1.0 Flush Test of LF6-8" Piezometer

1.1 Test Description

On April 10, 2002 at 10:00 hours 1032 gallons of water was flushed into Piezometer LF-6-8" along with 50 grams of Fluorescein dye. Approximately 5 feet of head was induced in the piezometer and presumably the gravel pit in which it is installed. It took 10 minutes to flush the water in. Water levels were measured by hand in MW-15 every 15 minutes for the first hour and hourly for 8 hours after. Transducers recording water levels every 5 minutes were installed in wells:

The following wells were sampled for dye hourly for the first 8 hours and then daily for 5 days:

Illinois Central Spring was sampled at its emergence with auto samplers on an hourly schedule.

No rain had occurred for several days before the flushing, nor in the 24 hours immediately following the flushing.

A copy of the test plan is included as attachment 1.

1.2 Test Results

Table 1 shows the dye results at the wells monitored and Table 2 (part 1 and part 2) shows the dye results at the spring along with the PCB results of select samples at the spring. Figures 1A, 1B, 1C, 1D, &1E show the datalogger results.

The water flushed into LF-6-8" reached the phreatic zone in one hour and raised the phreatic zone about 0.05 feet (less than an inch). Note that well KK-112, which is much closer to the IC Spring than the landfill and historically maintains a water level significantly above most of the landfill phreatic wells, also responded. This indicates the higher water level in this well is not caused by an isolated perched zone.

Distinct dye breakthrough curves were seen in wells 00-370 and 00-300A. The results in wells MW-5S, PZ-F, and NN-300 are less distinct. No dye was detected in 00-300. The apparent appearance of dye in all wells within an hour of flushing may be the hydraulic flushing of residual dye or an analytical anomaly. The first unambiguous appearance occurred in well 00-370 in 3 hours or less and peaked at 6 hours after flushing. The first unambiguous appearance in OO- 300A occurred 7 hours after flushing.

Figure 2 is a plot of the dye and PCB concentration as detected at lil inois Central spring. The first appearance of the dye occurred 11 hours after injection. A clear, single peak breakthrough curve peaking at 387 ppb at 13 hours after flushing is shown. This travel time of 11 hours is almost exactly what would be predicted for PCB peak storm travel time based on the average spring flow observed (240 gpm) during this test. This is different than previous dye test conducted from LF6 that showed a lag time of up to 7 hours between predicted and actual travel times and showed double peak breakthrough curves.

The PCBs at IC Spring were not significantly elevated during the time dye arrived at the spring. Although free product had been detected at the LF64 location within two weeks prior to the flushing, apparently little to no PCBs were mobilized during the flushing. Also note that the apparent mass recovery of dye at IC Spring was about 70 grams. With 50 grams injected, this indicates some measurement error, and also indicates that there was no significant storage of the dye injected.

Table 3 shows the water level in LF6-8 after water injection. The LF6 pit (gravel filled) dimensions are approximately 7' by 12'. Immediately after injection the water level had risen approximately 5 feet. The water took almost 9 hours to completely drain from the pit. This is an average outflow rate of about 2 gpm. Maximum outflow rate appears to be approximately 4 gpm.

2.0 Storm Tracer Test from LF64" Piezometer

2.1 Test Description

On May 6, 2002 300 grams of Fluorescein dye was injected with 10 gallons of flushing water in piezometer LF6-4" at 13:10 hours. This took place at the beginning of a 0.43" storm event. Dye was monitored at the spring on 10 minute intervals until 20:00 hours and thereafter hourly using auto samplers. Samples were also taken in wells:

MW-4s, PZ-F, NN-300, 00-300, 00-300A, 00-370

Samples were taken every 15 minutes until 19:30 hours. The hourly samples for dye at the spring were continued until 5/9/02 9:00 hours through 2 successive storm events.

Based on the dye results, select samples from the spring and wells were submitted for PCB analysis. Dataloggers measured water levels in LF64n, MW- 21, and MOO-6 on five minute intervals.

A copy of the test plan is included as attachment 2.

2.2 Test Results

Table 6 shows summary data for the IC Spring. Note that the total dye recovered was approximately 337 grams. With 300 grams injected this indicates a lessor amount of measurement error than for the flush test. Approximately one third of the dye injected on 5/6/02 was recovered as a result of the 5/7/02 storm and the remaining two thirds was recovered from the 5/8/02 event.

3.0 Conclusions and Implications for Further Work

3.1 Conclusions

• Some of the dye from the LF6 flush test was again confirmed to travel southwest into Valhalla Cemetary. Dye was not conclusively detected at well MW4S. Wells NN300 and 00300 did not receive unambiguous dye detections. These wells are screened in the 795-800 foot zone, while their paired uncreened and shallower neighbors (00300A and NN300A) and 00370 (another open hole well) received dye. This indicates th e dye may be traveling in an upper epikarstic zone.
• Flushing LF6 during non-storm conditions did not induce a PCB peak response at IC Spring. Some of the flushed water from LF6, when a high hydraulic head is induced, reaches the phreatic zone within one hour and induced a response even as far away as KK112.
• The outflow rate from the LF6 conduit averaged about 2 gpm.
• Based on tracer travel time and amount of hydraulic response during storms, the LF6 area is not the immediate contributing source to the high PCB pulse that manifests at the spring during storm flows.

3.2 Implications for Further Work

• Dye flushed from LF6 arrives in Valhalla wells near NN300A and 00370. This area of Valhalla should be further investigated to see if more PCBs can be found at these locations. Although high levels of PCBs were not associated with the LF6 flushes, the dye results may indicate that this area of Valhalla is near a major pathway from the southeast corner of the site where PCBs may be stored in other reservoirs. Samples taken above the water table from NN300 and NN300A during drilling had high concentration of PCBs.
• It appears that the dye travel from LF6 to the southwest may be occurring in an epikarst zone. Review of the drilling and videologs for these area reveals conduit development between 818 and 823 feet amsl. One well, 00300A, maintains a water level in this range and it received dye on all LF6 dye tests. Well 00370 also received dye during all dye test from LF6 and it is an open hole so the dye arriving at that well could be traveling at any elevation (less than 833 the top of rock at 00370). Well NN300A also received significant dye in other dye test (it was not sampled in this test). NN300A is an open hole well bottomed at about 810 feet amsl. Well 00370 was pump tested during 1999 and PCBs dropped during the pump test. This may indicate that the PCBs entering the well (like the dye) is coming from an epikarst fracture and being diluted by the phreatic water present in the well. The "A" series Valhalla wells should be modified to reflect only epikarst water levels (as 00300A does now), and isolate them from the phreatic water.
• Perform additional flush and tracer tests during non-storm conditions. Candidates include the shallow Valhalla wells after they are modified/isolated, PZF, and the railroad tracks.
• Consider additional shallow wells at the southeast corner of the landfill that could be located on existing geophysics anomalies and that may intercept the lower epikarst zones around the 823 elevation.