FortSheridan.Net


September 9, 1996

Ms, Colleen Reilly
BRAC Environmental Coordinator
3155 Blackhawk Drive, Suite 17
Fort Sheridan, Illinois 60037

Dear Ms. Reilly:

Please accept these comments on the Proposed Plan Landfills 6 and 7 Interim Action (ESE, August, 1996) on behalf of the Sierra Club, Great Lakes critical Lands Project The general comments below are intended to supplement the attached review of the Proposed Plan, the Fort Sheridan Landfills 6 and 7 Interim Action Draft Final Focused Feasibility Study and other relevant materials, prepared by report from Charles H. Norris, a hydrogeologist with Geo-Hydro, Inc. on our behalf.

We do not believe that the data provided in the Feasibility Study and the Phase I Draft Final Remedial Investigation-Risk Assessment (RI/RA), as summarized in the Proposed Plan, sufficiently supports either the evaluation or the selection of a preferred alternative at this time. We simply do not know enough about the waste in the landfills, or the Characteristics of the site, to make a decision which adequately protects public health and the environment. In fact, evidence in the materials suggests that the proposed alternative will be insufficient to achieve these goals.

However. while we do not feel that enough information has been collected to make a final - - remedial action decision at this time, we strongly recommend that the Army immediately begin collecting the leachate that is now flowing into lake Michigan to prevent further pollution of the regionís drinking water source. 'i-I

In our opinion, the proposal lacks important information in two key areas:
Uncertainties About the Nature of the Wastes We are concerned that sampling performed both before and during the Phase I RI/RA does not provide enough information to characterize the waste in the landfills or the groundwater flow which comes into contact with the waste, and therefore, the potential risks to the human health and the environment posed by landfills 6 and 7.

The on-site disposal of military wastes was subject to very few regulations. The Army does predict that several materials that may be harmful to human health may exist in landfills 6 and 7, but tests for these dangerous contaminants have not been done. The proposal indicates that many types of hazardous contaminants are likely to occur, including potentially carcinogenic and radioactive materials. Also, discrepancies between the leachate and generated gas compositions, as discussed in Mr. Norris' review, suggest that more sampling is necessary to determine the nature of the waste.

The Army has stated that the characterization of waste will not affect the capping alternative because a RCRA cap is used for both hazardous and non-hazardous waste. However, if the wastes do prove to be hazardous, we do not believe a cap is adequate to protect human health and the environment, given the likely infiltration of the waste by groundwater from the sides and the bottom of the landfills. Since the natural discharge for groundwater in this area is into Lake Michigan, these pollutants will be carried into the Lake Michigan if only a surface cap is applied. Uncertainties About the Characteristics of the Site

The sampling completed at the sites does not sufficiently characterize site geology and groundwater flow. Many assumptions are made on how successful the preferred system will be in collecting leachate based upon groundwater evaluations which are incomplete and misinterpreted, such as the slug tests, as discussed in Mr. Norris' review. More sampling is needed, and a proper determination of the site's geology and groundwater resources is a necessity for the sound evaluation of data from sampling and the determination of a preferred remediation alternative.

Finally, we are concerned that the remedial action chosen through this process may indeed become the final action taken. We want to be assured that the alternative chosen for remedial action is the one which will most effectively protect Lake Michigan and neighboring residents. Unfortunately, the proposed action is not based on adequate information to make such a decision at this time.

Sincerely,
Jolie DiMonte Krasinsky
Critical lands Organizer
Sierra Club Great Lakes Ecoregion Program
Protecting the People. Wildlife and Beauty of the Great Lakes

Critical lands Organizer cc:
Paul Lake, Project Manager, IEPA
Owen Thompson, Project Manager, USEPA
Fred Bates, President, Advocates for the Public Interest in Fort Sheridan
Ilene Figel, Senior Planner, Highland Park
Mark Rooney, city Manager, Highwood
======


A Review of: Fort Sheridan
Landfills 6 and 7 Interim Action
Draft Final Focused Feasibility Study Proposed Plan Landfills 6 and 7 Interim Action USEPA and IEPA comments Prepared on behalf of and under contract to:
Illinois Chapter of the Sierra Club
1 North LaSalle
Chicago IL 60602 by Charles H. Norris
Geo-Hydro, Inc.
1928 E. 14th Ave
Denver CO 80206 (303)322-3171
(303) 322-8649 (fax)
cnorrisghi@aol.com

Introduction The Illinois Chapter of the Sierra Club retained Geo-Hydro, Inc. to perform a brief review the Fort Sheridan Landfills 6 and 7 Interim Action Draft Final Focused Feasibility Study (FFS) prepared for the United States Army Corps of Engineers by Environmental Science & Engineering, Inc. ESE, July 2, 1996). In addition to the review of the FFS, Geo-Hydro, Inc. also reviewed the following documents: Proposed Plan Landfills 6 and 7 Interim Action ESE, August, 1996, Fact Sheet Excavation Alternative -- Landfills 6 and 7 Interim Action ESE, July, 1996), comments by BeryI Flom (Restoration Advisory Board, May 7, 1996) and the response to these comments (ESE, May 7, 1996), comments by the Illinois Environmental Protection Agency (IEPA, June 5, 1996) on the April 22, 1996 comments to the Draft version of the FFS (ESE, December 5, 1995) and the response to these comments ESE, July 8, 1996), the comments of the United States Environmental Protection Agency (USEPA, May 23, 1996) and the response to these comments ESE, July 8, 1996), and the response to comments ('EPA, February 9, 1996) on the Draft version of the FFS. Information and data relied upon to produce the FFS, but that were not included in the FFS, were generally not available for study at the time of these reviews. The objectives of these reviews were to determine if the geologic, hydrogeologic, geochemical, and chemical characterizations at the site of these landfills were adequate to complete a focused feasibility study and select along possible alternatives, to determine whether the methodologies employed to investigate the geology and hydrogeology of sites are valid and appropriate, and to establish whether the conclusions related to geology and hydrogeology in the FFS are supported by the available data and consistent with accepted scientific principles. This report presents the findings of those reviews. General Comment on the FFS One is struck in the FFS by the discrepancy between the qualitative descriptions of the waste in the sections on alternatives 2 and 3, those involve capping, and the descriptions of the waste quality in alternative 4, the alternative of excavation and removal. When discussing the capping alternatives, including the preferred alternative of using both RCRA and modified RCRA caps, the FFS uses language that leaves the impression that landfills 6 and 7 contain little more than standard, generic municipal waste that carries little long-term risk if kept in place. In contrast, the language used in the portions of the FFS that concerns excavation of the waste creates the impression of highly noxious or toxic materials, difficult to handle, treat or move, and that are the likely source of substantial environmental risk or damage. The discrepancy may originate with a predisposition toward capping rather than excavating and hauling the waste It may also originate from the recognition that the contents of these landfills are simply not reliably known. Conservative possibilities are visualized for the dig-and-haul alternative, where everything will be exposed and handled, whereas under the capping alternatives, the contents of the landfills are perceived as being largely immaterial to the implementation and success of the alternative. Regardless of its source, the discrepancy is inappropriate. A pre-existing preference for any one alternative should not guide the descriptive language of the alternatives, potentially biasing the decision-making process. Further, and seemingly unrecognized in the discussions of alternatives in the FFS, the more dangerous or hazardous the material is to excavate and haul from an urban setting, the less appropriate it is to leave the material in the urban setting, in an unlined, only partially confined facility. It is also increasingly likely that the problem will outlast the lifetime of the proposed cap(s) and other engineered structures and the cost, timing, and implementability of replacing these features have not been factored into the capping alternatives. Finally, a comment on the continued existing discharge into Lake Michigan and the risks associated with gas generation and migration is appropriate. Neither of these risks should continue unmonitored and unmitigated until an orderly implementation of one or more of the alternatives. Explosive concentrations of methane are reported to develop within storm drains. The FFS does not, however, indicate where such concentrations exist, how far from the landfill they are found, whether or not individual residences are in danger of the build-up of methane to explosive concentrations, or whether a monitoring program exists to track methane concentrations and migration in drains, residences or utility corridors. Also unaddressed in the FFS are existing programs or plans for programs, prior to implementation of an alternative, to monitor or mitigate the migration of other landfill-generated gases (vinyl chloride at least) that migrate with the methane and may be of danger at significantly lower concentrations. It is also unacceptable to continue the existing lake discharges without mitigation prior to implementation of one of the alternatives. It is known, based upon data within the FFS, that at least during periods of low flow, the storm drainage being discharged into the lake does not meet the applicable standards. The current plan is for this discharge to continue unabated for years while a selected alternative is implemented. An immediately implemented program should be developed to capture low-flow storm drainage and leakage from the landfills for treatment and discharge to NSSD, regardless of the eventually chosen alternative. Waste and Leachate Characterization The characterization of waste and the leachate at landfills 6 and 7 are inadequate to reasonably define and chose among alternative actions. At least one waste stream that can greatly affect the choice among alternatives, (presumably low level) radioactive wastes, is identified as expected to be present, yet no attempt to evaluate it or its presence in leachate or groundwater is reported. The descriptions of the waste are very generalized, and on page 38 the FFS provides. The following egregious non sequitor in discussion the composition of the waste: Sample results to date indicate constituent concentrations that are within typical ranges for MSW landfills operated during the period from 1950 to 1980. There has been no sampling of the solid materials in the landfill. The descriptions of the waste are to varying degrees inconsistent with leachate and generated-gas compositions. Further, the leachate analyses are not highly consistent with each other, nor consistent with the compositions of the landfill gases. Leachate or potential leachate has been collected and analyzed from three different settings. Analyses were obtained from samples collected a~ leachate seeps, from low-flow water collected at the energy dissipation structure flowing into Lake Michigan, and from fluid collected in gas venting wells installed in the landfill. Samples that may or may not be impacted by leachate have been analyzed from groundwater wells adjacent to the landfills. The data available for review are from various sampling events and the analyses are for different constituents. Quantifying conclusion about the character of the leachate and particularly about typical compositions of the leachate is correspondingly difficult. A major inconsistency between the leachate composition and the gas composition is observed in the presence or absence and the distribution of halogenated compounds. The gas contains high concentrations of vinyl chloride, a common degradation product of heavier, more complex chlorinated organic compounds, often de-greasers or solvents. The qualitative description of the possible waste streams indicates that such compounds may have been dumped at the landfills. In spite of the occurrence of gas-phase vinyl chloride, none of the leachate samples or storm drainage samples show detectable concentrations of vinyl chloride or of precursor compounds. In fact, the only detectable concentrations observed were found in groundwater between the two landfills. This is a major inconsistency that must be resolved before the FFS can be considered complete or before alternatives can be adequately defined or evaluated. The inconsistency may in part be the result of sampling the gas vent wells only for leachate that occurs in the wells and not producing landfill leachate into the wells prior to sampling. The current well bore leachate that is sampled may well represent the result of well-centered bioreaction rather than represent leachate within the landfill away from the gas vent well. Consequently, projecting treatment costs and processes for any alternatives based upon the existing data may not be valid and are potentially highly inappropriate for actual leachate(s) that exists in the landfills.

The vinyl chloride concentrations in the landfill gas are consistent with concentrations that occur where co-disposal of chlorinated solvents occurred in high relative quantities with general municipal waste, quantities that can produce free-phase occurrence of these compounds in a landfill. Chlorinated solvents in free-phase are normally DNAPLs or "sinkers" that are heavier than water and will sink through the waste and leachate to move along and collect at the base of the landfill. In spite of the vinyl chloride evidence suggesting DNAPLs, no investigation has been undertaken to identify and locate these compounds. The presence or absence of DNAPL accumulations will impact the cost, effectiveness, and appropriateness of all alternatives except the no-action alternative. Until the possibility and extent of DNAPL accumulation is determined, the FFS is incomplete and the selection of a preferred alternative is inappropriate. The character of the leachate from landfill seeps and from the gas vents shows some general similarities; in particular, high TDS, high total iron, and low sulfate. These are distinctly different than the character of the liquid, interpreted as predominantly leachate, sampled at energy dissipation structure at the end of the storm drain feeding into Lake Michigan. Further, the increases in constituent concentrations between storm drainage above and below the landfills does not support the interpretation of 10 gpm infiltration of leachate into the storm drain. Unless and until the chemical differences can be explained, it must be concluded that the low-flow storm drain effluent is not primary leachate. It follows that any alternatives using the contrary conclusion as a foundation are potentially not valid. A number of additional inconsistencies in the FFS are observed in the discussions of possible impacts of leachate migration into and through groundwater. One of the more obvious of these is the discussions in the FFS centers on the concentrations of sulfate in groundwater. The observation that high concentrations of sulfate are found in wells interpreted as upgradient of the both landfills is used as an argument that landfill leachate is not responsible for high sulfate seen in some monitoring wells. This logic ignores the significance of the leachate chemical data. The leachates show low sulfate concentrations, not high concentrations. If one wants to track potential leachate impacts in the surrounding native materials, the key would be anomalously low sulfate concentrations, not anomalously high concentrations. In fact, monitoring well pair LF6MW04S and LF6MW04D installed between the landfills clearly show responses indicative of leachate migration through the soils between the landfills, and the shallower of these wells shows both vinyl chloride concentrations that exceed standards and the existence of one precursor compound of vinyl chloride. An additional inadequacy in the FFS is found in the discussion of gas generation rates in the landfill. The final paragraph in the discussion, on page 93, dismisses as unknown the most critical potential change in the landfill that will result from changes due to implementing one of the action alternatives. Considering that the greatest current risk associated with the landfills is gas-transported contaminants, it is absolutely unacceptable to accept as unknown the impact of changing leachate levels on gas generation rates. whether generation rates increase or decrease is particularly critical to the evaluation of the capping alternatives, where active gas collection is to be undertaken late relative to the action of lowering the leachate levels. Geology and Hydrogeology There area number of serious known or potential errors in the FFS related to geologic and hydrogeologic descriptions and their impacts on the implementation of the various proposed or preferred alternatives. Individually they may not constitute fatal flaws, but collectively they raise serious concerns regarding the overall validity of the conceptual understanding of the sites, and therefore of the validity of the selected alternative. The overall perception of the area of Fort Sheridan as one of low-permeability clay sediments and encased isolated lenses of silt, sand and gravel, with slow rates of transmission for ground water, is not supported by the topographic character of the area itself, the data within the FFS, or by the background ground water quality. The very existence of the multiple ravines that cut deeply and sharply into the clay ridge of the easternmost lake-border moraine over the short distance between Lake Michigan and the Skokie River is evidence of efficient transmission of ground water through and under the ridge. These ravines advance landward of the lake primarily through undercutting or sapping in response to efficient groundwater flow through fracture systems or interconnected silt and sand stringers, not primarily through the downcutting of surface drainage. The ravines not only influence groundwater due to seepage into the ravines, as observed in Section 1.2.1.4 Hydrology, they were created by the same seepage alluded to in the reference to bluff instability in the same Section. The existence of the system of interconnected, secondary porosity is also documented by observations of entirely different phenomena at different scales. The discussions in the FFS of the results of the laboratory and slug testing for permeability correctly note that the combined data support an interpretation of a flow system with a fracture (or other) secondary flow network. The implications of this observation, however, are not explored. The ground water migration rate calculations provided in Section 1.2.4.2. Groundwater and on Figure 1-8 are based not upon flow through a secondary system but on an assumed matrix effective porosity of 1004. If, for example, the fracture density produces a secondary porosity of 1%, the travel time is 10-fold less than that represented in the FFS. The groundwater composition also reflects both the effectiveness of the groundwater system and documents the speed of groundwater flow through the fracture networks. The high sulfate concentrations observed in the groundwater are the direct result of chemical changes to wetland soils induced by urbanization and development (oxidation of sulfide-bearing wetland soils as water tables are lowered) and reflect travel times through the ridge system measured in decades, not centuries. The potentiometric map presented in Figure 1-7 does not represent all of the data in the FFS and available for use. In particular, the heads for G-101, G-102, and LF7MW01 do not appear to have been considered in construction the map.

The head for LW7MW01 is between 20 and 25 feet the below the surface mapped in Figure 1-7. The head for G-l01 is between four and 13 feet below the mapped surface, and head of G-102 is between nine and 14 feet below the mapped surface. If these heads were included in the map on Figure 1-7, the complexity of the head distributions in and around landfills 6 and 7 are far more apparent. Including the head of G-101 alone in the mapped potentiometric surface would clearly demonstrate that the reduced head at GV-1 is far more locally restricted than suggested by the map in Figure 1-7. This in turn suggests that the ability to effectively use the existing, deep storm drain system or the gas vent wells is likely to be far less effective at draining leachate than is implied in the FFS. The relatively low head values shown in the other two excluded wells suggest that the "sink" observed at GV-1 may not necessarily or solely be due to the deep storm drain, but may alternatively or supplementally have a geologic component. Until this possibility is fully evaluated, none of the action alternatives can be evaluated adequately and even such fundamental components as monitoring well locations along probable migration paths cannot be determined The FFS refers in several places to monitoring well LF6MW0I as being an upgradient well with respect to the landfills. While this well is the most headward well in the ravine, the site characterization does not establish this well as upgradient. A comparison of the head level at LF6MW01 and the topography displayed on Figure I-4 suggests that the well may likely be downgradient of the landfills, at least at some times. The land surface at landfill 6 is depicted on Figure 14 as a bowl, with several closed depressions contoured. The elevations of the depressions nearest LF6MW01 are between 664 feet and 658 feet, some S to 14 feet above the head in LF6MW01. During periods of heavy recharge, one would expect mounding of leachate under the depressions, potentially to the land surface, based upon drainage observations discussed in the FFS. At such nines, LF6MW01 would be downgradient, not upgradient of the landfills The hydrogeologic conceptualization of the landfills relies heavily upon permeabilities obtained from slug tests from only three wells at the landfills and two wells elsewhere on Ft. Sheridan. of the three site wells, only one tested in situ soil materials below the ravine and none tested in situ materials adjacent to the ravine. The interpretations of the three site slug tests, obtained from the Remedial Investigation ESE, 1992), show that certainly in one case (LF'7MW04S), and possibly a second case (LF6MW04D), the interpretative model selected is inappropriate based upon the response of the well to the test. The data should be re-evaluated correctly. The interpretation of the 10 gpm low-flow storm drainage as leachate infiltration into the underlying storm drain is unsupported by either chemical or hydrogeologic/geologic data. This interpretation requires 13 inches of some 33 inches of average precipitation infiltrate annually and drain through the landfill into the storm drain. This would have to be in addition to precipitation that must infiltrate to provide leachate that is observed as seepage from the flanks and east face, drainage into shallow storm and surface drains, and flow into surrounding and underlying native soils. The 13 inches stated, let alone the undetermined total infiltration required, stretches credulity. Further, there appears to have been no effort to evaluate even qualitatively the seasonal fluctuations of heads, a direct indication of the infiltration component of water balance. Similarly, the MODFLOW model made no attempt to match any transient behavior of the landfill, an absolutely critical step in establishing the validity of any numerical modeling. The modeling done in support of the 10 gpm interpretation appears at best circular in reasoning and, as described, the structure of the model seems inadequate to actually test the concept. The flow net provided in Figure 1-8 is distinctly in conflict with the parameterization of the "successful" model. The former shows a flow pattern through landfill material and native soils that have little contrast in hydraulic properties (little or no deflection of potentiometric lines or flow lines at the contact between the materials), whereas the model had a 500-fold contrast between the landfill waste and the native soils. However, the model is not documented in the FFS for critical review, and this is a major deficiency for the study. The 10 gpm interpretation is so questionable and so fundamental to the evaluation of the various capping alternatives that it cannot be accepted based upon anything less than full critical review of all supporting evidence, including the modeling that is referenced. Conclusions and Recommendations The data within the FFS is not sufficient to support either the evaluation of the proposed alternatives or the selection of a preferred alternative. The inadequacy is observed among geologic, hydrogeologic, geochemical and chemical data. In addition to the insufficiency of the data within the ITS, conclusions drawn from the included data are frequently not supported by the data and can be at odds with accepted scientific principles. Prior to proceeding with any action alternatives, additional characterization is required for groundwater systems and the geology surrounding the site. Sampling and characterization of the solid waste should he done. Leachate sampling from the landfill mass should be undertaken to supplement the characterization of the fluid within the gas vent wells. Radiologic assessment of the various media at the site should be undertaken. A sampling program to determine the extent of the probable DNAPLs should be undertaken. Finally a realistic integration of the existing and new data should be performed before any evaluation of action alternatives is undertaken. In the interim, leachate escaping from the landfills from either lateral and face seeps or that may occur in storm drain effluent should be captured, treated as necessary, and appropriately discharged, rather than being permitted to dump into Lake Michigan. Also during the interim, gas levels (at least methane and vinyl chloride) in storm drains, utility corridors and residences should be monitored and vented as necessary as an ongoing mitigation program.
======


To: Sierra Club, Illinois Chapter
From: Charles H. Norris, Geo-Hydro, Inc.
Date: December 17, 1996

Re: Responses to Draft Responsiveness Summary. 17-1 The rationale by the authors of the response justifying the discrepancy between the qualitative descriptions of waste among the alternatives is at best circular. "The differences can be attributed to the differences in risk" is a restatement of the problem, not a resolution of it. Unless and until the waste contents of Landfills 6 and 7 are adequately characterized, it is inappropriate to presume relative risks, or to use alternative-dependent descriptions to bias perceptions of the risk. The acknowledgment by the authors of the response that these wastes must be presumed to include hazardous and/or industrial wastes (response to comment #1-1) does not absolve the government from the responsibility of determining the specific nature of the waste, rather it underscores the need for such determination. Some MSW landfills from this era that were the sites of co-disposal are among the Superifund sites.

The authors acknowledge the cost of cap replacement has not been included, because NCP doesn't require it, since it is presumed to last "indefinitely." This is a presumption that is insupportable. The comparison of the proposed cap at this site with that at an off-site facility is deliberate misdirection. Such a site would not likely be in an urban area, would not be adjacent to Lake Michigan, would have hill engineering containment of the waste, and would be a facility acknowledged and sited for its purpose, a hazardous or special waste disposal facility. Further, as pointed out by the authors of the responses, the off-site facility would have the additional advantages of requiring up-front payment, by way of disposal fees, for facility maintenance, and the government would clearly still be liable for the waste and would be unable to "walk" at some arbitrary or budgetary point in the future. Finally, the proffered computation of discounting cap repairs calls to mind Mark Twain s observations regarding the relationship between prevaricators and statisticians. A more meaningful calculation is that a cap repair today that costs $3,000, 000 will cost 7.7 times as much, more than $23,000,000, in 30 years, using the suggested 7% rate for discounting/inflation. 17-2 The authors choose not to respond to the concerns of lake releases in this comment, but rather to emphasize the extortionary nature of the proceedings. Four months have gone by since the observation was made that unabated discharge to the lake should be stopped immediately. The government responds that it doesn't even anticipate stopping the lake dumping until at least a year after the final selection of the capping alternative as "interim" remedy. The longer selection of this alternative takes, the longer the lake dumping will be permitted to continue. Until the 'EPA agrees to their monitoring pro grain, they will not even monitor storm drain discharges.

======

Sierra Club Great Lakes Ecoregion Program
Protecting the People. Wildlife and Beauty of the Great Lakes
September 9, 1996
Ms, Colleen Reilly BRAC Environmental Coordinator 3155 Blackhawk Drive, Suite 17 Fort Sheridan, Illinois 60037 Dear Ms. Reilly: Please accept these comments on the Proposed Plan Landfills 6 and 7 Interim Action (ESE, August, 1996) on behalf of the Sierra Club, Great Lakes critical Lands Project The general comments below are intended to supplement the attached review of the Proposed Plan, the Fort Sheridan Landfills 6 and 7 Interim Action Draft Final Focused Feasibility Study and other relevant materials, prepared by report from Charles H. Norris, a hydrogeologist with Geo-Hydro, Inc. on our behalf. We do not believe that the data provided in the Feasibility Study and the Phase I Draft Final Remedial Investigation-Risk Assessment (RI/RA), as summarized in the Proposed Plan, sufficiently supports either the evaluation or the selection of a preferred alternative at this time. We simply do not know enough about the waste in the landfills, or the Characteristics of the site, to make a decision which adequately protects public health and the environment. In fact, evidence in the materials suggests that the proposed alternative will be insufficient to achieve these goals. However. while we do not feel that enough information has been collected to make a final - - remedial action decision at this time, we strongly recommend that the Army immediately begin collecting the leachate that is now flowing into lake Michigan to prevent further pollution of the region's drinking water source. 'i-I In our opinion, the proposal lacks important information in two key areas: Uncertainties About the Nature of the Wastes We are concerned that sampling performed both before and during the Phase I RI/RA does not provide enough information to characterize the waste in the landfills or the groundwater flow which comes into contact with the waste, and therefore, the potential risks to the human health and the environment posed by landfills 6 and 7. The on-site disposal of military wastes was subject to very few regulations. The Army does predict that several materials that may be harmful to human health may exist in landfills 6 and 7, but tests for these dangerous contaminants have not been done. The proposal indicates that many types of hazardous contaminants are likely to occur, including potentially carcinogenic and radioactive materials. Also, discrepancies between the leachate and generated gas compositions, as discussed in Mr. Norris' review, suggest that more sampling is necessary to determine the nature of the waste. The Army has stated that the characterization of waste will not affect the capping alternative because a RCRA cap is used for both hazardous and non-hazardous waste. However, if the wastes do prove to be hazardous, we do not believe a cap is adequate to protect human health and the environment, given the likely infiltration of the waste by groundwater from the sides and the bottom of the landfills. Since the natural discharge for groundwater in this area is into Lake Michigan, these pollutants will be carried into the Lake Michigan if only a surface cap is applied. Uncertainties About the Characteristics of the Site The sampling completed at the sites does not sufficiently characterize site geology and groundwater flow. Many assumptions are made on how successful the preferred system will be in collecting leachate based upon groundwater evaluations which are incomplete and misinterpreted, such as the slug tests, as discussed in Mr. Norris' review. More sampling is needed, and a proper determination of the site's geology and groundwater resources is a necessity for the sound evaluation of data from sampling and the determination of a preferred remediation alternative. Finally, we are concerned that the remedial action chosen through this process may indeed become the final action taken. We want to be assured that the alternative chosen for remedial action is the one which will most effectively protect Lake Michigan and neighboring residents. Unfortunately, the proposed action is not based on adequate information to make such a decision at this time. Sincerely, Jolie DiMonte Krasinsky Critical lands Organizer cc: Paul Lake, Project Manager, IEPA Owen Thompson, Project Manager, USEPA Fred Bates, President, Advocates for the Public Interest in Fort Sheridan Ilene Figel, Senior Planner, Highland Park Mark Rooney, city Manager, Highwood ====== A Review of: Fort Sheridan Landfills 6 and 7 Interim Action Draft Final Focused Feasibility Study Proposed Plan Landfills 6 and 7 Interim Action USEPA and IEPA comments Prepared on behalf of and under contract to: Illinois Chapter of the Sierra Club 1 North LaSalle Chicago IL 60602 by Charles H. Norris Geo-Hydro, Inc. 1928 E. 14th Ave Denver CO 80206 (303)322-3171 (303) 322-8649 (fax) cnorrisghi@aol.com Introduction The Illinois Chapter of the Sierra Club retained Geo-Hydro, Inc. to perform a brief review the Fort Sheridan Landfills 6 and 7 Interim Action Draft Final Focused Feasibility Study (FFS) prepared for the United States Army Corps of Engineers by Environmental Science & Engineering, Inc. ESE, July 2, 1996). In addition to the review of the FFS, Geo-Hydro, Inc. also reviewed the following documents: Proposed Plan Landfills 6 and 7 Interim Action ESE, August, 1996, Fact Sheet Excavation Alternative -- Landfills 6 and 7 Interim Action ESE, July, 1996), comments by BeryI Flom (Restoration Advisory Board, May 7, 1996) and the response to these comments (ESE, May 7, 1996), comments by the Illinois Environmental Protection Agency (IEPA, June 5, 1996) on the April 22, 1996 comments to the Draft version of the FFS (ESE, December 5, 1995) and the response to these comments ESE, July 8, 1996), the comments of the United States Environmental Protection Agency (USEPA, May 23, 1996) and the response to these comments ESE, July 8, 1996), and the response to comments ('EPA, February 9, 1996) on the Draft version of the FFS. Information and data relied upon to produce the FFS, but that were not included in the FFS, were generally not available for study at the time of these reviews. The objectives of these reviews were to determine if the geologic, hydrogeologic, geochemical, and chemical characterizations at the site of these landfills were adequate to complete a focused feasibility study and select along possible alternatives, to determine whether the methodologies employed to investigate the geology and hydrogeology of sites are valid and appropriate, and to establish whether the conclusions related to geology and hydrogeology in the FFS are supported by the available data and consistent with accepted scientific principles. This report presents the findings of those reviews. General Comment on the FFS One is struck in the FFS by the discrepancy between the qualitative descriptions of the waste in the sections on alternatives 2 and 3, those involve capping, and the descriptions of the waste quality in alternative 4, the alternative of excavation and removal. When discussing the capping alternatives, including the preferred alternative of using both RCRA and modified RCRA caps, the FFS uses language that leaves the impression that landfills 6 and 7 contain little more than standard, generic municipal waste that carries little long-term risk if kept in place. In contrast, the language used in the portions of the FFS that concerns excavation of the waste creates the impression of highly noxious or toxic materials, difficult to handle, treat or move, and that are the likely source of substantial environmental risk or damage. The discrepancy may originate with a predisposition toward capping rather than excavating and hauling the waste It may also originate from the recognition that the contents of these landfills are simply not reliably known. Conservative possibilities are visualized for the dig-and-haul alternative, where everything will be exposed and handled, whereas under the capping alternatives, the contents of the landfills are perceived as being largely immaterial to the implementation and success of the alternative. Regardless of its source, the discrepancy is inappropriate. A pre-existing preference for any one alternative should not guide the descriptive language of the alternatives, potentially biasing the decision-making process. Further, and seemingly unrecognized in the discussions of alternatives in the FFS, the more dangerous or hazardous the material is to excavate and haul from an urban setting, the less appropriate it is to leave the material in the urban setting, in an unlined, only partially confined facility. It is also increasingly likely that the problem will outlast the lifetime of the proposed cap(s) and other engineered structures and the cost, timing, and implementability of replacing these features have not been factored into the capping alternatives. Finally, a comment on the continued existing discharge into Lake Michigan and the risks associated with gas generation and migration is appropriate. Neither of these risks should continue unmonitored and unmitigated until an orderly implementation of one or more of the alternatives. Explosive concentrations of methane are reported to develop within storm drains. The FFS does not, however, indicate where such concentrations exist, how far from the landfill they are found, whether or not individual residences are in danger of the build-up of methane to explosive concentrations, or whether a monitoring program exists to track methane concentrations and migration in drains, residences or utility corridors. Also unaddressed in the FFS are existing programs or plans for programs, prior to implementation of an alternative, to monitor or mitigate the migration of other landfill-generated gases (vinyl chloride at least) that migrate with the methane and may be of danger at significantly lower concentrations. It is also unacceptable to continue the existing lake discharges without mitigation prior to implementation of one of the alternatives. It is known, based upon data within the FFS, that at least during periods of low flow, the storm drainage being discharged into the lake does not meet the applicable standards. The current plan is for this discharge to continue unabated for years while a selected alternative is implemented. An immediately implemented program should be developed to capture low-flow storm drainage and leakage from the landfills for treatment and discharge to NSSD, regardless of the eventually chosen alternative. Waste and Leachate Characterization The characterization of waste and the leachate at landfills 6 and 7 are inadequate to reasonably define and chose among alternative actions. At least one waste stream that can greatly affect the choice among alternatives, (presumably low level) radioactive wastes, is identified as expected to be present, yet no attempt to evaluate it or its presence in leachate or groundwater is reported. The descriptions of the waste are very generalized, and on page 38 the FFS provides. The following egregious non sequitor in discussion the composition of the waste: Sample results to date indicate constituent concentrations that are within typical ranges for MSW landfills operated during the period from 1950 to 1980. There has been no sampling of the solid materials in the landfill. The descriptions of the waste are to varying degrees inconsistent with leachate and generated-gas compositions. Further, the leachate analyses are not highly consistent with each other, nor consistent with the compositions of the landfill gases. Leachate or potential leachate has been collected and analyzed from three different settings. Analyses were obtained from samples collected a~ leachate seeps, from low-flow water collected at the energy dissipation structure flowing into Lake Michigan, and from fluid collected in gas venting wells installed in the landfill. Samples that may or may not be impacted by leachate have been analyzed from groundwater wells adjacent to the landfills. The data available for review are from various sampling events and the analyses are for different constituents. Quantifying conclusion about the character of the leachate and particularly about typical compositions of the leachate is correspondingly difficult. A major inconsistency between the leachate composition and the gas composition is observed in the presence or absence and the distribution of halogenated compounds. The gas contains high concentrations of vinyl chloride, a common degradation product of heavier, more complex chlorinated organic compounds, often de-greasers or solvents. The qualitative description of the possible waste streams indicates that such compounds may have been dumped at the landfills. In spite of the occurrence of gas-phase vinyl chloride, none of the leachate samples or storm drainage samples show detectable concentrations of vinyl chloride or of precursor compounds. In fact, the only detectable concentrations observed were found in groundwater between the two landfills. This is a major inconsistency that must be resolved before the FFS can be considered complete or before alternatives can be adequately defined or evaluated. The inconsistency may in part be the result of sampling the gas vent wells only for leachate that occurs in the wells and not producing landfill leachate into the wells prior to sampling. The current well bore leachate that is sampled may well represent the result of well-centered bioreaction rather than represent leachate within the landfill away from the gas vent well. Consequently, projecting treatment costs and processes for any alternatives based upon the existing data may not be valid and are potentially highly inappropriate for actual leachate(s) that exists in the landfills. The vinyl chloride concentrations in the landfill gas are consistent with concentrations that occur where co-disposal of chlorinated solvents occurred in high relative quantities with general municipal waste, quantities that can produce free-phase occurrence of these compounds in a landfill. Chlorinated solvents in free-phase are normally DNAPLs or "sinkers" that are heavier than water and will sink through the waste and leachate to move along and collect at the base of the landfill. In spite of the vinyl chloride evidence suggesting DNAPLs, no investigation has been undertaken to identify and locate these compounds. The presence or absence of DNAPL accumulations will impact the cost, effectiveness, and appropriateness of all alternatives except the no-action alternative. Until the possibility and extent of DNAPL accumulation is determined, the FFS is incomplete and the selection of a preferred alternative is inappropriate. The character of the leachate from landfill seeps and from the gas vents shows some general similarities; in particular, high TDS, high total iron, and low sulfate. These are distinctly different than the character of the liquid, interpreted as predominantly leachate, sampled at energy dissipation structure at the end of the storm drain feeding into Lake Michigan. Further, the increases in constituent concentrations between storm drainage above and below the landfills does not support the interpretation of 10 gpm infiltration of leachate into the storm drain. Unless and until the chemical differences can be explained, it must be concluded that the low-flow storm drain effluent is not primary leachate. It follows that any alternatives using the contrary conclusion as a foundation are potentially not valid. A number of additional inconsistencies in the FFS are observed in the discussions of possible impacts of leachate migration into and through groundwater. One of the more obvious of these is the discussions in the FFS centers on the concentrations of sulfate in groundwater. The observation that high concentrations of sulfate are found in wells interpreted as upgradient of the both landfills is used as an argument that landfill leachate is not responsible for high sulfate seen in some monitoring wells. This logic ignores the significance of the leachate chemical data. The leachates show low sulfate concentrations, not high concentrations. If one wants to track potential leachate impacts in the surrounding native materials, the key would be anomalously low sulfate concentrations, not anomalously high concentrations. In fact, monitoring well pair LF6MW04S and LF6MW04D installed between the landfills clearly show responses indicative of leachate migration through the soils between the landfills, and the shallower of these wells shows both vinyl chloride concentrations that exceed standards and the existence of one precursor compound of vinyl chloride. An additional inadequacy in the FFS is found in the discussion of gas generation rates in the landfill. The final paragraph in the discussion, on page 93, dismisses as unknown the most critical potential change in the landfill that will result from changes due to implementing one of the action alternatives. Considering that the greatest current risk associated with the landfills is gas-transported contaminants, it is absolutely unacceptable to accept as unknown the impact of changing leachate levels on gas generation rates. whether generation rates increase or decrease is particularly critical to the evaluation of the capping alternatives, where active gas collection is to be undertaken late relative to the action of lowering the leachate levels. Geology and Hydrogeology There area number of serious known or potential errors in the FFS related to geologic and hydrogeologic descriptions and their impacts on the implementation of the various proposed or preferred alternatives. Individually they may not constitute fatal flaws, but collectively they raise serious concerns regarding the overall validity of the conceptual understanding of the sites, and therefore of the validity of the selected alternative. The overall perception of the area of Fort Sheridan as one of low-permeability clay sediments and encased isolated lenses of silt, sand and gravel, with slow rates of transmission for ground water, is not supported by the topographic character of the area itself, the data within the FFS, or by the background ground water quality. The very existence of the multiple ravines that cut deeply and sharply into the clay ridge of the easternmost lake-border moraine over the short distance between Lake Michigan and the Skokie River is evidence of efficient transmission of ground water through and under the ridge. These ravines advance landward of the lake primarily through undercutting or sapping in response to efficient groundwater flow through fracture systems or interconnected silt and sand stringers, not primarily through the downcutting of surface drainage. The ravines not only influence groundwater due to seepage into the ravines, as observed in Section 1.2.1.4 Hydrology, they were created by the same seepage alluded to in the reference to bluff instability in the same Section. The existence of the system of interconnected, secondary porosity is also documented by observations of entirely different phenomena at different scales. The discussions in the FFS of the results of the laboratory and slug testing for permeability correctly note that the combined data support an interpretation of a flow system with a fracture (or other) secondary flow network. The implications of this observation, however, are not explored. The ground water migration rate calculations provided in Section 1.2.4.2. Groundwater and on Figure 1-8 are based not upon flow through a secondary system but on an assumed matrix effective porosity of 1004. If, for example, the fracture density produces a secondary porosity of 1%, the travel time is 10-fold less than that represented in the FFS. The groundwater composition also reflects both the effectiveness of the groundwater system and documents the speed of groundwater flow through the fracture networks. The high sulfate concentrations observed in the groundwater are the direct result of chemical changes to wetland soils induced by urbanization and development (oxidation of sulfide-bearing wetland soils as water tables are lowered) and reflect travel times through the ridge system measured in decades, not centuries. The potentiometric map presented in Figure 1-7 does not represent all of the data in the FFS and available for use. In particular, the heads for G-101, G-102, and LF7MW01 do not appear to have been considered in construction the map. The head for LW7MW01 is between 20 and 25 feet the below the surface mapped in Figure 1-7. The head for G-l01 is between four and 13 feet below the mapped surface, and head of G-102 is between nine and 14 feet below the mapped surface. If these heads were included in the map on Figure 1-7, the complexity of the head distributions in and around landfills 6 and 7 are far more apparent. Including the head of G-101 alone in the mapped potentiometric surface would clearly demonstrate that the reduced head at GV-1 is far more locally restricted than suggested by the map in Figure 1-7. This in turn suggests that the ability to effectively use the existing, deep storm drain system or the gas vent wells is likely to be far less effective at draining leachate than is implied in the FFS. The relatively low head values shown in the other two excluded wells suggest that the "sink" observed at GV-1 may not necessarily or solely be due to the deep storm drain, but may alternatively or supplementally have a geologic component. Until this possibility is fully evaluated, none of the action alternatives can be evaluated adequately and even such fundamental components as monitoring well locations along probable migration paths cannot be determined The FFS refers in several places to monitoring well LF6MW0I as being an upgra dient well with respect to the landfills. While this well is the most headward well in the ravine, the site characterization does not establish this well as upgradient. A comparison of the head level at LF6MW01 and the topography displayed on Figure I-4 suggests that the well may likely be downgradient of the landfills, at least at some times. The land surface at landfill 6 is depicted on Figure 14 as a bowl, with several closed depressions contoured. The elevations of the depressions nearest LF6MW01 are between 664 feet and 658 feet, some S to 14 feet above the head in LF6MW01. During periods of heavy recharge, one would expect mounding of leachate under the depressions, potentially to the land surface, based upon drainage observations discussed in the FFS. At such nines, LF6MW01 would be downgradient, not upgradient of the landfills The hydrogeologic conceptualization of the landfills relies heavily upon permeabilities obtained from slug tests from only three wells at the landfills and two wells elsewhere on Ft. Sheridan. of the three site wells, only one tested in situ soil materials below the ravine and none tested in situ materials adjacent to the ravine. The interpretations of the three site slug tests, obtained from the Remedial Investigation ESE, 1992), show that certainly in one case (LF'7MW04S), and possibly a second case (LF6MW04D), the interpretative model selected is inappropriate based upon the response of the well to the test. The data should be re-evaluated correctly. The interpretation of the 10 gpm low-flow storm drainage as leachate infiltration into the underlying storm drain is unsupported by either chemical or hydrogeologic/geologic data. This interpretation requires 13 inches of some 33 inches of average precipitation infiltrate annually and drain through the landfill into the storm drain. This would have to be in addition to precipitation that must infiltrate to provide leachate that is observed as seepage from the flanks and east face, drainage into shallow storm and surface drains, and flow into surrounding and underlying native soils. The 13 inches stated, let alone the undetermined total infiltration required, stretches credulity. Further, there appears to have been no effort to evaluate even qualitatively the seasonal fluctuations of heads, a direct indication of the infiltration component of water balance. Similarly, the MODFLOW model made no attempt to match any transient behavior of the landfill, an absolutely critical step in establishing the validity of any numerical modeling. The modeling done in support of the 10 gpm interpretation appears at best circular in reasoning and, as described, the structure of the model seems inadequate to actually test the concept. The flow net provided in Figure 1-8 is distinctly in conflict with the parameterization of the "successful" model. The former shows a flow pattern through landfill material and native soils that have little contrast in hydraulic properties (little or no deflection of potentiometric lines or flow lines at the contact between the materials), whereas the model had a 500-fold contrast between the landfill waste and the native soils. However, the model is not documented in the FFS for critical review, and this is a major deficiency for the study. The 10 gpm interpretation is so questionable and so fundamental to the evaluation of the various capping alternatives that it cannot be accepted based upon anything less than full critical review of all supporting evidence, including the modeling that is referenced. Conclusions and Recommendations The data within the FFS is not sufficient to support either the evaluation of the proposed alternatives or the selection of a preferred alternative. The inadequacy is observed among geologic, hydrogeologic, geochemical and chemical data. In addition to the insufficiency of the data within the ITS, conclusions drawn from the included data are frequently not supported by the data and can be at odds with accepted scientific principles. Prior to proceeding with any action alternatives, additional characterization is required for groundwater systems and the geology surrounding the site. Sampling and characterization of the solid waste should he done. Leachate sampling from the landfill mass should be undertaken to supplement the characterization of the fluid within the gas vent wells. Radiologic assessment of the various media at the site should be undertaken. A sampling program to determine the extent of the probable DNAPLs should be undertaken. Finally a realistic integration of the existing and new data should be performed before any evaluation of action alternatives is undertaken. In the interim, leachate escaping from the landfills from either lateral and face seeps or that may occur in storm drain effluent should be captured, treated as necessary, and appropriately discharged, rather than being permitted to dump into Lake Michigan. Also during the interim, gas levels (at least methane and vinyl chloride) in storm drains, utility corridors and residences should be monitored and vented as necessary as an ongoing mitigation program. ====== To: Sierra Club, Illinois Chapter From: Charles H. Norris, Geo-Hydro, Inc. Date: December 17, 1996 Re: Responses to Draft Responsiveness Summary. 17-1 The rationale by the authors of the response justifying the discrepancy between the qualitative descriptions of waste among the alternatives is at best circular. "The differences can be attributed to the differences in risk" is a restatement of the problem, not a resolution of it. Unless and until the waste contents of Landfills 6 and 7 are adequately characterized, it is inappropriate to presume relative risks, or to use alternative-dependent descriptions to bias perceptions of the risk. The acknowledgment by the authors of the response that these wastes must be presumed to include hazardous and/or industrial wastes (response to comment #1-1) does not absolve the government from the responsibility of determining the specific nature of the waste, rather it underscores the need for such determination. Some MSW landfills from this era that were the sites of co-disposal are among the Superifund sites. The authors acknowledge the cost of cap replacement has not been included, because NCP doesn't require it, since it is presumed to last "indefinitely." This is a presumption that is insupportable. The comparison of the proposed cap at this site with that at an off-site facility is deliberate misdirection. Such a site would not likely be in an urban area, would not be adjacent to Lake Michigan, would have hill engineering containment of the waste, and would be a facility acknowledged and sited for its purpose, a hazardous or special waste disposal facility. Further, as pointed out by the authors of the responses, the off-site facility would have the additional advantages of requiring up-front payment, by way of disposal fees, for facility maintenance, and the government would clearly still be liable for the waste and would be unable to "walk" at some arbitrary or budgetary point in the future. Finally, the proffered computation of discounting cap repairs calls to mind Mark Twain s observations regarding the relationship between prevaricators and statisticians. A more meaningful calculation is that a cap repair today that costs $3,000, 000 will cost 7.7 times as much, more than $23,000,000, in 30 years, using the suggested 7% rate for discounting/inflation. 17-2 The authors choose not to respond to the concerns of lake releases in this comment, but rather to emphasize the extortionary nature of the proceedings. Four months have gone by since the observation was made that unabated discharge to the lake should be stopped immediately. The government responds that it doesn't even anticipate stopping the lake dumping until at least a year after the final selection of the capping alternative as "interim" remedy. The longer selection of this alternative takes, the longer the lake dumping will be permitted to continue. Until the 'EPA agrees to their monitoring pro grain, they will not even monitor storm drain discharges. The authors also choose not to respond to the concerns expressed in the comment regarding existing (prior to implementation of the capping alternative) monitoring and remediation programs with respect to migration and buildup of landfill gases into the surrounding environment and structures. The implication is that until they are permitted to make their final selection, they will do nothing. It is also noted that the referenced section of the FFS, Section 1.2.3.8. does not address the concern, it defines the concern. 17-3 The authors apparently have failed to read "the radiological assessment report prepared by the State of Illinois Department of Nuclear Safety that was provided in the FFS as Appendix I," or they would not have misstated that that document concludes "there is no significant hazard resulting from radioactive materials potentially disposed of in Landfill 7." The report went no further than to conclude that there is no current exposure hazard at the site. Any hazard from the site with respect to radioactive waste at the site cannot be determined without characterization of the radioactive waste and any still-existing containment of that waste. The inability of the authors of the responses td understand the inappropriate juxtaposition that was cited, and its potential to mislead, is perhaps one of the more telling examples of the mind set with which they approach this project. While it is enlightening, it is not something that lends itself to resolution or to belaboring. However, other portions of the response do need to be addressed. First, the fact that may MSW landfills of this vintage were co-disposal sites for hazardous wastes does not reduce the concern - many such sites are now Superfund sites. Second, the suggestion that such materials at this site are immobile if present cannot be supported by the existing site data. The existing site data from different media are inconsistent. The best interpretation of the inconsistency is that the leachate sampling protocols for the gas vents are grossly inadequate. Sampling and analysis of the solid waste would assist in resolving the inconsistencies, as well as helping to define source terms for risk analysis and physical properties of the wastes. One of the inconsistencies observed was the detection of vinyl chloride in the effluent gas at the Landfill 7 gas vents, whereas there is no vinyl chloride or precursor compounds (chlorinated solvents, an expected material for this landfill) detected in the fluids collected at the gas vents. Perimeter air samples around the landfill did detect precursor compounds, and one monitoring well did detect vinyl chloride. The authors of the responses rationalize the lack of detection in water at the gas vent wells as being due to detection limits for the water that are too high to detect dissolved vinyl chloride that would be in equilibrium with the gaseous phase, referencing an analytical detection limit in the water of 2.0 mg/L. This is simply false. The reported detection limit in Table 1-7 is 0.010 mg/L, 200-fold lower than the stated number, and the analytical detection limit reported in Table 1.7 for vinyl chloride is 0.0026 mg/L, more than 760-fold lower than the number used in the response. Bottom line - at the reported gas concentrations, vinyl chloride or a predecessor compound should be observed in the liquid. That they aren't is a direct measure of the inadequacy of the leachate sampling protocol and the analyses to date. There is an indirect measure of the understanding by the government of the in adequacy of the existing leachate analyses. The work plan for this fall at the site included sampling the gas vents again, but this time purging the water from the vents first (normal procedure). The probability that this leachate is itself hazardous material is reflected in the fact that the purged liquid is to be dumped back down the vent, rather than properly disposing of it based upon its chemical characteristics. 17-5 In response to this comment, the authors reference the response to comment 1-32. Among the assertions in that response is the statement that "[t]he leachate collection system would intercept and collect DNAPL present. "This is without question an overstatement, if not a known erroneous statement. There is no evidence anywhere in the record that the storm drainage system is positioned or open in areas where DNAPL is or may be within the landfill. It is also important to recognize that the impacts of DNAPL is not primarily on the functionality of the landfill cap, but on the composition and treatability of the landfill leachates and landfill gases. Simple flaring will not destroy some chlorinated compounds. 17-6, 17-15 The authors of the responses maintain that mass balance cannot be used to verify that the low flow of 10 gpm may be leachate(s) from Landfills 6 and 7. In fact, this is one of the few methods available to independently verify the foundation of the house of cards that is based upon the assumption of current leachate drainage of 10 gpm. It is clear from the data provided in the FFS that the composition of the 10 gpm is highly inconsistent with the composition of any of the leachate compositions, but that it is fully consistent with the composition of the water from storm drains upgradient to the landfills. The chemical data overwhelmingly indicate that the water that constitutes the bulk of the low-period flow is not leachate from these landfills, but rather is storm-system drainage and "that unknown buried drains still exist that historically discharged to the ravine, or directly to the storm drain pipe..." (FFS, p24.) In response to the comments regarding the reasonableness of 13 equivalent inches of precipitation, the authors branch into what can best be considered a rambling discussion of ways that that can be accounted for without actually having precipitation infiltration of 13 inches/year. Among the more creative ideas is one that up to 3 gpm is groundwater flowing laterally into the waste, in spite of the fact that the landfills are everywhere in the FFS described as being characterized by outward gradients; i.e., groundwater flow from the landfill into the surrounding soils at rates of from 0.1 to 3 gpm, not inward into the landfill. (The outward gradient is used to justify the 10 feet of over-excavation of native soils in the dig-and-haul alternative.) The added losses to lateral leachate seeps is dismissed as being only that due to evaporation and transpiration. The losses to such mechanisms from these wet, swampy areas (App. I, IDNS) is not a trivial amount and should have been considered in performing a reasonable water balance for the site, and not dismissed out of hand. A check on the infiltration rates can be done by considering the seasonal head variations, as suggested in the comments. In response, rather than performing such a calculation, the authors choose merely to repeat data in the FFS, stating that seasonal head losses in the waste are about 5 feet. If the drainage porosity in the waste is as high as 10%, this seasonal drainage represents only about 6 inches of percolation, not 13 inches. 17-7 The authors of the responses acknowledge that the groundwater in LF6MW04S has been impacted by leachate migration in that area. 17-8 The authors indicate that sufficient information exists to determine whether increases or decreases in gas generation will result from lowering the leachate levels. Any protection that is to be afforded must come from the willingness to add to program costs by changing the sequence of engineering activities in response to unspecified additional, continuing monitoring during the installation of the cap. An example is to install the gas flare sooner than presently planned. [It is not clear what type of flare will destroy vinyl chloride.] The excuse of insufficient information is particularly hollow since characterization of the solid waste would provide the types of information (moisture contents, porosities, drainage behavior, bulk densities, etc.) that would permit the behavior of the gas generation at these landfills to be more reliably projected and the public to be protected proactively and not reactively. 17-9 Whether or not the present site characterization is consistent with a body of regional published data is irrelevant if it does not account for site specific conditions. And, while there may be a plethora of geologic data the authors of the responses choose to interpret in a given manner, the fact is that that data also fits an alternative conceptual model that demonstrates far less isolation of the landfills. The authors err when they opine that the erosional process of sapping does not involve an active groundwater flow system. Such flow systems define the essence of erosion by sapping as opposed to down-cutting. Such a flow system does not require continuous lenses of silt, sand or gravel, nor does the geology of the area support such systems. However, flow Systems with discontinuous lenses or layers of silt, sand, and gravel that are connected by fracture systems provide the necessary flow network to develop the geomorphology of the ravines. The authors are correct when they point out that were the sapping mechanism and the interconnected groundwater flow systems to be present and active, field evidence would include visible seeps and springs (even in the dry season), benching and terracing. It appears that the authors, however, have apparently never been down into the ravines, based upon the assertion that these features are not present. The presence of these features is apparent to anyone who takes the time to climb down and brush away the fallen leaves or push aside the hydrophilic vegetation at the seeps. These readily observed and characteristic features of sapping are why the Illinois Chicago Circle geology department brings field trips to the north shore ravines; they are one of the clearest examples available in the Midwest to demonstrate an erosional process normally associated with canyon-land development in the arid west. 17-10 The authors of the responses are undoubtedly aware that, while the rigorous characterization and evaluation of a dual porosity system would require far more data and different data than has yet been obtained for this site, the approximation suggested in the comment they are addressing is accurate to with a few percent and is all that is available from the data provided in the site documents. They are also unquestionably aware that, while "average or bulk values for the porous medium as a whole are most commonly used [as was done in the FFS characterization] in evaluating groundwater flow rates (Dominico and Schwartz, 1990)," neither Pat Dominico nor Frank Schwartz is advocating using the average bulk porosity of the entire system for computation of contaminant travel times. The three empirical observations called upon by the authors in support of no secondary porosity system are as applicable a system with secondary porosity as to one without. Both the first and second erroneously attribute head distributions solely to variations or contrasts in hydraulic conductivity. In fact, both phenomena discussed represent a balance between hydraulic conductivity and the flux of water through the system. Exactly the same head/gradient distributions exist even if the hydraulic conductivity is doubled, provided it is acknowledged that the mass of water moving through the system is also doubled. The significance to the site and the selected alternative is that the effectiveness of the cap and leachate drain for keeping water out of the waste and leachate production minimized are undermined if there is an unrecognized, contributing network of secondary porosity. Similarly, the now-reported lack of alteration areolae along fractures indicates only that oxygen has not historically moved through the fractures, not that water has and is not Moving through the fractures, As discussed in the original comment, the relatively high sulfate levels in the monitoring wells are suggestive of the oxidation of sulfide-rich soils associated with upland forest wetlands, as likely existed in the area prior to urbanization. Such soils would consume any oxygen in the groundwater before it could affect the tills adjacent to the conduit fractures. Thus, the system described in the comment not only is consistent with the sulfate concentrations observed as background at the site, but also is consistent with the introduced observation of lack of alteration areolae. The importance of the secondary porosity system is particularly significant once the leachate collection has stopped. Without withdrawing leachate from the landfills, they will resaturate and establish a gradient out of the landfill rather than into the landfill. Whether or not the secondary porosity has been yet evaluated with respect to average hydraulic conductivities, the travel times will be less through it than through an equivalent single-porosity system. Any attenuative properties of the natural system will also be reduced because of reduced contact and reduced contact time. 17-11, 17-12 The authors of the responses either don't understand the intent and Significance of the comment and the underlying data, or are deliberately attempting to cloud the issue. Wherever a vertical gradient is observed in or under the landfill, the gradient is downward. That is, the phreatic surface is higher than the head associated with the deeper interval. One can, therefore, project that the phreatic head at the G-101 location is as high or higher than the G-101 head, just as the LW6MW04S head is higher than the LW6MW04D head, as seen in Table 1-2. As mapped, the phreatic head at G-101 is interpreted at an elevation of about 642 feet. As pointed out in the response, the head at G-101, and thus the minimum phreatic head at that location, is about 653.5 feet, 10 or more feet higher than mapped. This means that the head gradient toward the manhole at the phreatic surface is significantly steeper than that mapped on Figure 1-7. Since the flux into manhole is whatever it is, the steeper gradient means that the hydraulic conductivity is lower than that implied by the Figure 1-7 interpretation. The manhole may be draining an area around itself but tat area is less than that displayed on Figure 1-7 and that suggests that drainage of both gas and liquid from the waste may be more difficult than inferred from the incorrect mapping. 17-13 The lack of head data for LF6 and the described standing and ponding water make the concern more than theoretical. Further, the distance of 100 feet from the waste boundary is not necessarily significant if there are preferred pathways from the waste to the well. If monitoring points have now been installed as part of the fall program, it may now be possible to determine at least what the dry-season gradient relationship is between leachate levels in LF6 and LF6MW01. 17-14 It is disturbing that, even upon review of the slug tests mentioned in the comment, no alternative interpretation was made. It is clear from the response of the wells to the slug test that the assumptions for the Bouwer and Rice interpretation method are not met by the wells. Were the Bouwer and Rice assumptions appropriate for the wells, the slug test data would have generated a semi-log plot of residual head (log) vs. time (linear) that is a straight line. It is the slope of this straight line that provides the solution in the Bouwer and Rice method. However, the data do not generate this straight line. Rather, they form curvilinear patterns with a continuously decreasing slope with time. Since the measured-data trends do not conform to trends generated by hydrologic systems assumed by the method, the site hydrologic system does not conform to the system assumed by the interpretation method. In the case of LF7MW04S, partial data reported in the groundwater classification document permitted a limited evaluation of the test results by the Cooper, et al. method (confined system). It appears the data conform more closely with the Cooper type curves than to the Bouwer and Rice straight line, with a hydraulic conductivity at least 5-fold greater and a lower storage coefficient. More important, however, than the revised hydraulic conductivity is the data-derived observation that even in the area of the beach and at shallow levels, the flow system appears to be a confined, not unconfined, system. The authors acknowledge that additional hydraulic conductivity determinations are necessary to characterize the soils adjacent to and beneath the landfills. It is unfortunate that it is not recognized that this data is needed before it is appropriate to make the final selection of the alternative for the landfills. 17-16 The limitation to the MODFLOW modeling and the down-grading of its significance in the response is appropriate and welcomed. It is somewhat surprising that the authors to the responses profess to not understand the critical nature of the 10 gpm figure with respect to the consideration of action alternatives, particularly given their emphasis on, and defense of it. The fundamental significance of the l0-gpm-as-leachate assumption is that it provides the foundation upon which the interpreted fluxes through the landfill and surrounding soils are built. 10 gpm is the maximum (and stretched) infiltration rate that can reasonably be postulated for the landfills. If the 10 gpm discharge were leachate, there could be very little left That is escaping through the flanks or sides of the ravine. Since the gradients are relatively high, that dictates a very low hydraulic conductivity, without the need for testing. If the hydraulic conductivity is very low, there will be very little lateral flow into the landfill once leachate levels are reduced, and leachate levels are kept low with low leachate pumping and treatment costs. Once active management ceases, the landfill will be slow to recharge and whatever the final head configuration, water fluxes and corresponding risks will be minimal. Correspondingly, if the 10 gpm discharge is not leachate, none of the other flux terms for the landfills are known. Lateral and downward flow could be significantly higher with correspondingly higher hydraulic conductivities. It would be more difficult to lower leachate levels and require more pumping and treatment to keep them low. Once management ceases, the drained wastes will recharge more quickly and outward fluxes will be greater. The authors are correct in observing that any differences in the flow system will affect all action alternatives. However, it will not affect them all equally. A major difference between the capping and the excavation alternatives is that the former is forever, the waste is always in the path of groundwater flow toward and through Wells Ravine. -- ====== Sierra Club Great Lakes Ecoregion Program Protecting the People. Wildlife and Beauty of the Great Lakes September 9, 1996 Ms, Colleen Reilly BRAC Environmental Coordinator 3155 Blackhawk Drive, Suite 17 Fort Sheridan, Illinois 60037 Dear Ms. Reilly: Please accept these comments on the Proposed Plan Landfills 6 and 7 Interim Action (ESE, August, 1996) on behalf of the Sierra Club, Great Lakes critical Lands Project The general comments below are intended to supplement the attached review of the Proposed Plan, the Fort Sheridan Landfills 6 and 7 Interim Action Draft Final Focused Feasibility Study and other relevant materials, prepared by report from Charles H. Norris, a hydrogeologist with Geo-Hydro, Inc. on our behalf. We do not believe that the data provided in the Feasibility Study and the Phase I Draft Final Remedial Investigation-Risk Assessment (RI/RA), as summarized in the Proposed Plan, sufficiently supports either the evaluation or the selection of a preferred alternative at this time. We simply do not know enough about the waste in the landfills, or the Characteristics of the site, to make a decision which adequately protects public health and the environment. In fact, evidence in the materials suggests that the proposed alternative will be insufficient to achieve these goals. However. while we do not feel that enough information has been collected to make a final - - remedial action decision at this time, we strongly recommend that the Army immediately begin collecting the leachate that is now flowing into lake Michigan to prevent further pollution of the region's drinking water source. 'i-I In our opinion, the proposal lacks important information in two key areas: Uncertainties About the Nature of the Wastes We are concerned that sampling performed both before and during the Phase I RI/RA does not provide enough information to characterize the waste in the landfills or the groundwater flow which comes into contact with the waste, and therefore, the potential risks to the human health and the environment posed by landfills 6 and 7. The on-site disposal of military wastes was subject to very few regulations. The Army does predict that several materials that may be harmful to human health may exist in landfills 6 and 7, but tests for these dangerous contaminants have not been done. The proposal indicates that many types of hazardous contaminants are likely to occur, including potentially carcinogenic and radioactive materials. Also, discrepancies between the leachate and generated gas compositions, as discussed in Mr. Norris' review, suggest that more sampling is necessary to determine the nature of the waste. The Army has stated that the characterization of waste will not affect the capping alternative because a RCRA cap is used for both hazardous and non-hazardous waste. However, if the wastes do prove to be hazardous, we do not believe a cap is adequate to protect human health and the environment, given the likely infiltration of the waste by groundwater from the sides and the bottom of the landfills. Since the natural discharge for groundwater in this area is into Lake Michigan, these pollutants will be carried into the Lake Michigan if only a surface cap is applied. Uncertainties About the Characteristics of the Site The sampling completed at the sites does not sufficiently characterize site geology and groundwater flow. Many assumptions are made on how successful the preferred system will be in collecting leachate based upon groundwater evaluations which are incomplete and misinterpreted, such as the slug tests, as discussed in Mr. Norris' review. More sampling is needed, and a proper determination of the site's geology and groundwater resources is a necessity for the sound evaluation of data from sampling and the determination of a preferred remediation alternative. Finally, we are concerned that the remedial action chosen through this process may indeed become the final action taken. We want to be assured that the alternative chosen for remedial action is the one which will most effectively protect Lake Michigan and neighboring residents. Unfortunately, the proposed action is not based on adequate information to make such a decision at this time. Sincerely, Jolie DiMonte Krasinsky Critical lands Organizer cc: Paul Lake, Project Manager, IEPA Owen Thompson, Project Manager, USEPA Fred Bates, President, Advocates for the Public Interest in Fort Sheridan Ilene Figel, Senior Planner, Highland Park Mark Rooney, city Manager, Highwood ====== A Review of: Fort Sheridan Landfills 6 and 7 Interim Action Draft Final Focused Feasibility Study Proposed Plan Landfills 6 and 7 Interim Action USEPA and IEPA comments Prepared on behalf of and under contract to: Illinois Chapter of the Sierra Club 1 North LaSalle Chicago IL 60602 by Charles H. Norris Geo-Hydro, Inc. 1928 E. 14th Ave Denver CO 80206 (303)322-3171 (303) 322-8649 (fax) cnorrisghi@aol.com Introduction The Illinois Chapter of the Sierra Club retained Geo-Hydro, Inc. to perform a brief review the Fort Sheridan Landfills 6 and 7 Interim Action Draft Final Focused Feasibility Study (FFS) prepared for the United States Army Corps of Engineers by Environmental Science & Engineering, Inc. ESE, July 2, 1996). In addition to the review of the FFS, Geo-Hydro, Inc. also reviewed the following documents: Proposed Plan Landfills 6 and 7 Interim Action ESE, August, 1996, Fact Sheet Excavation Alternative -- Landfills 6 and 7 Interim Action ESE, July, 1996), comments by BeryI Flom (Restoration Advisory Board, May 7, 1996) and the response to these comments (ESE, May 7, 1996), comments by the Illinois Environmental Protection Agency (IEPA, June 5, 1996) on the April 22, 1996 comments to the Draft version of the FFS (ESE, December 5, 1995) and the response to these comments ESE, July 8, 1996), the comments of the United States Environmental Protection Agency (USEPA, May 23, 1996) and the response to these comments ESE, July 8, 1996), and the response to comments ('EPA, February 9, 1996) on the Draft version of the FFS. Information and data relied upon to produce the FFS, but that were not included in the FFS, were generally not available for study at the time of these reviews. The objectives of these reviews were to determine if the geologic, hydrogeologic, geochemical, and chemical characterizations at the site of these landfills were adequate to complete a focused feasibility study and select along possible alternatives, to determine whether the methodologies employed to investigate the geology and hydrogeology of sites are valid and appropriate, and to establish whether the conclusions related to geology and hydrogeology in the FFS are supported by the available data and consistent with accepted scientific principles. This report presents the findings of those reviews. General Comment on the FFS One is struck in the FFS by the discrepancy between the qualitative descriptions of the waste in the sections on alternatives 2 and 3, those involve capping, and the descriptions of the waste quality in alternative 4, the alternative of excavation and removal. When discussing the capping alternatives, including the preferred alternative of using both RCRA and modified RCRA caps, the FFS uses language that leaves the impression that landfills 6 and 7 contain little more than standard, generic municipal waste that carries little long-term risk if kept in place. In contrast, the language used in the portions of the FFS that concerns excavation of the waste creates the impression of highly noxious or toxic materials, difficult to handle, treat or move, and that are the likely source of substantial environmental risk or damage. The discrepancy may originate with a predisposition toward capping rather than excavating and hauling the waste It may also originate from the recognition that the contents of these landfills are simply not reliably known. Conservative possibilities are visualized for the dig-and-haul alternative, where everything will be exposed and handled, whereas under the capping alternatives, the contents of the landfills are perceived as being largely immaterial to the implementation and success of the alternative. Regardless of its source, the discrepancy is inappropriate. A pre-existing preference for any one alternative should not guide the descriptive language of the alternatives, potentially biasing the decision-making process. Further, and seemingly unrecognized in the discussions of alternatives in the FFS, the more dangerous or hazardous the material is to excavate and haul from an urban setting, the less appropriate it is to leave the material in the urban setting, in an unlined, only partially confined facility. It is also increasingly likely that the problem will outlast the lifetime of the proposed cap(s) and other engineered structures and the cost, timing, and implementability of replacing these features have not been factored into the capping alternatives. Finally, a comment on the continued existing discharge into Lake Michigan and the risks associated with gas generation and migration is appropriate. Neither of these risks should continue unmonitored and unmitigated until an orderly implementation of one or more of the alternatives. Explosive concentrations of methane are reported to develop within storm drains. The FFS does not, however, indicate where such concentrations exist, how far from the landfill they are found, whether or not individual residences are in danger of the build-up of methane to explosive concentrations, or whether a monitoring program exists to track methane concentrations and migration in drains, residences or utility corridors. Also unaddressed in the FFS are existing programs or plans for programs, prior to implementation of an alternative, to monitor or mitigate the migration of other landfill-generated gases (vinyl chloride at least) that migrate with the methane and may be of danger at significantly lower concentrations. It is also unacceptable to continue the existing lake discharges without mitigation prior to implementation of one of the alternatives. It is known, based upon data within the FFS, that at least during periods of low flow, the storm drainage being discharged into the lake does not meet the applicable standards. The current plan is for this discharge to continue unabated for years while a selected alternative is implemented. An immediately implemented program should be developed to capture low-flow storm drainage and leakage from the landfills for treatment and discharge to NSSD, regardless of the eventually chosen alternative. Waste and Leachate Characterization The characterization of waste and the leachate at landfills 6 and 7 are inadequate to reasonably define and chose among alternative actions. At least one waste stream that can greatly affect the choice among alternatives, (presumably low level) radioactive wastes, is identified as expected to be present, yet no attempt to evaluate it or its presence in leachate or groundwater is reported. The descriptions of the waste are very generalized, and on page 38 the FFS provides. The following egregious non sequitor in discussion the composition of the waste: Sample results to date indicate constituent concentrations that are within typical ranges for MSW landfills operated during the period from 1950 to 1980. There has been no sampling of the solid materials in the landfill. The descriptions of the waste are to varying degrees inconsistent with leachate and generated-gas compositions. Further, the leachate analyses are not highly consistent with each other, nor consistent with the compositions of the landfill gases. Leachate or potential leachate has been collected and analyzed from three different settings. Analyses were obtained from samples collected a~ leachate seeps, from low-flow water collected at the energy dissipation structure flowing into Lake Michigan, and from fluid collected in gas venting wells installed in the landfill. Samples that may or may not be impacted by leachate have been analyzed from groundwater wells adjacent to the landfills. The data available for review are from various sampling events and the analyses are for different constituents. Quantifying conclusion about the character of the leachate and particularly about typical compositions of the leachate is correspondingly difficult. A major inconsistency between the leachate composition and the gas composition is observed in the presence or absence and the distribution of halogenated compounds. The gas contains high concentrations of vinyl chloride, a common degradation product of heavier, more complex chlorinated organic compounds, often de-greasers or solvents. The qualitative description of the possible waste streams indicates that such compounds may have been dumped at the landfills. In spite of the occurrence of gas-phase vinyl chloride, none of the leachate samples or storm drainage samples show detectable concentrations of vinyl chloride or of precursor compounds. In fact, the only detectable concentrations observed were found in groundwater between the two landfills. This is a major inconsistency that must be resolved before the FFS can be considered complete or before alternatives can be adequately defined or evaluated. The inconsistency may in part be the result of sampling the gas vent wells only for leachate that occurs in the wells and not producing landfill leachate into the wells prior to sampling. The current well bore leachate that is sampled may well represent the result of well-centered bioreaction rather than represent leachate within the landfill away from the gas vent well. Consequently, projecting treatment costs and processes for any alternatives based upon the existing data may not be valid and are potentially highly inappropriate for actual leachate(s) that exists in the landfills. The vinyl chloride concentrations in the landfill gas are consistent with concentrations that occur where co-disposal of chlorinated solvents occurred in high relative quantities with general municipal waste, quantities that can produce free-phase occurrence of these compounds in a landfill. Chlorinated solvents in free-phase are normally DNAPLs or "sinkers" that are heavier than water and will sink through the waste and leachate to move along and collect at the base of the landfill. In spite of the vinyl chloride evidence suggesting DNAPLs, no investigation has been undertaken to identify and locate these compounds. The presence or absence of DNAPL accumulations will impact the cost, effectiveness, and appropriateness of all alternatives except the no-action alternative. Until the possibility and extent of DNAPL accumulation is determined, the FFS is incomplete and the selection of a preferred alternative is inappropriate. The character of the leachate from landfill seeps and from the gas vents shows some general similarities; in particular, high TDS, high total iron, and low sulfate. These are distinctly different than the character of the liquid, interpreted as predominantly leachate, sampled at energy dissipation structure at the end of the storm drain feeding into Lake Michigan. Further, the increases in constituent concentrations between storm drainage above and below the landfills does not support the interpretation of 10 gpm infiltration of leachate into the storm drain. Unless and until the chemical differences can be explained, it must be concluded that the low-flow storm drain effluent is not primary leachate. It follows that any alternatives using the contrary conclusion as a foundation are potentially not valid. A number of additional inconsistencies in the FFS are observed in the discussions of possible impacts of leachate migration into and through groundwater. One of the more obvious of these is the discussions in the FFS centers on the concentrations of sulfate in groundwater. The observation that high concentrations of sulfate are found in wells interpreted as upgradient of the both landfills is used as an argument that landfill leachate is not responsible for high sulfate seen in some monitoring wells. This logic ignores the significance of the leachate chemical data. The leachates show low sulfate concentrations, not high concentrations. If one wants to track potential leachate impacts in the surrounding native materials, the key would be anomalously low sulfate concentrations, not anomalously high concentrations. In fact, monitoring well pair LF6MW04S and LF6MW04D installed between the landfills clearly show responses indicative of leachate migration through the soils between the landfills, and the shallower of these wells shows both vinyl chloride concentrations that exceed standards and the existence of one precursor compound of vinyl chloride. An additional inadequacy in the FFS is found in the discussion of gas generation rates in the landfill. The final paragraph in the discussion, on page 93, dismisses as unknown the most critical potential change in the landfill that will result from changes due to implementing one of the action alternatives. Considering that the greatest current risk associated with the landfills is gas-transported contaminants, it is absolutely unacceptable to accept as unknown the impact of changing leachate levels on gas generation rates. whether generation rates increase or decrease is particularly critical to the evaluation of the capping alternatives, where active gas collection is to be undertaken late relative to the action of lowering the leachate levels. Geology and Hydrogeology There area number of serious known or potential errors in the FFS related to geologic and hydrogeologic descriptions and their impacts on the implementation of the various proposed or preferred alternatives. Individually they may not constitute fatal flaws, but collectively they raise serious concerns regarding the overall validity of the conceptual understanding of the sites, and therefore of the validity of the selected alternative. The overall perception of the area of Fort Sheridan as one of low-permeability clay sediments and encased isolated lenses of silt, sand and gravel, with slow rates of transmission for ground water, is not supported by the topographic character of the area itself, the data within the FFS, or by the background ground water quality. The very existence of the multiple ravines that cut deeply and sharply into the clay ridge of the easternmost lake-border moraine over the short distance between Lake Michigan and the Skokie River is evidence of efficient transmission of ground water through and under the ridge. These ravines advance landward of the lake primarily through undercutting or sapping in response to efficient groundwater flow through fracture systems or interconnected silt and sand stringers, not primarily through the downcutting of surface drainage. The ravines not only influence groundwater due to seepage into the ravines, as observed in Section 1.2.1.4 Hydrology, they were created by the same seepage alluded to in the reference to bluff instability in the same Section. The existence of the system of interconnected, secondary porosity is also documented by observations of entirely different phenomena at different scales. The discussions in the FFS of the results of the laboratory and slug testing for permeability correctly note that the combined data support an interpretation of a flow system with a fracture (or other) secondary flow network. The implications of this observation, however, are not explored. The ground water migration rate calculations provided in Section 1.2.4.2. Groundwater and on Figure 1-8 are based not upon flow through a secondary system but on an assumed matrix effective porosity of 1004. If, for example, the fracture density produces a secondary porosity of 1%, the travel time is 10-fold less than that represented in the FFS. The groundwater composition also reflects both the effectiveness of the groundwater system and documents the speed of groundwater flow through the fracture networks. The high sulfate concentrations observed in the groundwater are the direct result of chemical changes to wetland soils induced by urbanization and development (oxidation of sulfide-bearing wetland soils as water tables are lowered) and reflect travel times through the ridge system measured in decades, not centuries. The potentiometric map presented in Figure 1-7 does not represent all of the data in the FFS and available for use. In particular, the heads for G-101, G-102, and LF7MW01 do not appear to have been considered in construction the map. The head for LW7MW01 is between 20 and 25 feet the below the surface mapped in Figure 1-7. The head for G-l01 is between four and 13 feet below the mapped surface, and head of G-102 is between nine and 14 feet below the mapped surface. If these heads were included in the map on Figure 1-7, the complexity of the head distributions in and around landfills 6 and 7 are far more apparent. Including the head of G-101 alone in the mapped potentiometric surface would clearly demonstrate that the reduced head at GV-1 is far more locally restricted than suggested by the map in Figure 1-7. This in turn suggests that the ability to effectively use the existing, deep storm drain system or the gas vent wells is likely to be far less effective at draining leachate than is implied in the FFS. The relatively low head values shown in the other two excluded wells suggest that the "sink" observed at GV-1 may not necessarily or solely be due to the deep storm drain, but may alternatively or supplementally have a geologic component. Until this possibility is fully evaluated, none of the action alternatives can be evaluated adequately and even such fundamental components as monitoring well locations along probable migration paths cannot be determined The FFS refers in several places to monitoring well LF6MW0I as being an upgradient well with respect to the landfills. While this well is the most headward well in the ravine, the site characterization does not establish this well as upgradient. A comparison of the head level at LF6MW01 and the topography displayed on Figure I-4 suggests that the well may likely be downgradient of the landfills, at least at some times. The land surface at landfill 6 is depicted on Figure 14 as a bowl, with several closed depressions contoured. The elevations of the depressions nearest LF6MW01 are between 664 feet and 658 feet, some S to 14 feet above the head in LF6MW01. During periods of heavy recharge, one would expect mounding of leachate under the depressions, potentially to the land surface, based upon drainage observations discussed in the FFS. At such nines, LF6MW01 would be downgradient, not upgradient of the landfills The hydrogeologic conceptualization of the landfills relies heavily upon permeabilities obtained from slug tests from only three wells at the landfills and two wells elsewhere on Ft. Sheridan. of the three site wells, only one tested in situ soil materials below the ravine and none tested in situ materials adjacent to the ravine. The interpretations of the three site slug tests, obtained from the Remedial Investigation ESE, 1992), show that certainly in one case (LF'7MW04S), and possibly a second case (LF6MW04D), the interpretative model selected is inappropriate based upon the response of the well to the test. The data should be re-evaluated correctly. The interpretation of the 10 gpm low-flow storm drainage as leachate infiltration into the underlying storm drain is unsupported by either chemical or hydrogeologic/geologic data. This interpretation requires 13 inches of some 33 inches of average precipitation infiltrate annually and drain through the landfill into the storm drain. This would have to be in addition to precipitation that must infiltrate to provide leachate that is observed as seepage from the flanks and east face, drainage into shallow storm and surface drains, and flow into surrounding and underlying native soils. The 13 inches stated, let alone the undetermined total infiltration required, stretches credulity. Further, there appears to have been no effort to evaluate even qualitatively the seasonal fluctuations of heads, a direct indication of the infiltration component of water balance. Similarly, the MODFLOW model made no attempt to match any transient behavior of the landfill, an absolutely critical step in establishing the validity of any numerical modeling. The modeling done in support of the 10 gpm interpretation appears at best circular in reasoning and, as described, the structure of the model seems inadequate to actually test the concept. The flow net provided in Figure 1-8 is distinctly in conflict with the parameterization of the "successful" model. The former shows a flow pattern through landfill material and native soils that have little contrast in hydraulic properties (little or no deflection of potentiometric lines or flow lines at the contact between the materials), whereas the model had a 500-fold contrast between the landfill waste and the native soils. However, the model is not documented in the FFS for critical review, and this is a major deficiency for the study. The 10 gpm interpretation is so questionable and so fundamental to the evaluation of the various capping alternatives that it cannot be accepted based upon anything less than full critical review of all supporting evidence, including the modeling that is referenced. Conclusions and Recommendations The data within the FFS is not sufficient to support either the evaluation of the proposed alternatives or the selection of a preferred alternative. The inadequacy is observed among geologic, hydrogeologic, geochemical and chemical data. In addition to the insufficiency of the data within the ITS, conclusions drawn from the included data are frequently not supported by the data and can be at odds with accepted scientific principles. Prior to proceeding with any action alternatives, additional characterization is required for groundwater systems and the geology surrounding the site. Sampling and characterization of the solid waste should he done. Leachate sampling from the landfill mass should be undertaken to supplement the characterization of the fluid within the gas vent wells. Radiologic assessment of the various media at the site should be undertaken. A sampling program to determine the extent of the probable DNAPLs should be undertaken. Finally a realistic integration of the existing and new data should be performed before any evaluation of action alternatives is undertaken. In the interim, leachate escaping from the landfills from either lateral and face seeps or that may occur in storm drain effluent should be captured, treated as necessary, and appropriately discharged, rather than being permitted to dump into Lake Michigan. Also during the interim, gas levels (at least methane and vinyl chloride) in storm drains, utility corridors and residences should be monitored and vented as necessary as an ongoing mitigation program. ====== To: Sierra Club, Illinois Chapter From: Charles H. Norris, Geo-Hydro, Inc. Date: December 17, 1996 Re: Responses to Draft Responsiveness Summary. 17-1 The rationale by the authors of the response justifying the discrepancy between the qualitative descriptions of waste among the alternatives is at best circular. "The differences can be attributed to the differences in risk" is a restatement of the problem, not a resolution of it. Unless and until the waste contents of Landfills 6 and 7 are adequately characterized, it is inappropriate to presume relative risks, or to use alternative-dependent descriptions to bias perceptions of the risk. The acknowledgment by the authors of the response that these wastes must be presumed to include hazardous and/or industrial wastes (response to comment #1-1) does not absolve the government from the responsibility of determining the specific nature of the waste, rather it underscores the need for such determination. Some MSW landfills from this era that were the sites of co-disposal are among the Superifund sites. The authors acknowledge the cost of cap replacement has not been included, because NCP doesn't require it, since it is presumed to last "indefinitely." This is a presumption that is insupportable. The comparison of the proposed cap at this site with that at an off-site facility is deliberate misdirection. Such a site would not likely be in an urban area, would not be adjacent to Lake Michigan, would have hill engineering containment of the waste, and would be a facility acknowledged and sited for its purpose, a hazardous or special waste disposal facility. Further, as pointed out by the authors of the responses, the off-site facility would have the additional advantages of requiring up-front payment, by way of disposal fees, for facility maintenance, and the government would clearly still be liable for the waste and would be unable to "walk" at some arbitrary or budgetary point in the future. Finally, the proffered computation of discounting cap repairs calls to mind Mark Twain s observations regarding the relationship between prevaricators and statisticians. A more meaningful calculation is that a cap repair today that costs $3,000, 000 will cost 7.7 times as much, more than $23,000,000, in 30 years, using the suggested 7% rate for discounting/inflation. 17-2 The authors choose not to respond to the concerns of lake releases in this comment, but rather to emphasize the extortionary nature of the proceedings. Four months have gone by since the observation was made that unabated discharge to the lake should be stopped immediately. The government responds that it doesn't even anticipate stopping the lake dumping until at least a year after the final selection of the capping alternative as "interim" remedy. The longer selection of this alternative takes, the longer the lake dumping will be permitted to continue. Until the 'EPA agrees to their monitoring pro grain, they will not even monitor storm drain discharges. The authors also choose not to respond to the concerns expressed in the comment regarding existing (prior to implementation of the capping alternative) monitoring and remediation programs with respect to migration and buildup of landfill gases into the surrounding environment and structures. The implication is that until they are permitted to make their final selection, they will do nothing. It is also noted that the referenced section of the FFS, Section 1.2.3.8. does not address the concern, it defines the concern. 17-3 The authors apparently have failed to read "the radiological assessment report prepared by the State of Illinois Department of Nuclear Safety that was provided in the FFS as Appendix I," or they would not have misstated that that document concludes "there is no significant hazard resulting from radioactive materials potentially disposed of in Landfill 7." The report went no further than to conclude that there is no current exposure hazard at the site. Any hazard from the site with respect to radioactive waste at the site cannot be determined without characterization of the radioactive waste and any still-existing containment of that waste. The inability of the authors of the responses td understand the inappropriate juxtaposition that was cited, and its potential to mislead, is perhaps one of the more telling examples of the mind set with which they approach this project. While it is enlightening, it is not something that lends itself to resolution or to belaboring. However, other portions of the response do need to be addressed. First, the fact that may MSW landfills of this vintage were co-disposal sites for hazardous wastes does not reduce the concern - many such sites are now Superfund sites. Second, the suggestion that such materials at this site are immobile if present cannot be supported by the existing site data. The existing site data from different media are inconsistent. The best interpretation of the inconsistency is that the leachate sampling protocols for the gas vents are grossly inadequate. Sampling and analysis of the solid waste would assist in resolving the inconsistencies, as well as helping to define source terms for risk analysis and physical properties of the wastes. One of the inconsistencies observed was the detection of vinyl chloride in the effluent gas at the Landfill 7 gas vents, whereas there is no vinyl chloride or precursor compounds (chlorinated solvents, an expected material for this landfill) detected in the fluids collected at the gas vents. Perimeter air samples around the landfill did detect precursor compounds, and one monitoring well did detect vinyl chloride. The authors of the responses rationalize the lack of detection in water at the gas vent wells as being due to detection limits for the water that are too high to detect dissolved vinyl chloride that would be in equilibrium with the gaseous phase, referencing an analytical detection limit in the water of 2.0 mg/L. This is simply false. The reported detection limit in Table 1-7 is 0.010 mg/L, 200-fold lower than the stated number, and the analytical detection limit reported in Table 1.7 for vinyl chloride is 0.0026 mg/L, more than 760-fold lower than the number used in the response. Bottom line - at the reported gas concentrations, vinyl chloride or a predecessor compound should be observed in the liquid. That they aren't is a direct measure of the inadequacy of the leachate sampling protocol and the analyses to date. There is an indirect measure of the understanding by the government of the in adequacy of the existing leachate analyses. The work plan for this fall at the site included sampling the gas vents again, but this time purging the water from the vents first (normal procedure). The probability that this leachate is itself hazardous material is reflected in the fact that the purged liquid is to be dumped back down the vent, rather than properly disposing of it based upon its chemical characteristics. 17-5 In response to this comment, the authors reference the response to comment 1-32. Among the assertions in that response is the statement that "[t]he leachate collection system would intercept and collect DNAPL present. "This is without question an overstatement, if not a known erroneous statement. There is no evidence anywhere in the record that the storm drainage system is positioned or open in areas where DNAPL is or may be within the landfill. It is also important to recognize that the impacts of DNAPL is not primarily on the functionality of the landfill cap, but on the composition and treatability of the landfill leachates and landfill gases. Simple flaring will not destroy some chlorinated compounds. 17-6, 17-15 The authors of the responses maintain that mass balance cannot be used to verify that the low flow of 10 gpm may be leachate(s) from Landfills 6 and 7. In fact, this is one of the few methods available to independently verify the foundation of the house of cards that is based upon the assumption of current leachate drainage of 10 gpm. It is clear from the data provided in the FFS that the composition of the 10 gpm is highly inconsistent with the composition of any of the leachate compositions, but that it is fully consistent with the composition of the water from storm drains upgradient to the landfills. The chemical data overwhelmingly indicate that the water that constitutes the bulk of the low-period flow is not leachate from these landfills, but rather is storm-system drainage and "that unknown buried drains still exist that historically discharged to the ravine, or directly to the storm drain pipe..." (FFS, p24.) In response to the comments regarding the reasonableness of 13 equivalent inches of precipitation, the authors branch into what can best be considered a rambling discussion of ways that that can be accounted for without actually having precipitation infiltration of 13 inches/year. Among the more creative ideas is one that up to 3 gpm is groundwater flowing laterally into the waste, in spite of the fact that the landfills are everywhere in the FFS described as being characterized by outward gradients; i.e., groundwater flow from the landfill into the surrounding soils at rates of from 0.1 to 3 gpm, not inward into the landfill. (The outward gradient is used to justify the 10 feet of over-excavation of native soils in the dig-and-haul alternative.) The added losses to lateral leachate seeps is dismissed as being only that due to evaporation and transpiration. The losses to such mechanisms from these wet, swampy areas (App. I, IDNS) is not a trivial amount and should have been considered in performing a reasonable water balance for the site, and not dismissed out of hand. A check on the infiltration rates can be done by considering the seasonal head variations, as suggested in the comments. In response, rather than performing such a calculation, the authors choose merely to repeat data in the FFS, stating that seasonal head losses in the waste are about 5 feet. If the drainage porosity in the waste is as high as 10%, this seasonal drainage represents only about 6 inches of percolation, not 13 inches. 17-7 The authors of the responses acknowledge that the groundwater in LF6MW04S has been impacted by leachate migration in that area. 17-8 The authors indicate that sufficient information exists to determine whether increases or decreases in gas generation will result from lowering the leachate levels. Any protection that is to be afforded must come from the willingness to add to program costs by changing the sequence of engineering activities in response to unspecified additional, continuing monitoring during the installation of the cap. An example is to install the gas flare sooner than presently planned. [It is not clear what type of flare will destroy vinyl chloride.] The excuse of insufficient information is particularly hollow since characterization of the solid waste would provide the types of information (moisture contents, porosities, drainage behavior, bulk densities, etc.) that would permit the behavior of the gas generation at these landfills to be more reliably projected and the public to be protected proactively and not reactively. 17-9 Whether or not the present site characterization is consistent with a body of regional published data is irrelevant if it does not account for site specific conditions. And, while there may be a plethora of geologic data the authors of the responses choose to interpret in a given manner, the fact is that that data also fits an alternative conceptual model that demonstrates far less isolation of the landfills. The authors err when they opine that the erosional process of sapping does not involve an active groundwater flow system. Such flow systems define the essence of erosion by sapping as opposed to down-cutting. Such a flow system does not require continuous lenses of silt, sand or gravel, nor does the geology of the area support such systems. However, flow Systems with discontinuous lenses or layers of silt, sand, and gravel that are connected by fracture systems provide the necessary flow network to develop the geomorphology of the ravines. The authors are correct when they point out that were the sapping mechanism and the interconnected groundwater flow systems to be present and active, field evidence would include visible seeps and springs (even in the dry season), benching and terracing. It appears that the authors, however, have apparently never been down into the ravines, based upon the assertion that these features are not present. The presence of these features is apparent to anyone who takes the time to climb down and brush away the fallen leaves or push aside the hydrophilic vegetation at the seeps. These readily observed and characteristic features of sapping are why the Illinois Chicago Circle geology department brings field trips to the north shore ravines; they are one of the clearest examples available in the Midwest to demonstrate an erosional process normally associated with canyon-land development in the arid west. 17-10 The authors of the responses are undoubtedly aware that, while the rigorous characterization and evaluation of a dual porosity system would require far more data and different data than has yet been obtained for this site, the approximation suggested in the comment they are addressing is accurate to with a few percent and is all that is available from the data provided in the site documents. They are also unquestionably aware that, while "average or bulk values for the porous medium as a whole are most commonly used [as was done in the FFS characterization] in evaluating groundwater flow rates (Dominico and Schwartz, 1990)," neither Pat Dominico nor Frank Schwartz is advocating using the average bulk porosity of the entire system for computation of contaminant travel times. The three empirical observations called upon by the authors in support of no secondary porosity system are as applicable a system with secondary porosity as to one without. Both the first and second erroneously attribute head distributions solely to variations or contrasts in hydraulic conductivity. In fact, both phenomena discussed represent a balance between hydraulic conductivity and the flux of water through the system. Exactly the same head/gradient distributions exist even if the hydraulic conductivity is doubled, provided it is acknowledged that the mass of water moving through the system is also doubled. The significance to the site and the selected alternative is that the effectiveness of the cap and leachate drain for keeping water out of the waste and leachate production minimized are undermined if there is an unrecognized, contributing network of secondary porosity. Similarly, the now-reported lack of alteration areolae along fractures indicates only that oxygen has not historically moved through the fractures, not that water has and is not Moving through the fractures, As discussed in the original comment, the relatively high sulfate levels in the monitoring wells are suggestive of the oxidation of sulfide-rich soils associated with upland forest wetlands, as likely existed in the area prior to urbanization. Such soils would consume any oxygen in the groundwater before it could affect the tills adjacent to the conduit fractures. Thus, the system described in the comment not only is consistent with the sulfate concentrations observed as background at the site, but also is consistent with the introduced observation of lack of alteration areolae. The importance of the secondary porosity system is particularly significant once the leachate collection has stopped. Without withdrawing leachate from the landfills, they will resaturate and establish a gradient out of the landfill rather than into the landfill. Whether or not the secondary porosity has been yet evaluated with respect to average hydraulic conduc tivities, the travel times will be less through it than through an equivalent single-porosity system. Any attenuative properties of the natural system will also be reduced because of reduced contact and reduced contact time. 17-11, 17-12 The authors of the responses either don't understand the intent and Significance of the comment and the underlying data, or are deliberately attempting to cloud the issue. Wherever a vertical gradient is observed in or under the landfill, the gradient is downward. That is, the phreatic surface is higher than the head associated with the deeper interval. One can, therefore, project that the phreatic head at the G-101 location is as high or higher than the G-101 head, just as the LW6MW04S head is higher than the LW6MW04D head, as seen in Table 1-2. As mapped, the phreatic head at G-101 is interpreted at an elevation of about 642 feet. As pointed out in the response, the head at G-101, and thus the minimum phreatic head at that location, is about 653.5 feet, 10 or more feet higher than mapped. This means that the head gradient toward the manhole at the phreatic surface is significantly steeper than that mapped on Figure 1-7. Since the flux into manhole is whatever it is, the steeper gradient means that the hydraulic conductivity is lower than that implied by the Figure 1-7 interpretation. The manhole may be draining an area around itself but tat area is less than that displayed on Figure 1-7 and that suggests that drainage of both gas and liquid from the waste may be more difficult than inferred from the incorrect mapping. 17-13 The lack of head data for LF6 and the described standing and ponding water make the concern more than theoretical. Further, the distance of 100 feet from the waste boundary is not necessarily significant if there are preferred pathways from the waste to the well. If monitoring points have now been installed as part of the fall program, it may now be possible to determine at least what the dry-season gradient relationship is between leachate levels in LF6 and LF6MW01. 17-14 It is disturbing that, even upon review of the slug tests mentioned in the comment, no alternative interpretation was made. It is clear from the response of the wells to the slug test that the assumptions for the Bouwer and Rice interpretation method are not met by the wells. Were the Bouwer and Rice assumptions appropriate for the wells, the slug test data would have generated a semi-log plot of residual head (log) vs. time (linear) that is a straight line. It is the slope of this straight line that provides the solution in the Bouwer and Rice method. However, the data do not generate this straight line. Rather, they form curvilinear patterns with a continuously decreasing slope with time. Since the measured-data trends do not conform to trends generated by hydrologic systems assumed by the method, the site hydrologic system does not conform to the system assumed by the interpretation method. In the case of LF7MW04S, partial data reported in the groundwater classification document permitted a limited evaluation of the test results by the Cooper, et al. method (confined system). It appears the data conform more closely with the Cooper type curves than to the Bouwer and Rice straight line, with a hydraulic conductivity at least 5-fold greater and a lower storage coefficient. More important, however, than the revised hydraulic conductivity is the data-derived observation that even in the area of the beach and at shallow levels, the flow system appears to be a confined, not unconfined, system. The authors acknowledge that additional hydraulic conductivity determinations are necessary to characterize the soils adjacent to and beneath the landfills. It is unfortunate that it is not recognized that this data is needed before it is appropriate to make the final selection of the alternative for the landfills. 17-16 The limitation to the MODFLOW modeling and the down-grading of its significance in the response is appropriate and welcomed. It is somewhat surprising that the authors to the responses profess to not understand the critical nature of the 10 gpm figure with respect to the consideration of action alternatives, particularly given their emphasis on, and defense of it. The fundamental significance of the l0-gpm-as-leachate assumption is that it provides the foundation upon which the interpreted fluxes through the landfill and surrounding soils are built. 10 gpm is the maximum (and stretched) infiltration rate that can reasonably be postulated for the landfills. If the 10 gpm discharge were leachate, there could be very little left That is escaping through the flanks or sides of the ravine. Since the gradients are relatively high, that dictates a very low hydraulic conductivity, without the need for testing. If the hydraulic conductivity is very low, there will be very little lateral flow into the landfill once leachate levels are reduced, and leachate levels are kept low with low leachate pumping and treatment costs. Once active management ceases, the landfill will be slow to recharge and whatever the final head configuration, water fluxes and corresponding risks will be minimal. Correspondingly, if the 10 gpm discharge is not leachate, none of the other flux terms for the landfills are known. Lateral and downward flow could be significantly higher with correspondingly higher hydraulic conductivities. It would be more difficult to lower leachate levels and require more pumping and treatment to keep them low. Once management ceases, the drained wastes will recharge more quickly and outward fluxes will be greater. The authors are correct in observing that any differences in the flow system will affect all action alternatives. However, it will not affect them all equally. A major difference between the capping and the excavation alternatives is that the former is forever, the waste is always in the path of groundwater flow toward and through Wells Ravine. -- Evan Craig