumping test results are discussed below.

2.2.6.1 Data Collection

Selected wells in the vicinity of the landfill were monitored for water level before, during, and after pumping. IDEM continuously monitored flow and conductivity at the ICS treatment plant. MOO-I 6 was sampled for PCBs and tracer dyes. LF-6 was again flushed with 300 gallons of water on November 16, 2001. PCBs and tracer dyes were also monitored at ICS. Samples were collected at the spring at various intervals (hourly or every 4 hours) during the duration of the pumping test and analyzed for dye and PCBs.

2.2.6.2 MW-16 Pumping Test Results

Table B-7 briefly summarizes MW-16 pumping test information. Figure A-16 shows time-series plots of several data sets collected during the MW-16 pumping test. These data include ICS flow (mean hourly discharge) and PCB and fluorescein tracer concentrations at MW-16 and ICS.

Similar to the MW-21 test, ICS flow was noted to drop almost immediately upon the initiation of pumping (see Figure A-16). The spring flow reduction observed during the pumping test was approximately equal to the well pumping rate. Flow increased by an amount about equal to the pumping rate upon the termination of the test on November 19, 2001.

Water levels were monitored continuously in MOO-6, MW-8S, MW-15, MW-17, MS-1, NN-625, and SP-1. Drawdown in the monitored wells ranged from 1.15 feet (MW-8S) to 2.165 feet (SP- 1) at the end of the test. Maximum drawdowns observed were slightly greater and occurred about 45 hours before the completion of the test.

Drawdowns observed in the monitoring wells during the MW- 16 pumping test can be used to evaluate aquifer conditions. Figure A- 17 is semilogarithmic plot of drawdown versus time for the MW- 16 pumping test. Similar to the MW-21 test, the slopes of the plotted lines are roughly parallel, suggesting relatively homogeneous bulk aquifer transmissivity.

Figure A-18 shows hydrographs of selected monitoring well water levels prior to, during, and after the test. All monitored wells open to the 795 to 800 feet amsl elevation zone showed drawdown and - recovery associated with pumping at MW-16. During the test, SP-I, NN-700, and LF-6 received additional slugs of 100 to 300 gallons of water to flush any additional tracer dye from the well bore. All monitored wells showed an instantaneous rise in water level associated with the flushing at SP-1 and NN-700. No similar response was noted during the flushing of the epikarst well LF-6 on November 19, 2001.

The rapid response of both the ICS flow and the water levels in monitoring wells screened in the 795- to 800-foot amsl elevation zone to MW-16 pumping indicates that wells open to this zone are hydraulically well-connected and that this zone is hydraulically well-connected to ICS.

Figure A-19 shows drawdowns measured during the MW-16 pumping test 70 to 73 hours after the initiation of pumping. Relatively small drawdown values were observed at wells in the MW-4 area (MW-20, MW-21, MW-5S, and MW-4I). Viacom (2002c) suggests that these data indicate proximity to a large conduit that may act as a local recharge source to these wells.

The PCB concentration at MW-16 was about 8 ppb upon the initiation of pumping. The concentration declined to a minimum of about 0.6 ppb and then increased and stabilized at 2.5 to 3 ppb (see Figure A-16). Viacom (2002c) notes that these results are generally consistent with the short-term pumping test conducted in Spring 2001 at this well.

The PCB concentration at ICS was about 15 ppb prior to pumping but increased to levels of about 20 ppb within 5 hours (see Figure A-16). Throughout the duration of the test, PCB concentrations varied considerably, but no reduction in the PCB concentrations occurred as a result of the pumping. Viacom (2002c) notes an apparent cyclic trend in the data during the test, possibly attributable to volatilization or loss of PCBs from selected samples as a result of varying periods of time that samples remained in autosampler devices after collection.

Samples for analysis of tracer dyes were collected at both ICS and MW-16 during the pumping test. Tracer dyes were sampled for at the pumping well to determine if any residual dyes previously injected in the aquifer on October 31, 2001, would be drawn to the well during pumping. Three additional dye flushes were performed during the pumping test in an effort to determine if MW-16 would capture the flushed dye from NN-700, SP-1, or LF-6.

Upon initiation of pumping, no significant levels of any of the injected dyes were detected at MW-16. One very low-level detection of apparently cosine occurred 2 hours after pumping was initiated. Low-level detections of apparently Rhodamine WT and fluorescein also occurred during the first 8 hours. However, results were not consistent from sample to sample, and no breakthrough curve was observed.

Beginning 8 hours after pumping, fluorescein was consistently detected in samples from MW-16. One prominent fluorescein peak occurred 28 hours after the initiation of pumping (see Figure A-16). Viacom (2002c) interpreted the source of this dye to be a continual slow release from the epikarst at LF-6.

Sporadic detections of cosine were also apparent at MW-16 after 8 hours of pumping. NN-700 was flushed with 100 gallons of water 47.5 hours after the initiation of pumping. After this time. cosine detections in the pumping wells were more consistent, but no breakthrough curve was observed (Viacom 2002c).

No phloxine B was detected at MW-16 even after an additional dye flush at SP-1 with 300 gallons of water on November 14, 2001.

A second flush of fluorescein dye was initiated at LF-6 on November 16, 2001, at 3:30 p.m. A fluorescein spike at MW-16, probably attributable to this injection, occurred at MW-16 8 hours after this flush (see Figure A-16). This indicates that at least some of epikarst drainage from LF-6 may be captured at MW-16, at least during nonstorm periods.

Samples for tracer dye analysis were also collected at ICS throughout the MW-16 pumping test. Fluorescein was the only tracer dye that was consistently detected (see Figure A-16). During the initial portion of the test, fluorescein concentrations at ICS were decreasing from the original October 31, 2001, dye injection; the subsequent November 8, 2001, dye injection; and again from the November 8 flushing event. Dye concentration was observed to increase slightly upon the initiation of pumping at MW-16 and the subsequent ICS flow reduction. Concentrations gradually declined until the termination of the pumping test on November 19, 2001.

A pronounced fluorescein breakthrough curve was observed at ICS beginning at 12:00 p.m. on November 19, 2001, or about 4.5 hours after the pumping test was terminated. This breakthrough curve is most probably attributable to the November 19 dye flush at LF-6. The data suggest a travel time from LF-6 of about 68 hours. Average spring flow during the 68-hour period was about 34 gpm The equation presented in Section 2.2.4.3 suggests a travel time of about 60 hours at this flow rate. The observed dye flow travel time was slightly longer than the time predicted by the equation but strongly suggests that the breakthrough curve at ICS beginning on November 19 was the direct result of the LF-6 dye flush on November 16.

Based on the MW-16 well pumping rate of 17 gpm and the average ICS flow rate during the period of dye discharge at ICS (about 50 gpm), much more of the dye flushed from LF-6 on November 16 was discharged at ICS than at MW-16. Most of the fluorescein dye flushed at LF-6 was not captured at MW-16.

2.2.6.3 MW-16 Pumping Test Conclusions

Based on the pumping test data, Viacom (2002c) draws the conclusions below.

• MW-16 is not on, near, or hydraulically well connected to a major conduit that carries a large amount of PCBs to ICS. The PCB content of the pumped well discharge was only 2 to 3 ppb and was even lower than the 5- to 7-ppb levels obtained during the previous MOO-2 1 pumping test. Overall, the test results indicate that location would be less efficient than MW-21 for capture of landfill area Groundwater and PCBs. The well is not suitably located for PCB recovery.
• None of the data suggest that phreatic drainage from the SP-l area to the ICS is influenced by MW-16. Phloxine B was not recovered in the pumping well. The phreatic drainage from the NN-700 area may be influenced by this well but not to a significant extent.
• A minor portion of the epikarst drainage from the LF-6 area may be influenced by MW-16, but most drainage flows directly to ICS. PCBs mobilized from the southeast corner of the landfill in the epikarst may be descending to a portion of the phreatic zone that is not impacted by pumping wells at the landfill.

2.2.7 Sargent's Pond Injection Test - November 2001

2.2.7.1 Data Collection

2.2.7.2 Injection Test Results

2.2.8 Short-Term MW-4S and MW-4I Pumping Tests - February 2002

Viacom also conducted a short-term pumping test at MW-4I in February 2002. The goals of the test were to determine if the well's hydraulic characteristics had changed since the August 2001 pumping test and whether high concentrations of PCBs exist near the well. The results of the pumping test indicate that well hydraulics had changed, probably because the well was still being partially plugged with bentonite during the August 2001 test. The February 2002 test also indicated high PCB- concentration water is not present in the vicinity of MW-4s (Viacom 2002c).

3.1 GROUNDWATER ANALYTICAL DATA

3.1.1 Residential Well Sample Results

3.1.2 Monitoring Well Sample Results

On April 4, 2001, Viacom conducted a pumping test at MW-19 (see Figure A-3). Viacom collected 10 groundwater samples from the well at 15-minute intervals during the pumping test. PCB concentrations detected in the samples ranged from 2.4 to 5.3 ppb (see Table B-8). Tetra Tech also collected two split groundwater samples with Viacom during the pumping test. PCB concentrations in Tetra Tech's split samples were 2.1 and 5.8 ppb; PCB concentrations in Viacom's split samples were 2.7 and 5.1 ppb, respectively.

On April 9, 2001, Viacom conducted a pumping test at MW-16 (see Figure A-17). Viacom collected 15 groundwater samples from the well at 15- and 30-minute intervals during the pumping test. PCB concentrations detected in the samples ranged from 3.5 to 5.2 ppb (see Table B-9). Tetra Tech also collected one split groundwater sample with Viacom during the pumping test. PCB concentrations detected in Tetra Tech's and Viacom's samples were 2.7 and 4.3 ppb, respectively.

On April 17, 2001, Viacom conducted a pumping test at MW-18. Viacom collected 10 groundwater samples from the well at l5- and 30-minute intervals during the pumping test. PCB concentrations detected in the samples ranged from 3.9 to 13 ppb (see Table B-10). Tetra Tech also collected one split groundwater sample with Viacom during the pumping test. PCB concentrations detected in Tetra Tech's and Viacom's samples were 8.7 and 9.1 ppb, respectively.

On July 12,2001, Viacom conducted a pumping test at MW-4I. Viacom collected 10 Groundwater samples from the well at 15-minute intervals during the pumping test. PCB concentrations detected in the samples ranged from 41 to 2,700 ppb (see Table B-ll). In addition, Viacom monitored ICS during the pumping test to determine if there was any impact on PCB concentrations. Viacom collected 12 Groundwater samples using an ISCO autosampler stationed at the spring during the time interval of 24 to 108 hours after the pumping test had ceased at MW-4I. PCB concentrations detected in the ICS samples ranged from 1.2 to 4.7 ppb. The lowest PCB concentrations were detected in the earlier samples and gradually increased during the time interval; however, no correlation between the high PCB values at the pumping well and the low PCB values at ICS can be made. It appears that the pumping did not have an impact on surface water PCB values at ICS. Table B-12 summarizes the ICS analytical data.

On August 1, 2001, Viacom conducted an additional pumping test at MW-4I. Viacom collected 20 Groundwater samples from the well at various time intervals during the pumping test. PCB concentrations detected in the samples ranged from 66 to 4,000 ppb (see Table B-13). Once again, Viacom monitored ICS during the pumping test to determine if there was any impact on PCB concentrations. Viacom collected 27 samples from ICS using an ISCO autosampler during 4- and 8-hour intervals beginning on July 31, 2001 (before the test), and ending on August 6, 2001, after the pumping test had stopped at MW-4I. PCB concentrations detected in the ICS samples ranged from 2.5 to 8.8 ppb. As was the case during the July 12, 2001, pumping test at MW-4I, no correlation between the high PCB values at the pumping well and the low PCB values at ICS can be made. It appears that the pumping did not have an impact on surface water PCB values at ICS. Table B-14 summarizes the ICS analytical data.

c

On September 27,2001, Tetra Tech and Viacom split two Groundwater samples collected during drilling activities at monitoring wells MW-20 and MW-21. An additional split Groundwater sample was also collected from well MW4I to determine if the drilling activities impacted PCB concentrations in this well, which is located between MW-20 and MW-21 on the east side of the site. PCB concentrations of 2.7 and 25 ppb were detected in the Tetra Tech and Viacom split samples collected from MW-20, and PCB concentrations of 3.0 and 4.0 ppb were detected in the Tetra Tech and Viacom split samples collected from MW-21, respectively. Additionally, the Tetra Tech and Viacom split sample collected from MW-4I contained PCB concentrations of 11 and 31 ppb, respectively. Based on comparison of these PCB sample analytical results to the August 1, 2001, PCB concentrations detected in pumping test samples collected from MW-4I, it appears that drilling activities did not impact the area around MW-4I.

3.1.3 Piezemeter Sample Results

On August 9, 2001, Tetra Tech and Viacom collected one split groundwater sample from piezometer PZ-D. PCB concentrations of 160 and 2,300 ppb were detected in the Tetra Tech and Viacom split samples, respectively. On August 16, 2001, Tetra Tech and Viacom collected split groundwater samples from piezometer PZ-D. PCB concentrations of 280 and 440 ppb were detected in the Tetra Tech and Viacom split samples collected from PZ-D, respectively.

3.2 NONSTORM SURFACE WATER ANALYTICAL DATA

• The ICS contained PCBs at concentrations ranging from 2.9 to 20 ppb.
• Composite samples collected from the Quarry A, B. and C springs contained PCBs at concentrations ranging from 0.37 to 1.8 ppb.
The Slaughterhouse spring did not contain detectable PCB concentrations during the monitoring period. PCB sample analytical results ranged from below the detection limit (BDL) (less than 0.1 ppb) to 0.11 ppb; however, the 0.11 ppb result was flagged "U" by the laboratory to indicate that the result was the reporting limit.
• Sargent's Pond contained PCB at a concentration of 0.76 ppb.

Tetra Tech has continuously updated figures of Viacom's interim monitoring data for the ICS and Quarry A, B. and C springs as the data has been released by Viacom. Figures A-21 and A-22 show the sample analytical results for ICS and Quarry A, B, and C springs for the interim monitoring period of 2000 and 2001. As the figures show, PCBs continue to be discharged from Lemon Lane Landfill through groundwater conduits to the ICS. Figures A-21 and A-22 do not show a decreasing trend of PCB concentrations since completion of the source control RA at the landfill in Fall 2000.

3.3 STORM SURFACE WATER ANALYTICAL DATA

On October 24, 2001, Tetra Tech collected four surface water samples during a storm event. Two influent samples and one overflow sample were collected at the ICS treatment plant. One surface water sample was also collected from the ICS channel springs near the plant. All samples contained PCBs at concentrations ranging from 0.14 to 210 ppb. Table B-16 summarizes sample analytical results.

On November 27 and 30, 2001, Earth Tech collected 12 surface water samples during two storm samples contained PCBs at concentrations ranging from 4.6 to 15 ppb.Table B-17 summarizes sample analytical results.

On December 14, 17, and 18, 2001, Earth Tech collected 18 surface water samples during two storm events. Thirteen influent and five overflow samples were collected at the ICS treatment plant. All samples contained PCBs at concentrations ranging from 11 to 63 ppb. Table B-18 summarizes sample analytical results.

On January 31, 2002, Earth Tech collected four surface water influent samples during two storm events at the ICS treatment plant. All samples contained PCBs at concentrations ranging from 10 to 110 ppb. In addition, TSS results ranged from 21 to 356 parts per million (ppm). Table B-19 summarizes sample analytical results.

On March 3, 2002, Earth Tech collected two surface water influent samples during a storm event at the ICS treatment plant. Both samples contained PCBs at 6.8 and 9.5 ppb. Table B-20 summarizes sample analytical results.

On March 25 and 26, 2002, Earth Tech collected four surface water influent samples during a storm event at the ICS treatment plant. All samples contained PCBs at concentrations ranging from 2.8 to 3.2 ppb. In addition, TSS results ranged from 9 to 38 ppm. Table B-21 summarizes sample analytical results.

4.0 ICS TREATMENT PLANT DATA

4.1 ICS FLOWS

4.2 MASS OF PCBs CAPTURED BY ICS TREATMENT PLANT

To calculate the storm-flow PCB mass, the PCB mass was the calculated for storm flows exceeding 500 gpm based on ICS flow rates and PCB influent sample results. The PCB concentrations for the mid-storm samples collected by Earth Tech during each storm were used to calculate the cumulative PCB mass for each hour of every storm event. The sum of the PCB mass for all storm events yields the total PCB mass entering the treatment plant during flows exceeding 500 gpm. To estimate the PCB mass captured, the PCB mass released during overflows (see Section 4.3) was subtracted from the total PCB mass entering the treatment plant.

Appendices C and D summarize the mass of PCBs captured by the plant from October 2001 through March 2002 during nonstorm and storm events, respectively. The total mass of PCBs captured by the plant during nonstorm events with flows of less than 500 gpm is estimated to be 688.45 grams (approximately 1.52 pounds [see Appendix C]), or 100 percent of the PCB mass entering the plant. The total mass of PCBs captured by the plant during storm events is estimated to be 2,759.91 grams (approximately 6.1 pounds [see Appendix D]), or 64.31 percent of PCB mass entering the plant during storm flows greater than 500 gpm. Section 4.5 below discusses the amount of sediment potentially contaminated if the ICS treament plant was not constructed and this captured mass of PCBs was distributed downstream of ICS.

4.3 MASS OF PCBs RELEASED BY ICS TREATMENT PLANT

4.4 ICS TREATMENT PLANT SLUDGE ANALYTICAL DATA

• Toxicity characteristic leaching procedure (TCLP) barium = l.O milligram per liter (mg/L)
• PCB Aroclor 1248 = 42 milligrams per kilogram (mg/kg)

All other analyses (except pH and Cashpoint) yielded nondetect results reported as less than the reporting limit for each analyte. Table B-23 summarizes sample analytical results.

• TCLP barium=1.02 mg/L
• PCB Aroclor 1242 = 21 mg/kg
• PCB Aroclor 1248 = 34 mg/kg
• PCB Aroclor 1260 = 2.0 mg/kg

All other analyses (except pH, Cashpoint, paint filter, percent moisture, and total solids) yielded nondetect results reported as less than the reporting limit for each analyte. Table B-24 summarizes sample analytical results.

4.5 POTENTIAL SEDIMENT CONTAMINATION DOWNSTREAM OF ICS

5.0 SUMMARY AND FINDINGS

The investigations completed by Viacom in 2001 have provided a better understanding of the conduit system connecting the Lemon Lane Landfill to ICS. The continuous monitoring data collected by Viacom during 2001 provided information on groundwater levels in several monitoring wells and piezometers as well as water quality in the springs near the site. The ICS treatment plant data provided valuable information for calculating the mass of PCBs captured by the plant and released during storm events that resulted in overflows. The investigations completed in 2001 suggest the findings below.

• The shallow epikarst zone in the landfill is connected to the epikarst zones in the Valhalla Cemetery south of the landfill (as evidenced by the major hydrologic pathway from LF-6 into Valhalla Cemetery, whose flow is predominately in the epikarst zone).
• The bedrock conduits above the 850-foot amsl elevation are somewhat connected within the landfill but do not appear to directly impact flow at ICS.
• Wells on the east side of the landfill are not effective at intercepting PCBs flushed from the epikarst as pumping tests in these wells did not significantly impact the flow rates or PCB concentrations at ICS.
• It is possible to affect flow rates at ICS by pumping the phreatic zone.
• Dye trace testing at the landfill confirmed that the Quarry B spring is not an underflow of ICS.
• The dye test results suggest that the capture of PCBs as they are flushed may be most effective either at Valhalla Cemetery or at the extreme southern portion of the landfill adjacent to the Illinois Central Railroad tracks.
• Pumping of wells in southern and eastern portions of the landfill did not capture much of the PCBs in the bedrock. Also, the wells yielded only about 20 gpm on a consistent basis under nonstorm conditions.
• Sargent's Pond has a low leakage rate at elevations up to 843 feet amsl.

The continuous monitoring data provided by Viacom indicate that the water quality (specifically, PCB concentrations) at ICS has not decreased significantly since completion of remediation in 2000. PCB concentrations at ICS typically range between 5 and 20 ppb during nonstorm conditions. Storm-water PCB concentrations are significantly higher (see Sections 4.2 and 4.3). The PCB concentration in the Quarry A, B. and C springs has remained about the same (about 2 ppb) after completion of remediation at Lemon Lane Landfill.

The ICS treatment plant has been effective in capturing PCB-contaminated water from the ICS, and from October 10, 2001, through March 2002, the mass of PCBs captured was approximately 7.6 pounds. During storm events, flows many times exceed 1,500 gpm and water accumulates in two storage tanks in addition to the SRS building. When the capacity of the storage tanks is exceeded, the water overflows into the ICS channel to Clear Creek. The mass of PCBs released during storms in the same time period was approximately 3.37 pounds.

Viacom is continuing investigation to better understand conduit horizons at the landfill. Changes are planned at the ICS treatment plant to increase storage and improve the monitoring of PCBs treated by the plant. Viacom is continuing monitoring at the landfill and will be initiating a long-term groundwater monitoring program. Details of this program will be presented in the next status report for the Lemon Lane Landfill.

To summarize, an understanding of karst zones at the Lemon Lane Landfill is improving and (in addition to the continuous monitoring data) will allow the management of contaminated water at the landfill. The ICS treatment plant operation is improving, and the plant is capturing all PCBs released during nonstorm events. With the planned increase in storage, more of the ICS flow during storm events should be treated by the plant in the future and not discharged to Clear Creek downstream of the treatment plant.

REFERENCES

Davies, Gareth J. 2001. Personal Communication Regarding Rhodamine WT Degradation. Between Gareth J. Davies, Cambrian Ground Water Company, Oak Ridge, Tennessee, and John Bassett, Earth Tech, Inc., and Mike McCann, Viacom, Inc., October.

Palmer, Arthur N. 1986. "Prediction of Contaminant Paths in Karst Aquifers." Environmental Problems in Karst Terranes and Their Solutions Proceedings. National Water Well Association. October 28 through 30. Bowling Green, Kentucky. Pages 32 through 53.

Palmer, Arthur N. 1999. "A Statistical Evaluation of the Structural Influence on Solution-Conduit Patterns." Karst Modeling. Karst Waters Institute, Special Publication 5. Pages 187 through 195.

Quinlan, James F., and others. 1983. "Ground-Water Hydrology and Geomorphology of the Mammoth Cave Region, Kentucky, and of the Mitchell Plain, Indiana." Field Trips in Midwestern Geology. Vol. 2: Field Trip Guide BooLfor the 1983 Annual Meeting of the Geological Society of America." Robert H. Shaver and Jack A. Sunderman, Editors. Indianapolis, Indiana. Pages 1 through 85.

Tetra Tech EM Inc. (Tetra Tech). 2001. "Revised Current Status Report for Groundwater, Surface Water, Soil, Sediment, and Fish Data. Lemon Lane Landfill Site, Monroe County, Indiana." May 31.

Viacom, Inc. (Viacom). 2001a. "Lemon Lane Karst Conduit Program Latest Results and Near Term Plans." April 16.

Viacom. 2001b. Letter Regarding Pump Test at Lemon Lane Landfill. From Michael R. McCarm, Bloomington Project Geologist. To Distribution List. March 30.

Viacom. 2002a. "Post-Excavation Groundwater Monitoring Results for October 19, 2001; Lemon Lane Landfill, Monroe County, Indiana." From Dorothy M. Alke, Project Director. To Distribution List. January 4.

Viacom. 2002b. "Short Term Pump Test at 4i/4s." February.

Viacom. 2002c. "Lemon Lane Landfill Karst Aquifer Test Reports from October to November 2001." March 13.