Summary of Dye Traces at Lemon Lane July-November 2002

There seem to be three categories of traces with regard to travel time from the site to IC spring.
They would be:

• Those tracer injections which are never detected.
• Tracer travel limes significantly longer than predicted.
• Tracer travel limes comparable to that preclicted by the empirically derived equation based on high PCB storm pulses.

Within the first category would be:

• The October 2001 injection of Phloxine B into SP-1.
• The October 2001 injection of Eosine into NN-700.
• The July 2002 injection of Fluorescein into MW-6.
• The October 2002 injection of F~uore-scein into MW-6.
• The October 2002 injection of Rhodamine WT into OO-370.
The second category would contain: The May 1996 injection of RWT into North Sink. The October 2001 injection of FLR into EF-6- 8".
• The May 2002 injection of FLR into LF-6-4 .
• The August 2002 injection of RWT into NN- 300A.
• The October 2002 injection of FLR into MW-16.

  The third category would contain:

  • The May 1996 injection of FLR into MW-7 area macropores.
  • The April 2002 injection of FLR into LF-6-8".
  • The July 2002 injection of RWT into MW-4i.
  • The August 9002 injection of RWT into PZ-F.
  • The October 2002 injections of RWT into MW-18

  Conclusions

• Traces that never produce break through curves at the spring represent areas that are not drained by major phreatic tributaries.
• Tracer travel times that are significantly longer than predicted by the PCB storm pulse regression represent locations that are "farther" from the spring, for example - North Sink, in terms of length of phreatic conduit. Another aspect of "distance" may also be lateral travel in the vadose zone.
• Locations where the tracer travel times match the PCB storm pulse are candidate locations for investigation as possible source areas for the PCB storm pulse.
• Despite the puzzle of 00370, the pattern emerges of phreatic convergence to the east side of Lemon Lane.
• Tracer locations in Valhalla have either produced no break through curves or delayed passage to the spring.
• This implies that the high PCB pulse does not originate in the Valhalla area, but lies north of the tracks under the Lemon Lane site itself.

Eosine Travel Chart

Epikarst Well Level Data
Non-Storm

• Pump Tests indicated that LF1 and LF5 have some partial connection and that LF2 is not connected to either
• EF5 usually has the higher level during non- storm conditions. This indicates flow to the south
• Based on pump test data, LF5 appears relatecl to a bigger reservoir than LF1
• Assumed leakage rates from either during non- storm conclitions appear small (< 1 gallon per hour)

Phreatic Well Level
Storm Data

• Well level vs flow at IC Spring
  • Some wells have more divergence from linearity in the rising limb (MW6, MW19, MW4S). This may indicate local recharge.
  • Wells sometimes show large slope changes on the receding limb vs IC Flow.
• Well levels from data loggers (at higher base flows) also show the MW4 area as the low
• On the east side, MW4S and MW19 rise faster during early rising limb of storms as does MW6 in Valhalla

Epikarst Well Level
Storm Data

• LF 1, 2 and 5 respond to even small storms
• LF5 has shown the highest recorded water level around the site at almost 870 feet amsl during a storm. Suspect this would have been significantly eclipsed during the Oct 2000 event
• Relative levels show flow from LF5 towards LF1 during storms
• LF6 and 7 respond
• EF6/7 response does not appear related to the LF1, 2, 5 responses although it is possible the secondary responses could be related
• LF7 appeared to receive storm water from a quick run-off source possibly the cap
• The only source of cap run-off near LF7 is the south slope area, this implies flow to the north/northeast
• Relative levels between LF6 and 7 do not show a consistent flow direction

Well Temperature Responses

• A great deal of temperature measurements exist for wells around Lemon Lane
• Interpretation of the data is complicatecl by
  • a lack of literature on the subject
  • data quality and comparability issues
  • Location of sensor in the well
  • Bad instrument data
  • Large scale weather changes
• Our strategy was to start with known events and derive a theory for temperature changes consistent with known drivers

Known Facts about Well Temperatures

• Epikarst water under the landfill was much warmer than phreatic water.
• Storm water has a significantly different temperature than aquifer water during certain times of the year.
• Non-storm temperatures in wells do not follow surface air temperatures. They tend to have annual thermal cycles that are out of phase with surface air temperatures.
• The re-routing of site drainage to Sargent's Pond had a major impact on MW7
• Wells do not show large temperature stratification vertically

Well Temperature Responses During Pump Tests

• Long term pump tests
  • MW21: note MW17 response and general lack of responese elsewhere
  • MW16: larger response seen in MS1
• Short Term pump tests

Well Flush Test 12/11/02
Valhalla Wells

• 00370 did not show a temperature response when flushing MW6, yet 00387did. Did we miss it because of sampling interval in 370?
• 00387 showed a larger temperature response to the MW6 flushing than to 00370 flushing
• Velocities
  • MW6 to 00387 = 2.3 Ft/Min
  • 00370 to 00387= 5.6 Ft/Min
• This implies that there is a conduit between 00387 and 00370
• Tentative response at 41 with a velocity of 1.8 Ft/Min
• No noticeable response in MW18

Well Flush Test 12/12/02
East Side Wells

• No flush water arrived at MW17
• Water flushed into MW18 arrived at 4I within 25 minutes. Travel velocity of 3.4 Ft/Min
• Water flushed into MW4 area arrived at MW18 within 40 minutes
• MW19 did not receive a major hit of any of the flush water

Theories on Well Temperatures

• Well temperature is a result of a dynamic equilibrium and changes in mass flux of water through a well will induce a temperature change. This implies a continuous flow vertically in the well bore during a transient.
• A level change by itself will not induce a significant temperature change in a well
• A warm anomaly indicates an increase in epikarst flow
• Differences in well temperature response can be useful in determining differences in local flow responses and water sources
• A large storm temperature response is indicative of a large flow change during the storm which indicates the well is near an active storm induced pathway

Well Temperature Data Review
Non-Storm

• Significant warm anomaly at MW19/18
  • Warm anomaly at MW6
  • Cold spot at 4 series wells

Well Temperature Data Review
Storm Data

• Most wells that are in the main phreatic zone do respond during storms
  • Deep wells do not respond
  • Limited data on epikarst wells do not show major responses
  • Some main phreatic wells respond more than others
  • Direction of the response varies both with season and location
  • Some wells are obviously impacted by storm water
• MW4I and MW21 have shown some disproportionately large responses to larger storms and in some cases smaller events (MW4I)
  • Oct 2000 event MW4I
  • May 2002 event MW21
  • Late 1999 MW4I responses
• Wells 8s, SP1 and MW7 (prior to lining this area) respond at times like storm water run-off
• Wells MW6 and MW19 have larger responses at limes
  • MW6 did have relatively more water develop at the 815 horizon during drilling
• Direction of storm responses varies by season and is not always consistent from year to year.
  • See MW4I during 1999 and 2000

Epikarst Well Temperature Responses During Storms

• LF7 appeared to receive storm run-off water quickly
• LF1/5/2 do not show much of a response
• When LF1 did respond, it did not indicate storm water
• LF6 predisturbance appeared to show a muted storm water response

Well Temperature Review
Conclusions

• There are temperature anomalies around the landfill and in Valhalla at non-storm conditions
• Most wells do show a temperature response during storms. Most times these responses are opposite of what you would expect from storm water. The direction of the response seems to vary seasonally. No well we have to date appears to act like a well that sees a large passage of storm water during a storm consistently.
• Any flow changes in the well will induce a temperature change. These flow changes may be more vertical than horizontal
• Abnormally large temperature changes at a well during a storm relate to large flow changes through the well. It is unclear if this indicates a conduit or a local condition.
• MW4I and MW21 have shown some abnormal responses during large events. It is not clear how to interpret these responses.

Summary of Basin/Aquifer

• Where are the major recharge areas
  • Martin's Sink
  • East side neighborhood
  • West side of landfill
  • North sink area
  • Far east neighborhoods
• What zones are used for tow flow transport of water and PCBs
• 798-800 zone: no storage available
• 815-818 zone
• epikarst zones supply water during low flow
• What zones are used for high flow transport epiphreatic zones
• 820-823 zone
• When does Epikarst water move during a storm
  • Early flow responses indicated from well levels and PCB response at spring
• When does site water get to ICs during a storm
  • Water arrival is indicated by PCBs which is repeatable and determined by flows which indicative of phreatic transport

PCB Data Review

• Concentration vs Flow
• Concentration vs Concluctivity
• Mass Rate vs Flow
• Comprehensive review of concluit testing impact on low flow PCB levels at the ICS
• PCBs in Wells
• Recent Storm Samples

PCBs at ICS Non-Storm

• PCBs Concentration vs Flow
• PCBs Concentration vs Conductivity:
  • Note: Conductivity clata may be impacted by road salt in winter months
• PCB Mass Rate vs Flow:
  • Note mass rate, indicates that mass per minute is lower during dry periocis
  • Assuming a 1000 ppb source strength, flow of the source could be .25 to 2.5 gpm at non-storm conditions

PCBs at ICS During the Flush Tests

• LF6
  • No impact
• PZF
  • No impact
• NN300A
  • No impact

PCBs at ICS During the Long Term Pump Tests

• PCBs at ICS were impacted by MW21 pump test
  • Mass rates increased after adjusting for autosampler time
• PCBs at ICS were not significantly impacted by the MW] 6 pump test
  • Mass rates not significantly changed after adjusting for autosamp~er times

PCBs at ICS During the Short Term Pump Tests

• Data quality impacted by time in autosamplers
• Tests were too short

PCBs at MW4 Area Wells During Storm Sampling

• No well showed a significant increase in
• PCBs during the storm
• Most wells indicated dilution of PCBs when concluctivity ciroppecl
• MW18 has the highest pre-storm PCBs
• MW19 showed the largest concluctivity change during the storm

Proposed Epikarst Investigations

• Add new wells to the epikarst based on existing geophysics in the SE corner
• Perform more dye testing during storm/non-storm conditions to establish which epikarst wells mimic PCB travel times
• Sample new wells for PCBs
• For those locations that match PCB travel limes, perform pump tests to determine reservoir size and PCB removal potential
• Utilize liming match locations as injection locations for phreatic location testing
• Dye test storm run-off to identify source of flushing water and test for dyes in candidate epikarst wells

Proposed Phreatic Investigations

• Utilize found epikarst wells that match PCB timing as injection locations
• Inject brine to enhance geophysical signals
• Utilize time lapse electrical resistivity with brine injections to illuminate pathway that PCBs may take
• Drill pathway and perform sampling and pump tests

Feasibility Testing

• Long term pump test from new phreatic wells
• Short term/long term pumping from epikarst locations
• Storm run-off collection to eliminate flushing source
• If source is identified and cannot be efficiently removed, possible to try in-situ degradation
• Epikarst clean water collection/removal