Saratoga County Landfill - Supplemental Decision, September 22, 1998
Supplemental Decision, September 22, 1998
STATE OF NEW YORK : DEPARTMENT OF ENVIRONMENTAL CONSERVATION
In the Matter
- of -
the Application for a permit to construct and to operate a solid waste management facility pursuant to Environmental Conservation Law (ECL) Article 27, Title 6 of the Official Compilation of Codes, Rules and Regulations of the State of New York (6 NYCRR), Part 360
- by -
SARATOGA COUNTY LANDFILL
Project Application No. 5-4146-00018/00002-1
SUPPLEMENTAL DECISION OF THE DEPUTY COMMISSIONER
This Decision supplements a prior decision of September 3, 1996, by Deputy Commissioner David Sterman, who has since left the Department. Because Commissioner Cahill served as the Department's General Counsel during the time that the initial adjudicatory hearing took place in this matter, responsibility to make this Supplemental Decision has been delegated to me as the Deputy Commissioner for Natural Resources.
Saratoga County ("County" or "Applicant") seeks a Part 360 permit to construct and operate a landfill to receive municipal solid waste. The proposed facility would be located at the eastern end of Kobor Road in the Town of Northumberland ("Town"). The County is the lead agency under the State Environmental Quality Review Act ("SEQRA"), and has issued a final environmental impact statement.
After an adjudicatory hearing on five permit-related issues, the Department issued a Decision dated September 3, 1996, which directed that a landfill permit be issued to the County. That Decision was subsequently challenged by the Town, an intervening party that had opposed the project at the hearing. On January 8, 1998, the Appellate Division of State Supreme Court, Third Judicial Department, annulled the determination to issue the permit and remitted the matter to the Department for readjudication of the issue of soil permeability and its impact on the County's entitlement to a variance; the Court further stated that the Town should be afforded a reasonable opportunity to conduct soil testing at the landfill site. (Town of Northumberland v. Sterman, et al., AD 2d, 667 NYS 2d 505 (3rd Dept., 1998)). Soil permeability is relevant to whether the County should receive a variance from a Part 360 construction requirement that a minimum separation of five feet be maintained between the base of the landfill's constructed liner system and the seasonal high groundwater table.
The Town performed its testing this past spring and the hearing in this matter reconvened for nine days between June 22 and July 2, 1998. The proceedings since the court decision remanding this matter are summarized in the attached Supplemental Hearing Report of Administrative Law Judge ("ALJ") Edward Buhrmaster. That report, including its Conclusion and Recommendation, is adopted as my Decision in this matter, subject to my comments below.
Environmental Impact of Granting Variance
The recent hearing in this matter was concerned with whether granting the groundwater separation variance would have a significant adverse impact on the environment, and the County's entitlement to a variance. Soil permeability is relevant to this issue since it concerns the ability of contaminants to migrate from the landfill in the event of leakage or containment failure.
Overall, I agree with the ALJ that the Town's testing for soil permeability was flawed both in how the testing was performed and how the test data were analyzed. I also find that the overall horizontal and permeability values of the County, which are part of the application and were examined at the 1996 hearing, are much more accurate and representative of the known soil characteristics of the site. In fact, aspects of the Town's testing confirm that the soils beneath the landfill are tight, and would resist the rapid spread of contaminants to locations where they might have significant impacts. I also note that the overall landfill design incorporates a pore pressure relief system to prevent uplift and possible tearing of the liner, which provides reasonable assurance that the design is safe and appropriate at this location. There is no basis for the Town's contention that the soils beneath the landfill site are permeable, or that the landfill system is inadequately designed.
I hereby reaffirm the Department's prior decision to grant the groundwater separation variance. I further reaffirm that the ALJ's Report and Supplemental Report, taken in conjunction with the entire hearing record and the Final Environmental Impact Statement ("FEIS"), prepared by the County as lead agency, affords an adequate basis for my finding that the requirements of SEQRA contained in Environmental Conservation Law ("ECL") Section 8-0109 and Title 6 of the Official Compilation of Codes, Rules and Regulations of the State of New York ("6 NYCRR") Part 617 have been met and that pursuant to ECL Section 8-0109(8) and 6 NYCRR Part 617, consistent with social, economic and other essential considerations, including reasonable alternatives, the Applicant will minimize or avoid to the maximum extent practicable any significant adverse environmental impacts.
The Department Staff is directed to reissue the prior permit to construct and operate the landfill.
Peter S. Duncan
Albany, New York
Date: September 22, 1998
Background and Brief Project Description
Saratoga County ("the County" or "the Applicant") proposes to construct and operate a 23-acre, double-lined municipal solid waste landfill on part of a 350-acre tract at the eastern end of Kobor Road in the Town of Northumberland, Saratoga County. The landfill would be constructed in three phases. The first phase would involve about nine acres of landfill area providing 467,000 cubic yards of capacity. Total landfill capacity for the three phases would be 1,900,000 cubic yards. The project - - including the landfill footprint, access road, leachate storage tanks and buffer areas - - would encompass about 130 acres when fully developed.
To move ahead with the project, the County requested a permit to construct and operate a solid waste management facility. A permit application was made to the Department of Environmental Conservation ("the Department" or "DEC") and was reviewed pursuant to Title 7 of Article 27 of the Environmental Conservation Law ("ECL") and Part 360 of Title 6 of the Official Compilation of Codes, Rules and Regulations of the State of New York ("6 NYCRR Part 360"). After the application was determined complete, it was the subject of a series of hearings including a 28-day adjudicatory hearing in 1996 on five permit-related issues.
One of the adjudicated issues was whether the County should receive a variance from a construction requirement that a minimum separation of five feet be maintained between the base of the landfill's constructed liner system and the seasonal high groundwater table [6 NYCRR 360-2.13(d)]. As the presiding Administrative Law Judge ("ALJ"), I recommended that the variance be granted. This recommendation was adopted in DEC Deputy Commissioner David Sterman's decision, dated September 3, 1996, which directed that the landfill permit be issued.
This final decision was successfully challenged by the Town of Northumberland, an intervenor in the adjudicatory hearing. On January 8, 1998, the Appellate Division of State Supreme Court, Third Judicial Department, annulled the decision and remitted the matter to DEC to readjudicate issues of soil permeability relevant to the groundwater separation variance. (Town of Northumberland v. Sterman, et al., AD 2d, 667 NYS 2d 505 (3rd Dept., 1998))
The environmental impact of issuing this variance was first identified as a hearing issue in May of 1995, during an issues conference I held. The County had determined that the five-foot separation between the base of the landfill's liner system and the seasonal high groundwater table could not be maintained due to the presence of low-permeability soils which resulted in a high water table at the landfill site. According to the County and Department Staff, granting the variance would not have a significant adverse impact on the environment. The Town disagreed, claiming there would be a considerable risk of groundwater contamination.
In identifying the issue for adjudication, I granted the Town intervenor status and authorized it to conduct pump tests and dig test pits on the landfill site according to a work plan the Town was to furnish to the County. However, the plan was not developed - - and the testing did not take place - - because my ruling allowing testing was reversed by Deputy Commissioner Sterman in an interim decision dated July 14, 1995. Acting on an appeal by the County, which was joined by Department Staff, Sterman denied the Town site access on the basis that its recommended testing was not necessary to the issue's adjudication. Sterman said adjudication of the issue should focus on the County's soil permeability measurements, since the Town had raised some doubt about their reliability.
The Appellate Division directed that the groundwater separation variance be reconsidered after the Town was afforded a "reasonable opportunity" to conduct on-site testing, in essence reversing Sterman's prior determination that such testing was not necessary. According to the Court's decision, "Inasmuch as there was every indication that the proposed testing would have produced relevant, probative evidence bearing on an adjudicable issue, and the Commissioner posited no reasonable basis for denying the Town the brief site access it sought for that purpose, his determination cannot be said to have a sound basis in reason (see, Matter of County of Monroe v. Kaladjian, 83 NY2d 185, 189). Accordingly, it must be annulled, along with the decision granting the groundwater separation variance (and the permit, which depended thereon), and the matter remitted for redetermination of the groundwater safety issue after the Town is afforded access to the site to perform testing." [Id. at 667 NYS 2d at 507.]
On January 15, 1998, the County submitted a letter requesting that the Department's Office of Hearings and Mediation Services establish a schedule relating to additional consideration of the soil permeability issue. On February 4, after discussions with the parties, I issued rulings establishing the framework under which the Town's testing would occur and the hearing would resume.
Consistent with my deadline, the Town submitted a work plan for its site investigation on April 1. The plan identified the types and possible numbers of intended tests, as well as testing protocol and equipment. It confirmed that the County and DEC Staff would be permitted to observe all testing being performed and would be given split samples of any samples taken.
The County and Department Staff submitted letters commenting on the work plan; however, the plan was not subject to either of their approvals. While both the County and Staff noted that the scope of testing identified in the plan went beyond what the Town had proposed initially, there was no objection to any of the proposed tests provided that they were performed during the window of time reserved by the ALJ.
The Town wanted to assure that testing was conducted when the water table was high, the frost had left the ground, and the soil was firm enough to support the heavy equipment it would be using. The Town proposed to reserve the period from April 27 to May 15 for its testing, and these dates were adopted in my scheduling ruling.
The County indicated that, to the extent the Town wanted to install or develop its own on-site wells for pump testing, excavation could occur before April 27, subject to mutually acceptable arrangements.
The Town commenced its testing program on April 27 on the landfill site. It excavated 17 test pits, tested samples of the excavated soil, and installed 11 wells in which it then conducted slug and pump tests.
Consistent with the timetable established by my rulings, the Town prefiled its two witnesses' direct testimony on June 17, 1998. I did not order that the County or Department Staff prefile testimony since they indicated they would not be doing any field work of their own, and if they presented any witnesses, their testimony would be rebuttal in nature.
The Adjudicatory Hearing
The adjudicatory hearing reconvened on June 22, 1998, and continued on consecutive weekdays through July 2, 1998, the nine days reserved in my scheduling ruling.
The permit applicant, Saratoga County, was represented by Louis A. Alexander and Kevin M. Bernstein, Esqs., of Bond, Schoeneck and King, LLP, in Albany.
DEC Staff was represented by Steven L. Brewer, Esq., of DEC's Region 5 office in Ray Brook.
The Town of Northumberland was represented by Edward Lindner, Esq., of Saratoga Springs, and Laura Zeisel, Esq., of New Paltz.
The following witnesses testified for the parties:
For the Town, Dr. Thomas F. Zimmie, a professor of civil engineering at Rensselaer Polytechnic Institute, and owner and president of Civrotech Engineering, P.C., a consulting firm in Troy; and Dr. Kevin E. Brewer, a hydrogeologist now living in South Carolina and employed as a fellow scientist by Bechtel Savannah River, Inc. As consultants to the Town, Dr. Zimmie designed and oversaw the Town's testing of the landfill site, and Dr. Brewer analyzed data from pump tests conducted as part of the Town's site investigation, in addition to suggesting the pump test design.
For the County, Dr. Donald I. Siegel, an independent hydrogeological consultant who is also an earth sciences professor at Syracuse University.
For DEC Staff, Dale A. Becker, an engineering geologist in DEC's Ray Brook office, and Robert Phaneuf, a bureau chief and engineer in the Department's Division of Solid and Hazardous Materials.
Except for Mr. Phaneuf, all the witnesses had testified on the groundwater separation variance at the previous adjudicatory hearing in 1996. The record in this matter consists of the testimony heard and exhibits received during the 1996 hearing as supplemented by the testimony and exhibits from this nine-day continuation.
Because the matter was remanded so the Town could augment the record with the results of its own site testing, the Town had the burden of going forward when the hearing reconvened. However, the burden of proof on the variance issue has remained with the County as permit applicant, consistent with 6 NYCRR 624.9(b)(1).
When the hearing reconvened, each party was provided an opportunity for an oral opening statement. At the conclusion of the hearing, each party was allowed to submit a closing brief. Timely closing briefs were received from all parties, and the hearing record officially closed on August 3, 1998.
The parties were also allowed an opportunity to propose corrections to the transcript. Each party submitted proposed corrections, as did I. There were no objections to any of these corrections, and therefore I adopt them. The corrections are noted in pen in the original of each day's transcript.
Should the County receive a variance from a requirement [at 6 NYCRR 360-2.13(d)] that a minimum separation of five feet be maintained between the base of the landfill's constructed liner system and the seasonal high groundwater table?
"Every application for a variance must demonstrate that the proposed activity will have no significant adverse impact on the public health, safety, or welfare, the environment or natural resources and will be consistent with the provisions of the ECL and the performance expected from the application of 6 NYCRR Part 360." [6 NYCRR 360-1.7(c)(2)(iii)].
POSITIONS OF THE PARTIES
Position of the Town
Construction and operation of the landfill would have a significant adverse impact on the environment. The Town's slug and pump tests verify that the overall horizontal permeability of the soils at the site is approximately 1 x 10-4 centimeters per second (cm/sec). Because of this, contaminants leaking from the landfill could move offsite in a matter of months, and could travel to nearby ravines and into the Snook Kill in less than a year.
In addition, the vertical hydraulic conductivity at the proposed site is on the order of 10-5 cm/sec, apparently due to cracks in the site's brown clays. The Town's pump tests establish that contaminants leaking from the bottom of the landfill could migrate downward into bedrock in a matter of weeks, making site remediation very difficult.
The findings of fact from the previous hearing require substantial revision, and the variance, as well as the permit that depends on it, should be denied.
Position of the County and Department Staff
Construction and operation of the landfill would not have a significant adverse impact on the environment. The County's landfill design meets or exceeds the minimum design requirements of Part 360. The textural and permeability characteristics of the site soils are consistent with the intent of the regulations to locate landfills in low permeability, clay- and silt-rich soils. Although the landfill has an "in-gradient" design, meaning that its base would extend below the groundwater table, it has adequate engineering features, including a pore pressure relief system, to ensure effective environmental protection.
The County's measurements of soil permeability are scientifically reliable, while the Town's are not credible due to errors in testing procedure and data analysis. Deficiencies in well construction compromised both the pump and slug tests conducted by the Town. Data from the Town's test pits merely confirm what was previously known about the upper soil profile, and the Town's admittedly selective soil sampling techniques resulted in laboratory findings that are not representative of the soils at the test pit locations.
Vertical seams at or below the projected bottom of the landfill do not pose a threat to landfill safety, since they are limited in extent and contain low-permeability silt. Compacting and proof-rolling the landfill subgrade, as required by DEC regulation, will further impede the possible escape of contaminants. The Town has exaggerated how fast contaminants would travel from the site and down to the underlying bedrock, based on inaccurate assumptions about the soils' permeability.
The findings of fact from the previous hearing should be reconfirmed, and the variance should be granted as a basis for allowing the landfill permit to be reissued.
The record in this matter was reopened so the Town could introduce evidence from the testing it performed in late April and early May at the landfill site. This evidence was received against the backdrop of a record that had been developed in the 1996 adjudicatory hearing. As a result of that previous hearing, I prepared a report addressing the groundwater separation variance, one of five issues on the table at that time. The report is attached to Deputy Commissioner Sterman's September 3, 1996, final decision. The portion of the report addressing the groundwater separation variance (identified as Issue No. 4) begins on page 78. On pages 80 through 85 are 28 numbered Findings of Fact addressing the variance, and a discussion of the issue is contained on pages 88 through 107.
As indicated on page 107 of my prior report, I recommended that the variance from the groundwater separation requirement be granted. This was based on my conclusion that, despite not maintaining the five-foot groundwater separation, the landfill would not have a significant adverse impact on the public health, safety, or welfare, the environment or natural resources. That conclusion has not changed in light of the Town's testing, and I still recommend that the variance be granted. As before, my determination is based on the design of the landfill, the characteristics of the surrounding soil, the soil's low overall permeability, and the distances contaminants would need to travel to have significant adverse impacts.
The Town's new evidence primarily concerns soil characteristics and permeability, issues that were litigated extensively in the previous hearing based on information contained in the County's application. Now the record also includes results of the Town's site investigation which was done this past spring.
As a result of the Appellate Division decision, the Town was given a full opportunity to examine and test the soils at the landfill site. The testing occurred during the window of time requested by the Town, and was not impeded in any way by the County or DEC Staff.
The Town's site work included:
- The digging of 17 test pits in the area of the landfill footprint and the laboratory analysis of samples of the excavated soil;
- The drilling of 11 boreholes along the footprint perimeter, in which wells were then constructed; and
- The performance of slug and pump tests in the completed wells, to determine the overall permeability (or hydraulic conductivity) of the glaciolacustrine deposit in which the landfill would be situated.
The parties were divided about the results of the Town's testing. According to the Town, the testing confirmed its long-held position that the site is unsafe for a landfill. Dr. Zimmie said the test pits and lab tests confirmed the presence of significant amounts of silt and sand in the generally clayey soil. He said the slug and pump tests established that the soil is much more permeable than the County had determined.
Given his values for soil permeability, Dr. Zimmie said that contaminants leaving the landfill in the event of a containment failure could travel offsite laterally into ravines northwest of the site in less than one year, and from there to the Hudson River via the Snook Kill in a matter of hours or days. He said that contaminants could also move vertically into bedrock below the site in as little as seven weeks and certainly in less than a year.
The County and DEC Staff disputed the Town's claims of environmental risk, arguing that the Town's testing procedures failed to follow standard scientific and engineering requirements, resulting in unreliable data and incorrect conclusions. According to the County and DEC Staff, the Town improperly exaggerated contaminant travel times to suggest an environmental threat that does not actually exist. The County and DEC Staff argued that, in certain respects, the Town's testing actually confirmed their own characterization of the soils as rich with clay and silt, with negligible amounts of sand and no extensive pathways through which contaminants could move.
I agree with the County and DEC Staff that the Town's testing was riddled with errors such that its estimates of horizontal and vertical permeability cannot be considered reliable. Overall, I am persuaded by criticisms of the Town's investigation by Dr. Siegel, the County's expert, and Mr. Becker, DEC Staff's hydrogeologist.
As in the prior hearing, Mr. Becker's testimony as a disinterested Staff reviewer was particularly important given the differences of opinion between the County's and the Town's experts. An engineering geologist with DEC since 1989, Mr. Becker has reviewed at least seven permit applications for double-lined municipal solid waste and industrial landfills, including this one. He has a Master of Arts degree in geological sciences, and his graduate study included concentrations in glacial and engineering geology.
Mr. Becker was at the site for 11 days during the Town's testing period, observing almost all of the field work and taking thorough notes. His testimony was dispassionate, considered, and extensively supplemented with scientific and technical references.
On issues of well construction, soil classification and slug test analysis, Mr. Becker appeared especially well-versed.
Overall, the County and DEC Staff successfully demonstrated that the well construction techniques employed by the Town were flawed and likely contributed to results overestimating both the horizontal and vertical permeability of the soils. These flaws included inappropriately long screens which were not restricted to the key soil interval in the ten feet below the landfill bottom, extension of the monitoring wells' sand packs into the most weathered and permeable portion of the soil closest to the site surface, and failure to ensure adequate seals at the bottoms of the wells to prevent hydraulic communication with a relatively permeable interface zone between the varved clays and bedrock.
Flaws in well construction would have biased both the slug tests (to determine horizontal permeability) and the pump tests (to determine horizontal and vertical permeability) that were conducted in the Town's wells. However, there were other flaws in how data from the slug and pump tests were analyzed. More specifically, the Town's slug test analysis failed to account for the double straight line effect that most of its data exhibited, and therefore overestimated the horizontal permeability of the soil. Also, the pump tests did not run long enough to establish aquifer characteristics and comply with the chosen method of analysis. The wells were pumped dry before valid drawdown curves could be established, which apart from affecting analysis of the pump tests, confirmed the soils' relative "tightness" and the absence of extensive lateral water-bearing seams.
As DEC Staff's witness, Mr. Becker, convincingly demonstrated, the Town's biased sampling of material from the test pits overstated the presence of sand in the soils closest to the surface, as did the soil classifications assigned by the testing lab. According to the credible observations of both Mr. Becker and Dr. Siegel, the test pits merely confirmed what was previously known about the near-surface deposits and shallow stratigraphy of the site.
To put the Town's study results in a proper perspective, one needs to consider how its testing was done and how its results were analyzed.
As Dr. Zimmie explained in his prefiled testimony, the Town drilled 11 boreholes around the site perimeter in the general area where slug tests were conducted on behalf of the County. Since groundwater flows in a northerly direction under most of the landfill site, three well clusters were located along the northern border of the proposed footprint: No. 1, located off the north-central area of the footprint, near the intersection of proposed landfill phases 1 and 3; No. 2, located off the northwest corner of the landfill footprint, in the vicinity of County well No. 1; and No. 3, located off the northeast corner of the landfill footprint, in the vicinity of County well MW-H. Well Cluster No. 1 consisted of three wells in a straight line: PW-1, OB-1 and OB-2. Well Cluster No. 2 consisted of wells PW-2, OB-3 and OB-4, and well cluster No. 3 consisted of wells PW-3, OB-5 and OB-6.
In addition to the three well clusters along the northern edge of the landfill footprint, two wells were drilled on the southern border of the landfill footprint. MW-1 was located just off the southern boundary of the footprint, to the east of County well MW-E. MW-2 was located just off the southern boundary of the landfill footprint, southeast of County well MW-C.
- - Excessive Sand Packing
The Town used relatively long screens in its wells, from 15 to 35 feet in length, to measure the overall permeability of the entire soil profile beneath the proposed landfill. However, because of excessively long sand packs above the screened intervals, the area measured actually extended into a weathered upper zone just below the ground surface.
As Dr. Zimmie acknowledged under cross-examination, the tops of the sand packs in the Town's wells were all close to the ground surface, from 3.5 feet below the surface (in MW-2) to 7.5 feet (in PW-2, OB-3 and OB-4). The sand packs around the well intakes were much more permeable than the soil surrounding the wells, allowing water to move through them. In effect, the sand packs lengthened the well screens and extended the area being tested into the permeable upper weathered zone, as noted by Dr. Siegel and Mr. Becker. By taking the sand packs into account, about 6 to 9 feet were added to the well screens for each monitoring well, with the exception of MW-2, in which about 3 feet were added.
Extending the sand pack of its wells into the upper weathered zone seriously compromised and biased the Town's soil permeability measurements. According to Dr. Siegel and Mr. Becker, this zone has a higher permeability than the deeper clay deposits due to a higher concentration of silt, the impacts of frost and past farming practices, and naturally occurring features such as calcite concretions, root traces, and desiccation cracks.
Because the upper weathered zone is more permeable than the underlying soil, permeability tests that intercepted this zone would reflect its influence rather than that of the deeper deposits. As Dr. Siegel noted, the depth of the upper weathered zone varies. Although the upper seven and a half feet would be the most weathered, significant impacts of weathering are felt to lessening degrees as deep as fifteen feet from the ground surface, throughout the upper soil profile which, it bears emphasis, would be removed during landfill construction.
The extension of the sand pack into the upper weathered zone was a serious error. As Dr. Siegel testified, if one wants to install a monitoring or pumping well that isolates portions of a soil column, then the appropriate practice is to not extend the sand pack very far above the screen. Dr. Siegel said it is standard practice to isolate the zone one is trying to analyze from other sections, and that the extensive use of the sand pack precluded that isolation necessary for successful testing of the zone of interest to the Town.
To buttress his testimony, Dr. Siegel cited American Society for Testing and Materials (ASTM) Standard D-5092 (reapproved in 1995) entitled "Standard Practice for Design and Installation of Ground Water Monitoring Wells In Aquifers." This standard states that where there is a hydraulic connection between the zone to be monitored and the overlying strata (as there is here, between the deeper varved clays and the upper weathered zone) the upward extension of the sand (or filter) pack from the well screen into the overlying strata should be gauged to prevent seepage into the pack from the overlying units. [Exhibit 395, Section 8.4.1., p.82.]
The Town tried to counter the concern about influence from the upper weathered zone by claiming the soils around the sand pack were coated with a "well skin" as a result of drilling, and that since well development was limited to the screened interval, the portions of the bore holes above and below the well screen retained this skin and therefore did not influence the test results. However, as Mr. Becker testified, the Town's method of open hole drilling and the use of large amounts of water to flush the boreholes during and after drilling combined to partially develop the holes before the well screens and sandpacks were installed. Dr. Siegel agreed that the drilling procedure employed by the Town minimized any skin effect.
On behalf of the Town, Dr. Brewer argued that certain of the County's slug test results at the soil interval 10 to 12 feet below the ground surface (as noted in Table 7-1D of the permit application) demonstrated relatively low permeability, thereby disproving the County's and Staff's "weathered zone" theory. However, Dr. Siegel then referenced other tests from the same table which showed significantly higher permeabilities in the interval of 5 to 15 feet below the surface. Mr. Becker testified that the part of the soil profile most affected by weathering is the top 10 feet, which is consistent with the County's slug test data.
At any rate, as Dr. Siegel explained, the County did not do much permeability testing in the weathered upper soil zone because that zone is above the landfill subgrade and therefore would be removed during the facility's construction. While the upper weathered zone would intersect the side of the landfill liner, Mr. Becker explained that escape of leachate from the side of the landfill is not a concern because there is no hydraulic head on the side wall. In other words, even if there were a side leak, there would be no head or pressure forcing contaminants out through the side. As Mr. Becker emphasized, the landfill bottom is the area of concern in terms of any possible leakage.
- - Inadequate Well Bottom Seals
In addition to the sand pack opening a connection in the wells to the upper weathered zone, thin bentonite seals at the bottom of its deepest wells may also have affected the Town's permeability measurements. Bentonite is a clay material that expands as it is hydrated. The Town sealed well bottoms by dropping bentonite pellets into the wells. Upon contact with water, such pellets begin to swell, and the swelling continues as they go down.
Referring to well sketches he developed based on the Town's information, Mr. Becker testified that in MW-1 there was only eight tenths of a foot of bentonite separating the bottom of the sand pack and the glacial till. In MW-2 the seal was also eight tenths of a foot thick; in PW-2, three tenths of a foot; and PW-3, one foot. (In PW-1, the top of the till was actually above the bentonite seal.)
These measurements may have been off by at least several tenths of a foot, according to Mr. Becker, given difficulties even the best driller would have in gauging depths in wells as deep as those constructed by the Town. Even accounting for these difficulties, Mr. Becker testified that the Town's bentonite seals were of "marginal" thickness. He contrasted them to the three-foot bentonite seals that Part 360 requires toward the top of a well (above the sand pack) to prevent surface water from migrating down into the sides of the casing.
Mr. Becker said that the thin seals employed by the Town might have allowed hydraulic communication between the clay deposits and the more permeable interface zone characterized by till and fractured bedrock. As in the case of the upper weathered zone, any hydraulic communication in the wells with the higher-permeability interface zone would have biased the Town's test results, suggesting the clays beneath the bottom of the landfill liner are more permeable than they actually are.
The comparatively high permeability of the interface was confirmed by County slug testing in that zone, and the roughly one order of magnitude difference in the measured mean permeability between the Town's slug tests in PW-1 (where the sand pack was connected to the till) and tests in adjacent observation wells OB-1 and OB-2 would suggest that PW-1's intersection with the till unit resulted in a higher value at that location, as Mr. Becker explained.
The wells with bentonite bottom seals were in the range of between 20 and 50 feet deep. Speaking from his extensive background in well installation, Dr. Siegel said that if one is dealing with a well that is more than 12 feet deep, it is difficult to drop bentonite pellets down the well and expect that they will place themselves appropriately to seal off the annular space at the bottom. This, he said, is because the pellets begin to hydrate and expand as soon as they touch water, creating the risk of their bridging above the intended level as they go down the well. Dr. Siegel said that in deep wells one normally does not use bentonite pellets; instead one uses a bentonite slurry that is pumped along a tremie line (or hose) to where you want the seal to occur.
Dr. Siegel also stressed that while the bentonite begins to swell upon contact with water, it takes some time for the hydration to become complete. Dr. Zimmie said that in some cases the Town waited "maybe 15, 20 minutes" between placing the bentonite seal and then putting sand on top of it. Although he was not sure, Dr. Siegel said he did not think this would leave enough time for the pellets to fully hydrate.
Finally, Dr. Siegel noted how in constructing the wells, the Town first placed the bentonite at the bottom of its drilled holes, then installed its sand packs, and lastly put its well casings in. If bentonite had bridged in a hole before the casing was set, the placement of the casing and well screen could jar the bentonite loose, sending it to the bottom, where it would form a composite with the sand. Because this composite would lack integrity, it could result in hydraulic communication between the interface and the well screen, as Dr. Siegel explained.
Responding to the other parties' criticisms, the Town returned Dr. Zimmie to the stand. He testified that all five of the bedrock wells were sealed with at least a foot of bentonite, despite contrary information in the Town's own drilling logs, the reliability of which he did not contest. Dr. Zimmie also said that if the bentonite had bridged rather than form a proper seal, one would expect to have seen permeability measurements in the Town's wells in the region of 10-3 cm/sec. However, the highest permeability recorded for the interface zone in the County's application is 3 x 10-4 cm/sec, suggesting that Dr. Zimmie was overstating his case.
Whether or to what extent the bentonite seals of the Town's well bottoms were adequate is more difficult to determine than whether the sand packs at the tops of the wells were excessively long. The upward limit of the sand pack is more easily determined than the depth or integrity of a seal in the bottom of a well as deep as 50 feet.
The Town emphasized that the County and DEC Staff had no direct evidence of bentonite bridging or inadequate bottom sealing. However, under the circumstances, such things could be demonstrated only inferentially. It appeared unlikely that all of the seals had failed; however, the evidence was adequate to suggest that some failures were likely. Of course, as noted above, any failures would cast doubt on the results of permeability tests conducted in the affected wells.
- - Inappropriate Screen Lengths
Finally, in a discussion of well construction issues, one must also consider the very long screen lengths employed by the Town. In the PW-1 and PW-3 well clusters the Town used 25-foot screen lengths; in the PW-2 cluster, the screen length was 35 feet. At MW-1 the screen length was 35 feet and at MW-2 the screen length was 15 feet. However, as discussed above, screen length alone underrepresents the area that is evaluated, since the effective length of a screen also includes the additional length of the sand pack.
Dr. Zimmie claimed that the Town's intent was to screen the length of the varved clay deposit as much as it could, to give a more representative value for site permeability than the County calculated as part of its application. As Mr. Becker explained, this contrasted with the County's intent to screen wells in the clay unit over the comparatively narrow 10-foot interval that corresponds with the soil interval directly beneath the bottom of the landfill liner.
Mr. Becker said that in conducting its investigation, the County used screen lengths to evaluate that specific interval in accordance with a work plan reviewed and approved by DEC. He added that screening the 10 feet beneath the landfill liner avoids any of the near-surface features that could influence permeability tests, and allows one to look at the most critical horizons immediately below the liner, where contaminants would first escape into the environment. Under cross-examination, even Dr. Zimmie agreed generally with the statement that for purposes of evaluating potential contaminant transport, the permeabilities at or below the depth of the landfill liner are the most relevant.
Town's Slug Tests
Like the County did as part of its application, the Town performed slug tests to calculate the horizontal permeability of the soils at the landfill site. These tests, conducted under Dr. Zimmie's supervision by Continental Placer, Inc., were performed in each of the wells constructed by the Town, and the results are summarized in Attachment No. 10 to Dr. Zimmie's prefiled testimony (Exhibit No. 328).
Based on its slug testing, the Town figured that the horizontal permeability of the site soils is 1.3 x 10-4 cm/sec. This contrasts with the County's determination, also based on slug tests, that the clay unit of the landfill footprint and its associated downgradient area has a horizontal permeability of 4.8 x 10-6 cm/sec. Both the Town's and the County's estimates are based on a geometric mean averaging results from different data points.
Slug tests are performed by inserting a solid cylinder (in the Town's case, a metal rod) of a known volume into a hole or well, and then removing it. Both the insertion and subsequent removal of this "slug" displaces the water level in the well, and the rate of response as the water returns to its original level is measured throughout the test with a frequency sufficient to define a water level response curve. When the slug is inserted into the well, it is often referred to as a falling head test; when the slug is removed, it is often referred to as a rising head (or "bail") test.
In well cluster No. 1 (PW-1, OB-1 and OB-2) the Town performed three sets of falling and rising head tests. Determining that the consistency of the test results was "quite good," the Town did only two sets of tests in the remaining eight wells.
Because the Town performed its slug tests in wells that it had constructed, flaws in the construction would affect the test results, as Dr. Siegel and Mr. Becker explained. Dr. Siegel cited the example of a falling head test in which the "slug" is pushed down into the well, forcing the water level higher. Under normal conditions, he said, water would then drain through the well screen until the initial water table condition was achieved. However, assuming the screened soils are tight and the overlying unit is much more permeable, Dr. Siegel said backed-up water in the well would tend to move out to the more permeable zone on top. In such a circumstance, Dr. Siegel explained, the Town would be measuring the permeability of the upper weathered zone, and not the deeper, screened clays. Mr. Becker agreed that the slug test results were biased by the long upward extension of the wells' gravel packs.
- - Slug Test Analysis/Double Straight Line Effect
Not only were the Town's slug tests affected by well construction problems, they were biased by problems growing out of data analysis, as the County and DEC Staff both demonstrated. The Town's slug test data were analyzed by Continental Placer using two different software packages employing what is known as the Hvorslev method: a proprietary, not readily available package called "Hvor Program" developed by Dunn Geoscience and regularly used by Continental Placer for slug test analyses, and a commercial software product called "Super Slug." Finally, as a cross-check against the Hvorslev analysis, the data were analyzed using Super Slug and other methods known as Bouwer & Rice and Ferris Knowles.
As Dr. Siegel and Mr. Becker both explained, the Town's analyses did not account for the double straight line effect and therefore overestimated soil permeability. Basically, this effect is evident whenever the semi-log plot of the response curve of a slug test has a concave-upward curvature, rather than the straight line plot one expects if the test has been conducted properly. The early part of the concave-upward plot is a relatively steep curve, whereas the latter part of the curve is much shallower. If the plotted data exhibit this concave-upward curvature, as the Town's did, analysis of that data depends on how one fits a straight line through the plot to calculate a lag time factor which is then used to solve the permeability equation.
Mr. Becker described the double straight line effect using excerpts from a book by James Butler (Exhibit No. 377) and an article by Herman Bouwer, one of the developers of the Bouwer & Rice method (Ex. 356). In his book Dr. Butler emphasizes the importance of fitting a straight line to the semi-log plot of the slug test data. According to Dr. Butler, the Hvorslev method assumes that a plot of the logarithm of the normalized response data versus time will be linear. However, he also notes that it is not uncommon for storage mechanisms to affect the response data, in which case the data as plotted in a semi-log format will display a distinct concave-upward curvature.
As Dr. Butler explains, when a graph of the response data exhibits such a curvature, there may be considerable uncertainty about how to fit a straight line to the plotted data. He then indicates that the best approach is to fit a straight line to the data representing the shallow, latter part of the curve.
Butler's approach is consistent with that of Bouwer, who attributes the steep, initial part of the curve to the highly permeable sand pack which quickly sends water into a well immediately after the water level is lowered. Bouwer writes that when the water level in the permeable zone around the well has drained to the water level in the well itself, the flow into the well slows down and the plotted points begin to form the second, less steep, straight line which is more indicative of the flow from the undisturbed aquifer into the well. Hence, he continues, this second straight line should be used in calculating the soil permeability.
To illustrate the double straight line effect, Mr. Becker selected a few sample graphs from Dr. Zimmie's Exhibit 328, Attachment 9, using both the Hvorslev and Bouwer & Rice methods. In each of these curves, Mr. Becker highlighted the clear double straight line effect and showed how the line that was used by the Town to obtain the permeability results cuts across the data and does not take the effect into account. Mr. Becker said he reviewed all the Town's graphs and most of them showed the double straight line effect to some degree. This, he added, would be expected, given the low-permeability soil and the much more permeable sand pack around each well.
The Town's failure to account for the double straight line effect occurred despite cautions in the documentation for the computer software programs that the Town used. As Mr. Becker explained, the documentation for the Hvor program (Exhibit No. 378) notifies the user that the method of solution employs averages of all of the supplied data points, and therefore that the user must be careful in the use of field data and the interpretation of results when the data depart systematically from a straight line plot. According to the documentation, such a departure may occur in low-permeability materials when a sand pack is present and the well cuts across the water table surface, exactly the situation here.
The documentation for Super Slug (Exhibit No. 379) cautions that when data are first entered into the program, "Super Slug looks for the maximum drawdown value and uses it to calculate head ratio. In some tests, there may be one or several anomalously high drawdown values at the beginning of the test. Using the points to calculate head ratio may cause invalid results." The Super Slug manual then instructs the user how to edit data to account for the double straight line effect while calculating the head ratio.
Dr. Siegel also provided a lengthy description of the double straight line effect, saying the concave-upward curvature could be attributable to well construction problems or to a well intersecting zones of different permeabilities. Dr. Siegel concluded that the curvature of most of the Town's plotted lines indicated draining of the weathered zone around the sand pack. According to Dr. Siegel, by failing to account for the significance of the double straight line effect, the Town's slug tests measured the permeability of the weathered part of the soil column due to the manner of well construction.
The significance of the double straight line effect has previously been recognized by the Department. In a discussion on the amount of recovery one must have in a valid slug test, the Commissioner's April 14, 1993, ruling on a motion to reopen the hearing in Monroe County's application to construct and operate the Mill Seat Solid Waste Landfill (Exhibit No. 191) emphasizes that a reasonable assurance that data are representative is provided where the plotted points fall on a straight line sufficiently long enough to eliminate errors which could be caused by other well influences such as the permeability of the sand pack around the well screen.
As DEC Staff notes, the County accounted for the double straight line effect in calculating the permeability values included in the permit application. On page 3 of a letter dated October 4, 1994, which is included in Volume 2, Appendix Q, of the application, Mr. Becker said in comments to Smith and Mahoney, the County's consulting firm:
"In my review of the County slug test results, a number of the recovery curves exhibit a two-stage or double straight line effect. In order to obtain an overall permeability for these wells, permeability values were calculated for each straight line segment and then averaged. However, Bouwer suggests that with this type of recovery curve, the first segment probably represents a highly permeable zone, either the adjacent filter pack or developed area, immediately adjacent to the well, consequently, the second straight line segment is more indicative of the flow from the undisturbed aquifer into the well, this effect should be considered when calculating representative hydraulic conductivity values."
According to Mr. Becker, the County subsequently recalculated its permeability values based on that comment and the values in the final version of the application reflect that change.
The Town dismissed the evidence that its slug test analyses had not accounted for the double straight line effect. Dr. Zimmie described the criticism of the Town's slug test analyses as "nitpicking." He said "it doesn't matter much if you use the double-straight-line analysis or Bouwer & Rice or whatever," adding that the various methods used by the Town to analyze the slug test data produced "fairly consistent" results.
There was no question that had the Town accounted for the double straight line effect and used the more shallow part of the response curve, its calculated values for horizontal permeability would have been lower than the ones it presented at the hearing. However, it is unclear how much lower the numbers would be, since neither party offered testimony on this point. None of the parties took the Town's slug test data and recalculated permeability with the double straight line effect in mind.
The Town argues that, in both the 1996 and 1998 hearings, County and Department Staff witnesses performed numerous calculations in support of their points and required Town witnesses to do the same when they felt it would buttress their position. The Town argues that their failure to do so with regard to the Town's slug test analyses entitles one to infer that such analyses would not have changed the Town's results.
As the Town argues, it is a general rule of trial practice that an unfavorable inference may be drawn when a party fails to produce evidence that is within its control and which one would expect it to produce. However, it was enough for the County and Department Staff to bring out the flaws that were part of the Town's slug test analyses, and it was up to the Town, which generated the data and did the analyses in the first place, to show the flaws were not significant.
In the absence of recalculated permeability values, Dr. Zimmie's testimony that "it doesn't matter" whether one addresses the double straight line effect was not persuasive, especially in light of the documentation introduced through Mr. Becker which warns about the effect and its biasing influence. Since the Town's plotted data exhibited this effect which was not accounted for in its analysis, I cannot rely on the Town's slug tests for conclusions about soil permeability.
- - Presentation of Slug Test Results
In his prefiled testimony, Dr. Zimmie presented a table (Exhibit No. 328, Attachment No. 10) listing slug test results for the 11 Town wells as calculated using Hvorslev's method and the "Hvor program" software. From these results Dr. Zimmie calculated an arithmetic mean permeability for each of the 11 wells and ultimately a geometric mean permeability for the site: 1.3 x 10-4 cm/sec.
As Mr. Becker demonstrated, however, using a wider range of analyses than the Town did in Dr. Zimmie's prefiled testimony, one may calculate a lower permeability in the range of 4 x 10-5 cm/sec. In Exhibit No. 381, Mr. Becker incorporated results from each of the three methods used by the Town: Hvorslev, Bouwer & Rice, and Ferris Knowles. From these results he developed a geometric mean permeability for each slug test, then an average permeability for each of the 11 test holes, and finally an overall geometric mean of the hole averages.
Mr. Becker emphasized that he did not believe the Town's slug test results were representative of the site soils in light of how the wells were constructed and the data were analyzed. Also, he did not maintain that Exhibit No. 381 was necessarily the best way to present the Town's data, only that it was at least as representative as Dr. Zimmie's presentation.
Under cross-examination, Mr. Becker acknowledged he was aware of no published authority supporting the averaging of slug test results from three separate types of analysis into a single result. However, he had never said that his method was correct. What his testimony did suggest is that the Town weighted its presentation with numbers generally suggesting the highest soil permeability, and that the permeabilities determined using the various methods employed by the Town were not as consistent as Dr. Zimmie maintained.
Mr. Becker also said that he considered soils having a permeability in the range of 10-5 cm/sec to be relatively tight and "not inconsistent with the intent of Part 360 to site landfills in low-permeability soils." He said a soil permeability of 4 x 10-5 cm/sec, the figure from Exhibit No. 381, would produce a groundwater velocity no greater than 24 feet per year, which "is not a hindrance to the ability to either monitor this facility or remediate in the event of a containment failure."
In addition to slug tests, the Town performed pump tests to assess soil permeability. As Dr. Brewer explained in his prefiled testimony, a pump test traditionally consists of pumping water from an aquifer and measuring the change in water levels in the pumping well and nearby observation wells. Most common for environmental purposes (and the method that was used at this site) is a constant discharge test, where one attempts to keep the pumping rate constant throughout the test. By plotting the water level as a curve and matching the data points to a known solution, it is possible, under the correct circumstances, to determine values for both horizontal and vertical hydraulic conductivity.
The Town first proposed conducting pump tests at the time of the issues conference in this matter, more than three years ago. Although Deputy Commissioner Sterman denied permission at that time, the request was renewed during the 1996 adjudicatory hearing.
My previous hearing report discusses the merits of pump testing at this site, concluding that while it would provide relevant and material evidence, such testing was not necessary to a determination on the groundwater separation variance. Pointing out that the Part 360 regulations allow the use of pump tests to determine hydraulic conductivity, I also noted the lively debate among the parties about how useful such tests would be at this site:
"The general consensus was that pump tests are useful for determining hydraulic conductivity in relatively permeable soils. [Mr. Becker] said that at sites with low-permeability soils, the pumping well is drawn down too fast, the pumping rate cannot be sustained, and the test fails. The Town turned this around on Mr. Becker by noting that whether the test succeeded or failed would itself provide information on the soils' permeability. Mr. Becker conceded that if pump tests were successful, this would tend to support Dr. Zimmie's assertion that the soils are more permeable than the County has characterized them. He also acknowledged that inferences could be drawn from a successful pump test about the continuity of any seams or layers in the soil, and that properly conducted pump tests would provide information about the soils' hydraulic conductivity over a much larger area than an individual slug test." (Hearing Report, p.104.)
The Appellate Division decision from early this year finally gave the Town permission to conduct the pump tests it wanted. As the County and Department Staff had predicted, however, the tests failed because they could not be sustained long enough.
Ideally, pump tests should be continued until equilibrium conditions are established, that is, until the aquifer adjusts to the stresses of pumping and the cone of depression stabilizes. [See Groundwater and Wells, Second Edition, by Fletcher G. Driscoll, excerpted as Exhibit No. 361.] For tests in unconfined aquifers (aquifers with a water table), 72 hours are usually required, although this time can be reduced if equilibrium conditions are established sooner.
The County's position during the 1996 hearing was that, in the soils at the proposed landfill site, no aquifer test as proposed by the Town could run for 24 hours. As it turned out, none of the tests lasted that long. The wells were all pumped dry in less than 24 hours without equilibrium conditions having been established. Standing alone, the short test durations tend to underscore the relative tightness of the soils in and around the landfill footprint.
- - Pump Test Procedure and Duration
The Town's pump tests were performed at the three well clusters along the northern perimeter of the landfill footprint:
- Cluster No. 1, consisting of PW-1, OB-1 and OB-2, near the intersection of Phases 1 and 3;
- Cluster No. 2, consisting of PW-2, OB-3 and OB-4, off the northwest corner of the footprint; and
- Cluster No. 3, consisting of PW-3, OB-5 and OB-6, off the northeast corner of the footprint.
The Town had intended to use PW-1, PW-2 and PW-3 as its pumping wells; however, since insufficient water came out of PW-3 for proper well development, the Town instead used OB-6 as the pumping well in Cluster No. 3. Therefore, the Town's pumping wells were PW-1, PW-2 and OB-6, and the remaining wells were observation wells, spaced 5 and 10 feet from the pumping wells.
During the previous adjudicatory hearing, Dr. Zimmie had said he could perform a very adequate pump test using a low flow rate of one or maybe two gallons per minute, spacing observation wells at 10, 30, 50 and 100 feet from the pumping well. Spacing observation wells as much as 100 feet from the pumping well would imply the creation of a relatively wide cone of depression, which depends on the soils being quite permeable. However, the spacing of observation wells no more than 10 feet from the pumping well suggests that when the Town actually went ahead with its tests, it lacked confidence that the soils are as permeable as it had earlier contended, as Mr. Becker observed.
The parties agreed that the actual pumping rates were one gallon per minute for PW-1 in Well Cluster No. 1, and one-half gallon per minute for PW-2 in Well Cluster No. 2 and OB-6 in Well Cluster No. 3. Even at these low pumping rates, the May 13-14 test at Well Cluster No. 1 terminated after about 21 hours when the water level was drawn down to the level of the pump at 32 feet; the May 11 test at Well Cluster No. 2 terminated after about 3 hours and 20 minutes when the water level was drawn down to the level of the pump at 33 feet; and the May 15 test at Well Cluster No. 3 terminated after about 2 hours and 50 minutes when the water level was drawn down to the level of the pump at 32 feet. (A May 7 test at Well Cluster No. 1 ended after just more than eight hours due to generator failure.)
There was general agreement among the parties that the tests did not have to run any particular time to be successful; however, they had to run long enough to get the data from which one could develop drawdown curves and, from the curves, calculate soil permeability.
As Mr. Becker noted, the typical drawdown in the outermost observation wells at each cluster during the pump tests was in the range of two to five feet, suggesting that the cone of depression (or radius of influence) around each pumping well never extended significantly beyond 10 feet. This was much less than the fifty to maybe a couple hundred feet that Dr. Zimmie had projected as a radius of influence for the tests as proposed at the initial adjudicatory hearing. At the reconvened hearing, Dr. Zimmie admitted that the area influenced by the pump tests was only about 15 feet in radius. As Staff and the County argue, this suggests a very steep cone of depression, which indicates low permeability.
- - Evaluation of Pump Test Data
To calculate permeability from the pump test data, Dr. Brewer used an unconfined aquifer solution developed by Shlomo Neuman, which involves the use of non-equilibrium (or transient) equations. Non-equilibrium equations permit analysis of aquifer conditions that vary with time and involve storage. They may be used, with limitations, in unconfined aquifers.
In his prefiled testimony, Dr. Brewer said he used the Neuman solution because the data fit it very well, and the hydrogeology of the site suggested that the response would be more like an unconfined aquifer. All the parties agree that groundwater is present within the glaciolacustrine clay under unconfined conditions; in other words, the clay is not confined by a less permeable layer above it.
As Mr. Becker testified, the Neuman method of pump test analysis is described in the text Applied Hydrogeology by C. W. Fetter, Second Edition, at pages 191-193 (received as Exhibit No. 382). The Neuman method is a curve-matching process where the time-drawdown curve from the test is matched to type-curves, match points are selected, and certain values are taken from those plots and used to derive aquifer parameters, including hydraulic conductivity. As applied to unconfined aquifers, the method recognizes that there are two different stages to the aquifer response during the pumping test, the first being elastic storage due to compression of the aquifer and expansion of water as the head drops and the pressure changes, and the second being gravity drainage from the aquifer, from which one can derive the aquifer's specific yield.
As Fetter explains, two sets of type curves are used for the matching process, Type-A and Type-B curves:
"Type-A curves are good for early drawdown data, when instantaneous release of water from storage is occurring. As time elapses, the effects of gravity drainage and vertical flow cause deviations from the nonequilibrium type curve, which is accounted for in the family of Type-A curves. The Type-B curves are used for late drawdown data, when effects of gravity drainage are becoming smaller." (Fetter, p.192.)
Fetter writes that the type curves are used to evaluate the field data for time and drawdown, which are plotted on logarithmic paper of the same scale as the type curve. First, the latest time-drawdown data are superposed on the Type B curves, from which values for transmissivity and specific yield are determined. The early drawdown data are then superposed on the Type-A curve, and a value of transmissivity is calculated that should be about equal to that computed by the Type-B curve.
As Mr. Becker and Dr. Siegel agreed, the Town's key failure was not keeping its tests running long enough so that late-time drawdown data could be obtained and a specific yield could be calculated. In his prefiled testimony Dr. Brewer acknowledged that the Town's tests were analyzed with respect to early-time data only "as no late-time responses were evident in the test data." The reason for this, he later conceded, was that the tests did not run long enough.
Dr. Siegel explained the concept of late-time data and how it relates to the Neuman method used by Dr. Brewer. According to Dr. Siegel, when the Neuman method is used, the better data for testing unconfined aquifers are when the effects of gravity drainage are observed and one can obtain values for specific yield, which is the storage parameter that sustains the aquifer under pumping conditions at equilibrium. Dr. Brewer admitted he did not calculate a specific yield for each pumping test he ran. When pressed on whether knowing specific yield is important to determine the validity of a pump test in an unconfined aquifer, he responded, "I don't believe so, not always, though it can be useful."
Indeed, Dr. Brewer admitted that he could not calculate specific yield without incorporating late-time data in the technique that he used. Mr. Becker confirmed that without late time data, there could be no matching to the Type-B curves, and therefore one could not follow the procedures in Fetter and other textbooks to derive a specific yield. Based on that, Mr. Becker concluded that the Town's tests were incomplete, that there was not enough data to establish the required curves, match them to the type curves, and from that process derive an acceptable permeability value.
"In effect," said Becker, "what happened is exactly what both the County experts and I stated in the prior hearing and also actually at the issues conference, was that in these types of soils, essentially you pump the well dry before you develop a representative curve for the aquifer from which you can derive the aquifer characteristics. So, the method of analysis wasn't fulfilled because the curve was inadequate."
As Mr. Becker explained, the failure of the Town's tests to run an adequate time is why Dr. Brewer calculated such low values for storativity (or storage coefficient, representing the amount of water that comes out of the soil once you relieve some pressure). According to Mr. Becker, these values (in the range of 0.0015) represent the elastic storage and not the aquifer's specific yield, which is attributable to gravity drainage and represents the majority of the water released from an unconfined aquifer. According to Mr. Becker, specific yields usually are more in the range of 0.1 to 0.3.
- - Appropriateness of Pump Testing
My previous hearing report adopted arguments by Mr. Becker and Dr. Siegel that pump tests would very likely be unsuccessful in the site's low-permeability glaciolacustrine soils. The Town's tests appear to bear this out, since the wells were pumped dry before a representative time-drawdown curve could be developed for the aquifer.
Although Dr. Zimmie alleged that low-flow pump tests in silty clays are common, this contention was effectively refuted by Dr. Siegel. Dr. Zimmie was able to provide only two examples of such tests from the existing literature, both of which Dr. Siegel distinguished from the tests performed at the landfill site.
One example (Exhibit No. 331) involved the pumping of a thin, sand-rich confined aquifer permeable enough that it could be used as a limited source of water for household use. As Dr. Siegel explained, the study was properly done for the environment in which it was conducted. Although the pump tests ran for only 13 and 24 hours, Dr. Siegel explained that one does not have to pump a thin confined aquifer very much in order to get a significant response in the drawdown curve. The tests described in Exhibit 331 were conducted in river delta sediments along the U.S. Gulf Coast, in relation to a discernible continuous aquifer about 18 inches deep. These soils and aquifer characteristics differ markedly from those at the County landfill site.
Dr. Zimmie's second example (Exhibit No. 332) involved pump testing in unweathered, low-permeability glacial till in Iowa. Dr. Siegel contrasted the testing in that situation to the testing performed at the County landfill site. First, the Iowa testing did not involve a constant flow pump rate, as was attempted at the landfill site, but a constant drawdown flow rate where the water level in the pumping well was drawn down to and maintained at a constant point. Second, the Iowa testing did not involve the Neuman method used by Dr. Brewer, which is based on constant flow rates, but the Finite Element computer simulation model, from which the material properties in the drawdown could be modeled. Finally, in the Iowa testing, the unweathered till was "packed off" (or separated) from the weathered till by a thick bentonite seal, which was not the case here.
In summary, Dr. Siegel demonstrated that the articles cited by Dr. Zimmie discussed studies that were not comparable to the pump tests done at the landfill site, since they were either conducted in different soil environments, or used different test procedures and methods of analysis. Also, addressing Dr. Zimmie's contention that there are plenty of articles in Ground Water about low-flow pump tests in silty clays, Dr. Siegel confirmed based on a search of back issues that there were no such articles published in that journal during the last two years, other than one of the two papers produced by Dr. Zimmie. This seemed to confirm the County's and DEC Staff's position, held throughout this proceeding, that pump tests are not well-suited for the soil environment of the County's landfill site, and should generally be limited to aquifers one could develop as water supplies.
- - Other Pump Test Issues
As noted above, there were serious deficiencies in the Town's well construction methods, which would have affected both the slug and pump tests conducted by the Town. In addition, there were important deficiencies in how the pump tests were conducted, particularly with regard to the narrow spacing of observation wells five and ten feet from the pumping wells.
As Dr. Siegel explained, one needs to put observation wells far enough away from the pumping well to ensure that horizontal flow is maintained. A text by Kruseman and De Ridder, Analysis and Evaluation of Pumping Test Data (Exhibit No. 359) states that as a general rule the nearest piezometers (or observation wells) should be placed at a distance which is at least equal to the thickness of the aquifer, where it may be assumed that the flow is horizontal. Dr. Siegel noted that the aquifer, or saturated zone, that was tested by the Town is 30 to 40 feet thick, and therefore the first observation well should have been at least that far away. In fact, according to an EPA guidance manual on aquifer pump tests (Exhibit No. 360), if only one observation well is to be used, it is usually located 50 to 300 feet from the pumping well.
Dr. Brewer asserted that the spacing of the Town's wells was correct based on an ASTM standard attached to his prefiled testimony (Exhibit No. 333, Attachment No. 11). That standard, addressing testing of unconfined aquifers by the Neuman method, states that observation wells may be located at any distance from the control well within the area of influence of pumping.
According to Dr. Siegel, the standard implies that one meets the parameters that govern a proper test, one of the most important being horizontal water flow to the pumping well. Dr. Siegel said it was inconceivable that with the observation wells spaced so closely to the pumping well, one would not be inducing substantive vertical flow in the cone of influence.
Other important failures of the Town's pump tests are noted below:
- - Failure to Evaluate Recovery Data. The Town recorded but did not analyze data for the recovery of water levels in its pumping wells after its tests ended. This is contrary to the aquifer pump test protocol set out in Driscoll's Groundwater and Wells (excerpted in Exhibit No. 361), which states:
"Whenever possible, recovery data should be taken to verify the accuracy of pumping data. Often, the recovery data will be more reliable because no pumping is required . . . Recovery measurements should be recorded with the same frequency as those taken during the pumping portion of the aquifer test."
Dr. Brewer said the Town measured water recovery following the pumping phase for a period of time not equal to the pumping time, but did not analyze the data due to time constraints.
Dr. Siegel emphasized the importance of collecting and analyzing recovery data. According to Dr. Siegel, when the water level is recovering "you don't have to worry about the variability in the pumping rates because what the aquifer is seeing, when it's rebounding, is the net effect of all the pumping and it's common practice to want to have a good recovery curve because you could calculate the same parameters you do as from the actual drawdown curves and you can compare one versus the other and in my experience, the better value is from the recovery curve because there you have more control over what's happening, water basically is coming into the well and you don't have a mechanical object such as a pump that's influencing the system."
- - Variability of Flow Rates. Dr. Siegel's concern about the variability in pumping rates was realized in the pumping tests conducted by the Town. EPA guidance (Exhibit No. 360) states that the discharge in the pumping well should never be allowed to vary by more than plus or minus 5 percent, and that the lower the discharge rate, the more important it is to hold the variation to less than 5 percent. The guidance states that the variation of discharge rate has a large effect on permeability estimates calculated using data collected during a test.
While Dr. Brewer said in his prefiled testimony that the Town maintained excellent control of its pumping rates, the County forced him to concede examples in the first test in Well Cluster No. 1, conducted May 7, and in the test in Well Cluster No. 3, conducted May 15, when the variation in pumping rates exceeded plus or minus 10 percent, the ASTM standard that Dr. Brewer deemed applicable. In fact, Dr. Brewer conceded that between the first and second hour of the first test in Well Cluster No. 1, the pumping rate variability was about 20 percent. When he was later asked whether, given variations in the pumping rate of more than plus or minus five percent, he should have taken longer recovery measurements, Dr. Brewer responded, "I don't know."
- - Excessive Drawdown and Vertical Flows. As Dr. Siegel explained, the rapid dewatering of the pumping wells most definitely induced large vertical gradients of flow into the wells, violating the assumption of the Neuman method that flow to the pumping well is horizontal. As noted in the Federal Ground Water Manual, this assumption applies regardless of whether one uses equilibrium or transient equations to analyze aquifer conditions. Therefore, the Town's use of transient equations does not eliminate this concern. Even Dr. Brewer admitted that it was "not necessarily desirable" to dewater the aquifer around the pumping well, and that he could recall nothing in the literature which allows or recommends that practice in conducting an aquifer test.
- - Town's Defense of Pump Testing
The Town defended its use of pump tests by noting language in Part 360 that "hydraulic conductivities may be determined using pump tests," among other methods also identified as suitable for a site investigation. [See 6 NYCRR 360-2.11(a)(11).] However, there is no question that pump tests may be used to determine hydraulic conductivities; the issues are how and in what environments. Part 360 does not endorse the use of pump tests in all circumstances. In fact, the hazards of conducting such tests in low-permeability lake clays were made apparent in this case, where the soils' tightness caused the Town's wells to pump dry before sufficient data could be gathered.
Also, the Town argued that while Dr. Siegel claimed the lack of late-time data invalidated the Town's pump test results, Dr. Siegel favorably reviewed a paper by Dr. Brewer (Exhibit No. 340) on a one-hour low-flow pump test in which no late-time data were analyzed. However, as Dr. Siegel noted, the soils of the County's landfill site differ markedly from the setting tested in Dr. Brewer's paper, a relatively thin, unconfined alluvial aquifer in Montgomery County which is characterized by sand with a small amount of interbedded silts. In such high-permeability deposits, Dr. Siegel said the initial part of the Neuman curve might not even express itself, and one could successfully run a test in such a productive aquifer for only one hour.
One should stress that Dr. Siegel did not say that the lack of late-time data invalidates use of the Neuman method in all cases. He said it invalidated the use of the Neuman method at the County's landfill site. Why? Because, as Dr. Siegel explained, "One of the issues is whether the sand seams, or whatever one wants to call them, are continuous, whether the fractures are interconnected, these are fundamental issues at this site, and the way to find out is to do a pumping test so that you can sustain it long enough to see if there are any effects of either dewatering these particular layers or whether you hit some sort of barrier, or if the layers terminate and end."
- - Boundary Condition at Well Cluster No. 2
One of the County's arguments is that thin sand and silt layers in the site's soil column pinch out and are essentially sealed within clay. One of the Town's pump tests which seems to bear this out is the one conducted at Well Cluster No. 2, off the northwest corner of the landfill footprint.
Dr. Brewer said he did not evaluate this test because responses in the observation wells were not typical, analysis would have been complicated, and he had limited time. As Dr. Siegel demonstrated, however, the test could be analyzed rather simply and quickly, with results that undermine the Town's arguments.
Using data from the Town's observation wells, Dr. Siegel made plots (Exhibit No. 366) which he said show a sudden, dramatic increase in drawdown at a certain point in the pump test, indicative of a no-flow boundary where any permeable seams intersecting the wells had been dewatered. Dewatering would suggest that such seams are not laterally extensive, and therefore would not pose a threat of rapid contaminant transfer, in this case to ravines northwest of the landfill footprint, a particular concern of the Town at the first adjudicatory hearing.
- - Pumping Well Switch at Cluster No. 3
As noted above, the May 15 test at Well Cluster No. 3 involved a substitution of OB-6 for PW-3 as the pumping well. The reason for this, the Town claims, is that insufficient water came out of PW-3 for proper well development. This was verified by Mr. Becker's notes, which indicate that the well was drawn dry twice during development and was bringing up water at a rate of only a fifth of a gallon per minute when Mr. Becker arrived.
As Dr. Siegel testified, the problems in developing PW-3 and its subsequent replacement with OB-6 as a pumping well tend to confirm the low-permeability of the soil at that location. Also, as Mr. Becker noted, the pump test at the PW-3 location terminated in two hours and fifty minutes when the water level in OB-6 had been drawn below the pump at a depth of 32 feet. This rapid drawdown, at a pumping rate of only a half-gallon per minute, would tend to suggest soil tightness at the northeast corner of the landfill footprint.
Test Pit Excavation
In addition to its slug and pump tests, the Town excavated 17 test pits in the area of the landfill footprint to get an overall picture of the site geology. The pits were dug with a trackhoe operated by Town highway department personnel. A typical pit dimension was 4 to 6 feet wide by 10 to 14 feet long and 15 to 16 feet deep. During the excavations, Dr. Zimmie and his assistants made field notes to document the soil profile, as did Mr. Becker, who observed the test pit excavations on behalf of Department Staff. Dr. Zimmie's field notes constitute Attachment No. 3 to his prefiled testimony (Exhibit No. 328), and Mr. Becker's field notes were received as Exhibit No. 372.
As the County and Department Staff argue in their closing briefs, the Town's test pits provided no new information about the soil stratigraphy. For instance, Dr. Zimmie found that in all the test pits there were noticeable vertical channels, from a quarter to a half inch wide, filled with a whitish, rocky material which appeared to be some form of calcium carbonate. There was general agreement that these concretions were due to geochemical processes as water moved along narrow upper soil fractures opened by frost heaving and the roots of vegetation.
In addition, Dr. Zimmie noticed in his test pits thin rust-colored vertical channels which Dr. Siegel described as root traces formed as oxygen emitted from rootlets combined with pore water that was rich in dissolved iron.
The Town's discovery of the calcite concretions and the rust-colored channels does not provide new information since these features were previously noted in the County's application. Also, these features are largely irrelevant since they are generally limited to the top 15 feet of the soil profile, which will be removed during landfill construction. Dr. Zimmie claimed that a leak through the side of the landfill might intercept the calcite concretions; however, using an analogy to a leaking bathtub, Mr. Becker said there is no hydraulic head on a landfill side wall, so it is not an avenue for significant contaminant migration.
- - Soil Collection
As the Town removed soils from its test pits, it put them into spoils piles, generally separating soils from different depths. Seven soil samples taken from these piles were then subject to laboratory testing including grain size analyses to determine relative proportions of clay, silt, and sand.
As both Dr. Siegel and Mr. Becker explained, the Town's laboratory tests were biased in light of the manner in which soils were collected for sampling. Dr. Siegel observed Dr. Zimmie scraping selected layers from the spoils piles with a screwdriver to form some sort of composite sample. Similarly, Mr. Becker said that, after a pile was set up, Dr. Zimmie and his assistants "would then typically pick through the pile and look for chunks of soil generally that appeared to have a higher proportion of silt in them or in some cases chunks were removed from the pile and a screwdriver was used to scrape silt and fine sand, presumably, from the chunks into sample jars." Mr. Becker said this was not a typical soil collection technique because normally the objective of collecting soil samples from a test pit is to get ones that represent all soil units within the interval of concern.
Under cross-examination by the County, Dr. Zimmie admitted his gathering of soils did not obtain representative samples, and that the sampling was selective. He admitted using a screwdriver and knife to choose composite samples that were then jarred and sent to the lab.
In surrebuttal, Dr. Zimmie defended his atypical soil collection technique, saying he was not doing an investigation as part of a Part 360 landfill application, but an "anti-landfill" investigation. He said he was looking for permeable layers to verify the presence of silt and sand in them, adding that "I was not looking for clay. We would have accomplished nothing if we said we found clay on the site."
- - Lab Testing of Soil Samples
Dr. Zimmie said the lab testing was done because visual soil classifications are somewhat subjective and different observers may classify the same soil differently, especially soils on the silt-sand and clay-silt boundaries. The results of the Town's lab tests, as performed by Soil & Material Testing, Inc., are appended to Dr. Zimmie's testimony (Exhibit No. 328) as Attachment No. 4, and Dr. Zimmie's list of samples for potential testing was introduced as Exhibit No. 371. Comparing Dr. Zimmie's visual description of samples as documented on this list with the test results when these same samples were analyzed, Mr. Becker effectively discredited Dr. Zimmie's indications that the samples contained appreciable or significant quantities of sand.
More particularly, Dr. Zimmie's list contained the following descriptions for five samples: No. 1 ("fine sand & coarse silt"), No.2 ("silt & fine sand, very little clay"), No. 3 ("fine sand, coarse silt"), No. 11A ("clayey silt - silty clay & fine sand - coarse silt"), and No. 16A ("coarse silt, fine sand & trace clay"). However, as Mr. Becker testified, when each of these five samples was tested, the grain-size curve indicated that it was less than five percent sand, a negligible amount.
The Town has maintained throughout the proceeding that sand is a significant constituent of the varved clay soils, and that lateral layers of sand would provide the fastest means for contaminants to escape the landfill site in the event of a containment failure. However, Mr. Becker noted very little sand during 11 days observing the Town's test pit excavations and soil borings. Likewise, Dr. Siegel said he observed at the site "mostly clay with secondary silt," and not a lot of sand.
The lab classifications provided by Soils & Material Testing, Inc., for the seven tested soil samples all include the adjective "sandy." Three of the samples are described as "sandy lean clay," three as "sandy silt," and one as "sandy elastic silt." As Mr. Becker explained, however, these group names are misleading and not consistent with ASTM D 2487, which was used for the soil classifications. ASTM D 2487 (Exhibit No. 368) provides that a soil classification should have the adjective "sandy" only if the sample is more than 30 percent sand. Since all of the lab tests showed less than 5 percent sand, this adjective should not have been included, as Mr. Becker explained, since it tends to suggest more sand than was actually found in the samples.
Even with a selection process that was biased to exclude clay and emphasize the most permeable materials, three of the soil samples tested as predominantly clay and four as predominantly silt. The presence of significant amounts of silt in the varved clay deposits is not a new finding; it was acknowledged in the permit application and the previous hearing. Silt, though more permeable than clay, is not as permeable as sand. Clay with a secondary amount of silt is expected in a typical glaciolacustrine varved deposit, as both Dr. Siegel and Mr. Becker explained at this and the prior adjudicatory hearing. Far from being an impediment to landfill siting, silt is compatible with it, as evidenced by Part 360's preference for landfill siting in low-permeability, clay- and silt-rich soils.
Vertical Desiccation Features
Prior to this reconvening of the adjudicatory hearing, the Town's primary concern was the horizontal permeability of the site's soils. According to Dr. Zimmie, prior to the Town's site investigation it was assumed that the vertical permeability of the site soils was quite low, on the order of 1 x 10-7 cm/sec. This figure was derived from laboratory tests which the County conducted on Shelby tube samples of intact clay, and based on Dr. Zimmie's observations of the amount of clay on the site the figure seemed quite reasonable to him two years ago.
I wrote in my prior hearing report (on page 95) that the soils' low vertical permeability was acknowledged by Dr. Zimmie, and that, at any rate, vertical permeability is not very important relative to contaminant transport at this site. In my findings of fact I emphasized that contaminants reaching the end of a horizontal sand lens in the varved clay soils would have to move downward to find other relatively permeable layers to continue their lateral migration, and the very low vertical permeability of the varved clays would tend to impede the contaminants' progress. (Finding of Fact No. 24, p.84.)
The Town now claims that the vertical permeability is much greater than it had initially conceded, in effect interjecting a new issue into this case. Combining results of its two pump tests in Well Cluster No. 1, mid-way along the northern perimeter of the landfill footprint, with the one pump test in Well Cluster No. 3, off the northeast corner of the footprint, the Town now maintains that the vertical permeability of the site is two orders of magnitude greater than what the County calculated during its lab tests.
Dr. Zimmie attributed this new, higher permeability value to numerous vertical seams that he said he observed in both the Town's test pits and in four of its five soil borings as logged in Attachment No. 5 to his testimony. Likewise, Dr. Brewer attributed these "surprisingly high" values to vertical seams he observed in the drilling cores and which he said are recorded in the boring logs. To explain the discrepancy between the County's and the Town's vertical permeability estimates, Dr. Zimmie said these seams may not have been encountered in the County's small Shelby tube samples.
Dr. Zimmie said the vertical seams would provide a rapid avenue for contaminant transport underneath the landfill directly to bedrock, where contaminants would be extremely difficult to remediate. However, when asked to identify a vertical seam in the Town's boring logs that extended from a depth equivalent to the landfill bottom down through the entire clay unit to bedrock, he could not do so.
For Town Well MW-2, where bedrock was encountered at 22 feet, Dr. Zimmie noted vertical seams between 7 and 8 feet, at about 12 feet, between 14 and 16 feet, and between 18 and 20 feet. However, for Town Well MW-1, he agreed that the longest vertical seam in the boring log was six inches in length, and that there were no vertical seams between the 25.7 foot level down to bedrock at 49 feet. For Town Well PW-1, with the exception of an unexplained gray vertical line at the level of 30-32 feet, there were no other references in the boring logs to what might possibly be considered a vertical seam. Similarly, for Town Wells PW-2 and PW-3, no vertical seams were noted, although in PW-3 the well log describes a small angled silt layer near 23 feet and slightly tipping silt seams in the varved clay in the area between 26 and 28 feet, which Mr. Becker described as likely contorted bedding caused by turbidity currents in the glacial lake where the sediments were deposited.
Mr. Becker developed sketches from the Town's boring logs to illustrate what might be construed from the logs as vertical seams, but which Mr. Becker referred to as desiccation cracks. Mr. Becker said these desiccation (or shrinkage) cracks were formed long ago in geologic time at points when the moisture-laden clay dried, underwent a volume reduction and broke apart. He said such cracks are very common in glaciolacustrine clay deposits and exist in all similar deposits he has seen in the area of the project site. He said they are expected because the deposits were laid down in a glacial lake some 10,000 to 12,000 years ago "when they were obviously saturated and were subject to some degree of drying later on."
As Mr. Becker explained, all the documented vertical seams (or desiccation cracks) are observed above the transition zone from brown to gray clays. Generally speaking, the clays at the site are brown at the surface, mottled with gray as one goes deeper, and entirely gray at the farthest depths approaching bedrock. Brown coloration of the clays indicates soil oxidation at some point since the clay was deposited. Gray coloration indicates permanent waterlogging, which precludes the formation of desiccation cracks. The brown clay in the upper portion of the soil profile is stiffer or firmer than the underlying gray clay, since the brown clay was dried at some point and thereby became very hard.
According to Mr. Becker, the desiccation cracks were well-known to Department Staff at the time of the initial adjudicatory hearing and were evident in test pits revealing the top 10 feet of soil, which is most subject to wetting and drying cycles. Mr. Becker agreed that desiccation cracks can transmit water vertically in the uppermost portion of the soil profile, but added that they do not transport any significant quantities of water at depths deeper than about 10 feet. He offered several reasons for this:
"First of all, the density of the cracks is much less with depth, in other words, there are many more desiccation features near the surface due to the greater amount of wetting and drying than there are at depth. Secondly, desiccation features taper with depth, again, they start at the surface, where the drying actually begins, and then the shrinkage occurs downward from that point, so the maximum degree of shrinkage occurs near the surface, less at depth, so the width of the crack actually decreases with depth, it tapers downward. Finally, desiccation features are frequently filled with other material, which was the case at this site, and at depth they are filled with silt, as the Town's test pit and boring logs, in this particular instance, note. So, the ability to transmit water through that feature would be limited by the permeability of the silt essentially."
On cross-examination, Mr. Becker conceded that in the area around County Well MW-G, the brown clay extends all the way to bedrock. He said it is conceivable that this is also the case in an area extending several hundred feet north from that location. He also acknowledged that for at least part of the year, during the seasonal low groundwater condition, the water table would drop below the pore pressure relief system in the vicinity of MW-G, and any flow of leaking contaminants would be vertical, toward the bedrock.
In its closing brief the Town argues that this combination of factors makes the site unsuitable for a landfill. I do not agree with this assessment. As Mr. Becker pointed out on redirect, the brown clay is still a low permeability material. While the extension of brown clay to bedrock suggests the possibility of desiccation cracks that also extend to bedrock, all the testimony indicates that the cracks are more common closer to the surface, and as even Dr. Zimmie acknowledged, fracture density generally decreases with depth. Mr. Becker testified that not only do desiccation cracks taper with depth, but they are also filled with silt, which itself has relatively low permeability.
In addition, the 1988 version of the Part 360 regulations, which governs this application, provides an additional protection regarding any vertical crack in the brown clays beneath the landfill liner. More specifically, 6 NYCRR 360-2.13(i)(3)(ii) provides that in constructing the landfill, the subgrade (meaning that area upon which the facility will be placed) must be "proof-rolled" using procedures and equipment acceptable to the Department. According to Mr. Becker, proof-rolling means taking a vibratory compactor and rolling the surface to make it smooth and uniform. It reworks the material such that any desiccation cracks or features of that type would be reworked and sealed to some degree. As the County points out, Section 360-2.13(i)(3)(i) also requires that before placing any material over the subgrade, the project engineer must visually inspect the exposed surface to evaluate the suitability of the subgrade and ensure that the surface is properly compacted, which would provide some additional assurance of environmental protection.
While I do not find that desiccation cracks pose an environmental threat in this instance, my previous hearing report included as a possible condition for a groundwater separation variance "over-excavation of the soil as part of the landfill's construction and the subsequent re-compaction of the clay to break up any sand and silt lenses and, in effect, create another liner below the pore pressure relief system." (Hearing Report, p.106.)
The idea of over-excavating and then recompacting the soil was raised by Dr. Zimmie, who said it has been employed in Wisconsin, though not apparently for landfills with a double-liner design.
I did not recommend this practice in my initial report, and I am not recommending it now in relation to alleged desiccation cracks. Such a condition, to in effect create a third liner, is unnecessary in light of the demonstrated safety afforded by a design that incorporates a double liner and two leachate collection and removal systems. However, if the Commissioner has further concerns, I would recommend over-excavation and recompaction of the landfill subgrade as an alternative to denying the groundwater separation variance outright.
Denying the variance would not necessarily prevent the project from going forward, but it would require the County to import a substantial quantity of fill to construct the liner system subgrade and thereby ensure the five-foot separation between the liner system base and the seasonal high groundwater table.
Relation of New Evidence to Previous Record
The recent hearing in this matter is basically a continuation of a previous hearing on the same issue: the environmental impact of granting a groundwater separation variance. The recent site work by the Town has to be reviewed against the accumulated evidence in the record developed in 1996, and the results of the Town's permeability testing have to weighed against those of the County's. The County's hydrogeologic investigation is documented in Section 7.0 of its permit application and explained in my initial hearing report.
I previously concluded that, despite not meeting the groundwater separation requirement, the County landfill would not have a significant impact on the environment "given the design of the landfill, the characteristics of the surrounding soil, the soil's overall permeability, and the distances from the landfill footprint to nearby ravines and the Snook Kill aquifer." Of these considerations, the Town's site investigation primarily addressed soil characteristics and permeability.
The Town's testing basically confirms my previous finding that the soils at the site are glaciolacustrine clay which is banded with thin layers of silt or fine sand. The varved nature of the soils is confirmed by the Town's test pits. The predominance of clay and, to a lesser degree, silt, and the minor occurrence of sand are confirmed by the Town's lab tests, despite the bias introduced in Dr. Zimmie's sample collection process.
The soils appear to be homogeneous, displaying the same characteristics across the site. No large scale coarse-grained deposits such as channel fillings or sand beds have been identified. While there are vertical desiccation cracks, they narrow with depth and are filled with silt. Therefore, they are not good conduits for contaminant migration.
On the issue of soil permeability, the Town's test results differ dramatically from the County's. Given this difference, the question is which are more accurate. I adopt the County's values since they are more representative of the site's known soil composition, and because the County's testing survived scrutiny at the initial hearing, whereas the Town's testing was shown to be flawed.
Faced with extensive criticisms of its permeability testing, the Town, during the cross-examination of Mr. Becker, sought to turn the focus back to possible flaws in the County's permeability testing and analysis. However, at the direction of Deputy Commissioner Sterman, the County's permeability testing was examined exhaustively at the prior proceeding, and the Town had a full opportunity then to question witnesses, including Mr. Becker, on that work. Therefore, I disallowed pursuit of certain issues about the County's slug testing, finding that they were untimely and that the proper focus of this hearing was the Town's testing, what it demonstrated, and whether it was flawed.
Although landfill design was not a major emphasis of the recently-concluded hearing, it was a major focus of my initial hearing report. Landfill design is critical to the consideration of the groundwater separation variance since it is expected that the design will prevent the escape of contaminants in the first place. As argued by DEC Staff and confirmed in my prior hearing report:
"The Department will typically grant a groundwater separation variance where a landfill permit applicant has conducted an appropriate siting study resulting in the selection of a site with the desired low-permeability, silt- and clay-rich soils, and an associated high water table. In these instances, however, the applicant is also required to design a groundwater suppression system to prevent hydrostatic uplift pressures from "floating" the liner." (Hearing Report, p.90.)
As noted in my previous findings of fact, the County's design incorporates a double-composite liner and two leachate collection and removal systems (Findings of Fact Nos. 6-10, pages 81-82), as well as a groundwater suppression system also known as a pore pressure relief system (Findings of Fact Nos. 11-17, pages 82-83). This system would consist of a two-foot thick layer of very high permeability coarse sand interlain with a pipe network draining via gravity to manhole pump stations, the flow at which could either be released to nearby drainages or pumped back to the leachate storage tanks. (Finding of Fact No. 12, p.82.)
In addition to relieving hydrostatic pressure, the pore pressure relief system would act as a horizontal monitoring well under the entire landfill footprint. Water quality from the system's discharge points would be sampled on a quarterly basis during normal operation, and more frequently if the action leakage rate in the liner's secondary drainage layer is exceeded. (Finding of Fact No. 13, p.82.) As Mr. Becker noted in the initial proceeding, the addition of the pore pressure relief system makes the landfill at least as safe as one meeting the groundwater separation requirement, since it provides another layer of leak detection capability that would not be present in a standard landfill design. And of course the sooner any leak is detected, the sooner remediation can occur.
In his July 14, 1995, interim decision, Deputy Commissioner Sterman said that because the County's soil permeability measurements were directly linked to the engineering design of the pore pressure relief system, it was critical to verify the County's permeability data to ensure the County became aware of any potential redesign the system might need. As the County demonstrated at the prior proceeding, the system is designed to handle 325 gallons of inflowing water per minute, which Dr. Siegel said is fully 12 times more than it would receive under what he described as "foolishly worst-case" conditions. Both Dr. Siegel and Dr. Gregory Richardson, a geotechnical engineer who also testified for the County, agreed that the system was conservatively designed even if the varved soils have a horizontal permeability in the range of 1 x 10-4 cm/sec, as the Town has maintained throughout this proceeding.
Although mentioned explicitly by Deputy Commissioner Sterman, the adequacy of the pore pressure relief system to handle incoming groundwater was not a concern of the Town at the initial hearing. In fact, Dr. Zimmie, the Town's main expert, testified then that the system's two-foot thickness and the size of its pipes "appear to be fine."
However, when the hearing reconvened, the adequacy of the pore pressure relief system came under question based on problems alleged by Dr. Zimmie in the similar system underlying the recently-opened Finch, Pruyn & Co. landfill. That landfill receives paper sludge, borders the County site on the south, and has a design similar to that proposed for the County's landfill.
Asked by the County whether he was aware of any pore pressure relief system in New York State that had failed, Dr. Zimmie answered, "If you want my opinion, I'd say the Finch Pruyn one probably or it's about to, it's got problems." When I asked him to explain, Dr. Zimmie said he had seen some reports received by the Town which indicated that the secondary leachate collection system of the Finch/Pruyn landfill was "quite routinely" exceeding its action leakage rate of 20 gallons per day per acre.
"What exactly the problem is, I don't know," said Dr. Zimmie, "but they're exceeding the action leakage rate. We have a landfill next door that's not operating too good, I suspect because of the high permeability soils, I don't know, it seems kind of dangerous to approve another landfill, at least until you get the other one straightened out."
While the County did not pursue Dr. Zimmie's answer, DEC Staff presented a witness, Robert J. Phaneuf, to explain the situation at the Finch/Pruyn landfill. A registered engineer with extensive experience in landfill design and operation, Mr. Phaneuf is a section chief in DEC's Division of Solid and Hazardous Materials. He is familiar with the Finch/Pruyn landfill, including its liner and pore pressure relief systems.
As Mr. Phaneuf explained, starting in December 1997, the secondary leachate collection and removal system of the Finch/Pruyn landfill took some large spikes of leachate flow such that by January 11, 1998, the 30-day average of the action leakage rate was over 20 gallons per acre per day. Pursuant to 6 NYCRR 360-2.9(j)(4), this required that the Department be notified, and that the landfill operator take steps to try to identify the problem, since it could suggest a leak in the primary liner.
After an extensive investigation, it was determined that the anchor trench at the landfill perimeter was saturated and that stormwater that had built up in the trench was seeping in between the primary and secondary liner. In other words, the exceedence in the 30-day average was caused by stormwater from the anchor trench in effect short-circuiting the liner system, and not from a liner leak.
To address the problem, on or about June 1, 1998, Finch/Pruyn completed constructing a drainage trench around the perimeter of the anchor trench. Based on a June 25 inspection by Mr. Phaneuf and leachate monitoring data provided by Finch/Pruyn, Mr. Phaneuf testified that the drainage trench seems to be working.
Mr. Phaneuf removed any concerns left from Dr. Zimmie's testimony that the exceedence in the action leakage rate was due to high permeability soils or a failure in the pore pressure relief system at the Finch/Pruyn landfill. Mr. Phaneuf said there was no indication that the pore pressure relief system or its design was related to the issue of elevated flows into the secondary leachate collection and removal system, adding that the pore pressure relief system is designed "to allow groundwater to basically be removed from, and keep it from coming in contact with, the lower portions of the landfill's liner system. It would be very unlikely for that system to build up enough hydrostatic pressure to basically cause or contribute to elevated flows in the secondary leachate collection/removal system here. It would be almost impossible."
Mr. Phaneuf reported that, during his June 25 inspection, there was relatively low flow from the outfall of the Finch/Pruyn landfill's pore pressure relief system, even after rain the day before.
As noted above, the County maintains that its pore pressure relief system is adequately engineered to handle 325 gallons of inflowing water per minute. While the Town did not question this figure, Dr. Zimmie said that if the soil is as permeable as he thinks, a "dangerous situation" could develop in which the system receives more water than it can handle, causing uplift pressure on the liner and potentially a "blow out" if the water breaks through. Dr. Zimmie equated the tearing of the liner and resulting containment failure to "a bomb in the landfill, that's about what would happen, that it would be the bomb in the landfill, it would just blow out."
According to Dr. Zimmie, the County's calculations used to develop the pore pressure relief system should be rechecked, and the system possibly re-engineered for greater thickness, in light of the Town's permeability estimates. I disagree for two reasons. First, the Town's permeability values are not reliable, based on the flaws set out in this report. Second, the County's calculations already assume high soil permeabilities equivalent to those estimated by the Town.
Conceding that he could not remember how Dr. Siegel arrived at the 25 gallon per minute figure, Dr. Zimmie incorrectly assumed in his testimony that it was based on a permeability of 5 x 10-6 cm/sec. In fact, Dr. Siegel assumed a horizontal permeability of 10-4 cm/sec, and a vertical permeability of 10-6 cm/sec. These values suggest much more permeability than the County's own testing proved. Dr. Siegel also assumed the highest possible hydraulic gradient (or driving force to move the groundwater), even though the calculated natural hydraulic gradient is almost 100 times smaller, and should be diminished further as the water table drops due to the trenching at the Finch/Pruyn landfill and dewatering caused by the pore pressure relief system.
In summary, all the evidence suggests that the pore pressure relief system is adequately designed and will not be inundated with water. Functioning as designed to prevent upward pressure on the overlying liner, the system would avoid the possibility of a liner "blow-out" sending contaminants into the surrounding environment.
Rate of Contaminant Travel
Finally, in considering whether granting the groundwater separation variance would have a significant environmental impact, one needs to look at where and how fast contaminants would travel in the event of leakage or containment failure. At the initial hearing, this issue was considered in terms of the lateral movement of leachate away from the landfill in semi-permeable sand and silt seams. At the reconvened hearing, the issue was broadened, based on the Town's new estimate of vertical permeability. New concerns were raised about contaminants moving downward from the landfill along vertical seams and then penetrating the underlying bedrock, from which they could not easily be captured.
- - Horizontal TravelIn Table 7-2 of the permit application, the County calculated horizontal groundwater flow velocities ranging from 0.4 to 3 feet per year in the glaciolacustrine clay unit of the landfill site, assuming an average hydraulic conductivity of 4.8 x 10-6 cm/sec and a range of possible values for effective porosity.
In his prefiled testimony, Dr. Zimmie notes how he substituted the Town's hydraulic conductivity figure of 1 x 10-4 cm/sec, and came up with a horizontal flow velocity of up to 62.5 feet per year. For the most permeable layers, those having what he assumed to be a hydraulic conductivity of 5 x 10-4 cm/sec, Dr. Zimmie then calculated an even greater velocity, 312.5 feet per year.
These numbers of the Town are not reliable since they are based on flawed testing which overstates soil permeability. Also, the Town's rate of 62.5 feet per year assumes the value of effective porosity most favorable to the rapid flow of groundwater; as noted on page 7-41 of the County's application, the effective porosity of silty clay sediments is a source of considerable disagreement, which explains the wide range of flow velocities calculated by the County.
Finally, even assuming an average flow velocity of 62.5 feet per year, it would still take more than three years for contaminants to reach deep, water-bearing ravines which the Town says are 200 feet from the southeast corner of the landfill footprint, on property owned by Finch/Pruyn, and four years for contaminants to reach a ravine northwest of the landfill in which the Town says water is breaking out at a distance of 250 feet from the proposed landfill footprint. (The Town also voices concerns about water-bearing ravines it says are about 600 feet from the eastern edge of the landfill, and a stream located about 1000 feet to the north of the landfill, just beyond the tree line to the north of the County's MW-2 well cluster.) Presumably this would allow enough time for site remediation before any significant offsite impacts were felt.
As noted above, the Town says contaminants would move most quickly, at 312.5 feet per year, in the most permeable layers of the soil profile. However, this is a concern only if one believes such layers are laterally extensive. If these layers do not extend several hundred feet, directly to the areas of concern, the flow of contaminants would be impeded dramatically. In fact there is no evidence that the most permeable soil layers have any great continuity. As noted in my prior hearing report (Finding of Fact No. 24), such sandy layers in the varved clays tend to appear in thin lenses that pinch out over short distances. Generally being less than a half-inch thick, they are too thin to carry a large amount of contaminants.
In his prefiled testimony Dr. Zimmie said "it seems certain" that "appreciable" portions of the soil profile have a permeability of 5 x 10-4 cm/sec or greater if one assumes an overall site hydraulic conductivity of 1 x 10-4 cm/sec and a high percentage of clay on the site. This claim was challenged effectively by Dr. Siegel, who said that he saw no evidence of any laterally continuous layer having a permeability of 5 x 10-4 cm/sec. Also, as Dr. Siegel explained, Dr. Zimmie misrepresented what one does in solute transport calculations.
According to Dr. Siegel:
"What one does is determine a geometric mean on the hydraulic conductivity, apply hydraulic gradient, as appropriate, and then what that value will be when applied to Darcy's Law will be the average what's called advective horizontal flow of velocity, which engineers often call seepage velocity, another term is the average linear velocity, they're all the same thing, and this is what's used in solute transport calculations to determine where the center of the mass of any potential contaminant might be along the potential plume that might hypothetically evolve. Now, with respect to those layers within the sediment that might have higher permeabilities or lower permeabilities than the average, what is done is apply the dispersion coefficient to it . . . and it's my contention that the dispersivities along these flow paths are very small . . . What one doesn't do, I don't think, is come up with the geometric mean of what you think is happening and if you don't like what it gives, you make up a different number in order to get the flow to be faster."
At the initial hearing there was extensive testimony about the effect dispersion would have in spreading the leading edge of any contamination faster than the average velocity of groundwater moving along a flow path. Because water moves slowly in clayey soils, dispersion would not be a major factor at this site, as both Mr. Becker and Dr. Siegel explained. The clay would not only inhibit the flow of leachate, it would tend to filter contaminants from the groundwater. Even Dr. Zimmie admitted that, in general, leachate is fairly dilute. Therefore, it would take a considerable amount, more than just the leading edge, to have a significant effect on water quality.
While the Town's contaminant transport estimates are meant to suggest an environmental threat, the Town does not account for a number of safety factors emphasized in the County application and my prior hearing report. These include the early leak detection capacity afforded by the landfill design (see Findings of Fact Nos. 30 and 31 from my prior report), the network of surrounding monitoring wells (which, in the event of a leak, could be supplemented), and the monitoring of groundwater elevations and flow patterns (pursuant to special permit condition No. 10.)
As noted in my prior report, as long as groundwater is in contact with the pore pressure relief system, it will flow into the system and the only way contaminants could migrate away from the landfill would be by diffusion, a very slow process. (Findings of Fact Nos. 18 and 19, p.83). Even if the water table drops below any portion of the landfill for part of the year, there will be other periods when the water table is high and water is moving into, not away from, the pore pressure relief system.
In his pre-filed testimony Dr. Zimmie says that the monitoring wells would not necessarily detect contamination before it reaches adjacent drainage ravines. He claims that contaminants leaking from the midpoint of the western edge of the landfill could easily reach a ravine northwest of the site without ever being detected by monitoring well MW-1, and that other contaminants could also travel directly to ravines southeast of the site without being intercepted by any monitoring wells.
As Mr. Becker demonstrated, such scenarios are unrealistic in light of the final environmental monitoring network as shown on hydrogeologic drawing Sheet No. 10, part of the County's application. This sheet shows a monitoring well cluster (MW-5) located mid-way along the western edge of the landfill which would in all likelihood detect the first leak Dr. Zimmie hypothesized. This cluster does not currently exist, but would be installed prior to the landfill's construction.
As for leaks moving in other directions, Mr. Becker emphasized that the County's spacing of monitoring wells exceeds the requirement of 6 NYCRR Section 360-2.11(c)(1)(i)(b) that such wells not be spaced more than 500 feet apart along the downgradient perimeter of the facility. He said that with any monitoring network one can always draw lines between well clusters and say that contaminants could get between them. However, leachate does not travel in straight lines like an arrow; it creates a spreading plume that facilitates detection.
Dr. Zimmie said that, during his site work this past spring, he observed groundwater flowing continuously from drain tiles and pipes into a ravine west of the site. He did not know the extent of the tiles, but said they were of the type commonly used to drain farmland.
Dr. Zimmie claimed that his observation of these drain tiles emptying into a ravine near the landfill "raises the possibility of a dangerous contaminant pathway." However, as he conceded on cross-examination, the drain tiles are within a few feet of the land surface. At that level, they would be above the base of the landfill and unable to intercept contaminants.
Dr. Zimmie added that where the drain tiles exit into the ravine, they are probably about 20 feet below the landfill. Even assuming this is true, it is not a problem. As Mr. Becker explained:
"The landfill site is basically flat or relatively flat across most of the surface, except where you get close to a ravine. In clayey soils, typically, because when it rains, the water doesn't percolate down through the clay, it drains or runs off the site through surface runoff, you tend to get a number of deeply incised or deeply eroded drainages around the area, which is exactly what you have here. So, consequently, when you approach these ravines, you have a sharp change in the topography . . . Now, if you have a drainage tile that's running approximately parallel to the ground surface at a shallow depth, it's going to be above the landfill liner for the area of concern at the actual ravine itself, it may drop down into the ravine to where it's at a depth below the liner where it actually discharges, but from the standpoint of contaminant transport that's not what you're concerned with because that's just at the ravine itself."
As at the previous hearing, the Town attacked the County's estimate of average horizontal hydraulic conductivity, which bears upon its values for horizontal flow velocity. The Town contends that this permeability estimate of 4.8 x 10-6 cm/sec is biased by smearing of clay that likely occurred during the drilling of undeveloped test holes in which some of the County's slug tests were performed.
My initial hearing report discussed how the clay's blockiness and low plasticity would discount the possibility of a significant smearing problem. However, even if one limits consideration to the County's slug tests performed in developed, cased wells, this still results in an average permeability of 1 x 10-5 cm/sec, according to the Town's expert, Dr. Brewer. As Staff argues, this value is roughly two times higher than the County's, and would only increase the County's highest estimate of groundwater velocity in the clay unit from three to six feet per year.
- - Vertical Travel
Based on an estimate of vertical hydraulic conductivity of 7.5 x 10-8 cm/sec, the County concluded that groundwater flows vertically no faster than 0.31 feet per year in the clay underlying the landfill footprint area. (Application, page 7-56.) This was accepted by all parties at the initial adjudicatory hearing and confirmed in my findings of fact, where I noted that this very low vertical permeability would make it extremely difficult for contaminants to migrate downward from the base of the landfill. (Finding of Fact No. 83.)
Making the same assumptions as the County and changing the vertical hydraulic conductivity to the 4.4 x 10-5 cm/sec value calculated from the Town's PW-1 pump test, the resulting flow velocity is between 23 and 175 feet per year, according to Dr. Zimmie's pre-filed testimony. Since bedrock is generally located about 20 feet below the landfill liner, Dr. Zimmie writes that at these rates contaminants could reach bedrock in as little as 7 weeks, and certainly in less than a year.
Again, these flow velocities are not reliable since they are based on flawed testing. These errors in relation to the Town's pump tests, which were used to determine vertical permeability, are outlined above.
Depth to Bedrock
The Town implied through its cross-examination of Mr. Becker that the separation between bedrock and the landfill liner system is inadequate in the vicinity of the Town's boring MW-2. At this location, just off the southern boundary of the landfill's phase one footprint, the Town located bedrock at about 22 feet below the ground surface. Since this is about 10 feet higher than what the County's top of bedrock contour plan [Sheet No. 2, hydrogeological drawings] would have suggested, the Town then asked whether it was possible that the bedrock might also be 10 feet higher at the next dashed bedrock contour line to the north, which intersects the southwest corner of the landfill footprint at a marked elevation of 190 feet.
Mr. Becker agreed that it was "certainly possible" that the elevation was slightly higher in that direction, which led the Town to ask whether, if it was 10 feet higher (putting the bedrock at an elevation of 200 feet), there would be twenty feet of unconsolidated deposits underneath the constructed landfill liner. Mr. Becker responded that he believed there would still be 10 feet of separation between the bedrock and the base of the liner system, which he estimated to be at an elevation of 214 feet along the southern phase one perimeter.
The thrust of the Town's questions became apparent when it asked Mr. Becker whether he was concerned about the landfill's compliance with a Part 360 requirement of 20 feet of unconsolidated deposits beneath the bottom of the constructed liner system. However, as Mr. Becker quickly pointed out, this provision [6 NYCRR 360-2.12(a)(1)(v)], to which there are exceptions, is from the 1993 version of the Part 360 regulations. As all the parties understood at the start of this hearing, the County's application is governed by the construction requirements of the 1988 Part 360 regulations, because it was noticed as complete before the 1993 regulations took effect. [See 6 NYCRR 360-1.7(a)(3)(vi), 1993 regulations.]
Rebuffed on this argument, the Town then implied that the siting requirements of the 1988 regulations, and specifically 6 NYCRR 360-2.12(d)(1) which states a preference for sites underlain by thick sequences of homogeneous, clay and silt rich, low permeability materials, should be interpreted in light of the 1993 requirement, so that anything less than 20 feet would be deemed not thick enough for safety. This argument also was effectively turned back by Mr. Becker, who emphasized that the only requirement for separation between bedrock and a landfill's liner system in the 1988 regulations is contained in 6 NYCRR 360-2.13. This construction requirement [Section 360-2.13(e)], which also exists in the 1993 regulations, is that a minimum of ten feet of vertical separation be maintained between the base of the constructed liner system and bedrock.
Mr. Becker noted that northwest of MW-2, in the southwest area of the phase one footprint, there is an existing County well MW-C. As Staff argues, since the boring log for MW-C (Appendix K of the application) indicates that bedrock there is encountered at an elevation 188.8 feet, one may conclude that the bedrock deepens significantly as one moves northwest of MW-2 and onto the landfill footprint. Therefore, it is not reasonably necessary to conduct additional borings to assure the 10-foot separation requirement is met throughout the landfill footprint, as both Mr. Becker and Dr. Siegel agreed.
While the main purpose of this report is to address the evidence from the recently-concluded hearing, whether or not to grant the groundwater separation variance depends on an evaluation of an entire record the first part of which was developed in 1996. My initial report, attached to Deputy Commissioner Sterman's decision of September 3, 1996, discusses all matters raised during the previous hearing and contains 28 findings of fact (see pages 80-85) which bear directly on this issue. As the evidentiary record closed, I asked the parties to brief whether any of these findings require revision. Upon review of their papers, I find no reason to change any of the findings. Therefore, they are hereby reconfirmed.
The Town claims that my previous findings must be re-evaluated in light of the new evidence resulting from its testing. However, I conclude that the testing was flawed and therefore its results are not reliable. Although the Town says the site soils can no longer be characterized as having low permeability (Finding of Fact No. 1), I find that description to be appropriate. Because of their varved clay composition, they have an especially low vertical permeability, which alleviates any concern one might have about their thickness.
Proposals by County and DEC Staff
The County and DEC Staff propose that I modify or delete the last sentence in Finding of Fact No. 25. This finding reads as follows:
"Layers of silt are also part of the varved sediments beneath the site. These layers are thicker than the sand features, varying from one to three inches. In some cases, they may also be quite extensive, although there is no way to verify this based upon existing hydrogeological data. One silt layer may exist within several feet of the bottom of the landfill liner, extending in the direction of the ravines generally north and west of the site."
Staff proposes that the last sentence be deleted because the Town presented no new evidence to confirm the existence of the possible silt layer extending north and west to the ravines. The County proposes that the last sentence be modified to read as follows:
"Although the Town has suggested that one silt layer may exist within several feet of the bottom of the landfill liner, extending in the direction of the ravines generally north and west of the site, no evidence has been presented by the Town based on its on-site investigation that supports the existence of any such layer."
I see no basis to modify the original finding. While no new evidence was introduced about the alleged silt layers during the recent hearing, they were the subject of extensive testimony at the 1996 hearing, as noted on pages 100-102 of my prior hearing report. That testimony is still part of the record and I see no reason to change my prior discussion of it. Instead, I repeat my prior statement that even if there is a western silt seam as extensive as the Town suggests, it does not warrant denying the groundwater separation variance or requiring further investigation. As Mr. Becker testified at the initial hearing, there is no ideal landfill site comprised entirely of clay and no significant glaciolacustrine deposit that does not have some proportion of silt layers. A silt layer one to three inches thick, located three feet below the pore pressure relief system, is not significant with respect to contaminant transport.
FINDINGS OF FACT
I hereby reaffirm the findings of fact contained in my initial hearing report. Those findings and the discussion of the groundwater separation variance in that report are incorporated by reference to this document.
The failure to maintain a five-foot separation between the based of the constructed liner system and the seasonal high groundwater table will not have a significant adverse impact on the public health, safety, or welfare, the environment or natural resources. A variance from this requirement would be consistent with the provisions of the ECL and the performance expected from the application of 6 NYCRR Part 360.
The groundwater separation variance should be granted and the final Department permit reissued to the Applicant, Saratoga County.
Administrative Law Judge