NCD079044426_19980514_General Electric Co. Shepherd Farm_FRBCERCLA RD_30 Remedial Design-OCR

UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
REGION 4
61 Forsyth Street
Atlanta, Georgia 30303-3104
4WD-NSRB
Michael J. Bush, Manager
Environmental, Health, and Safety
GE Lighting Systems, Inc
301 O Spartanburg Highway
Hendersonville, NC 28792
SUBJ: GE/Shepherd Farm NPL Site
East Flat Rock, NC
Dear Mr. Bush:
May 14, 1998
RECEIVED
MAY 181998
SUPERFUND SECTIO~J
Enclosed are additional Agency comments on the response to 'the 30% RD
comments dated April 1998. Comments are also included on the Intermediate Design
Investigation Work Plan.
If you have any questions, please give me a call at 404/562-8824.
Sincerely,
~
iezelle S. Bennett
Remedial Project Manager
cc: Dave Mattison, NC DENR
Todd Hagemeyer, HSI Geotrans
Lynn France, COM
.,, •
Review of Response to Comments on Preliminary Design Report
Response to
Comments Page/Item Number
36/1
36/2
37/2
37/5
37/6
Discussion
I agree with most of this response; however, EPA considers any vertical
spreading of the plume in concentrations_ above performance standards as
a significant process.
The focus of the natural attenuation analysis is on the chlorinated
compounds. However, the presence of other constituents in ground water
may negate a ground-water remedial strategy based, in part, on natural
attenuation of organic contaminants. A Remedial Design Report must
define the remedial components required to address all the contaminants of
concern (COCs) . Since these contaminants include several metals which
are unaffected by mass removal processes, any EPA review or approval of
a natural attenuation component to a remedial action would have to
include a review of the natural attenuation of metals of concern, or any
otJ1er COCs. The Record of Decision (ROD) specifies the manner in
which tJ1e ground-water COCs were to be addressed; if monitored natural
attenuation is only deemed a viable remedial action for a subset of tJ1c
COCs, the ROD-specified ground-water remedy would not change.
This response appears adequate; however, the referenced Treatability
Study Report and laboratory microcosm studies will have to be reviewed
and commented upon before a remc<lial design process can be completed.
It is unclear whether or not the process initiated in the 30% Remedial
Design Report (contemplated combination natural attenuation and ROD-
specified remedial action) should be advanced to an intermediate design
stage without the complete documentation necessary for the EPA to
support (if supportable) tJ1e revised ground-water remedy.
There remains a concern about the potential inability of natural
attenuation to completely arrest contaminant migration into deeper parts
of the bedrock. EPA would not choose a remedy which allows for
continued dispersal of contaminants into uncontaminated areas without a
high degree of assurance that any such migration was limited in extent and
duration to, a degree which would be comparable to that attained by tJ1e
remedial alternative specified in the ROD.
The response is presented in tJiree parts, which arc herein addressed as
points a, b, and c: (a) It is unclear from the referenced page 7 response to
comment 25 what evidence there is regarding ground-water flow tJ1rough
the saprolite. Note that on page H-8 of the Preliminary Design Report,
tJ1crc is specific reference made to the presence of relict fractures. (b) The
slope.of the line may not change appreciably by excluding the WW-61
point, yet inclusion of that point in the analysis is technically incorrect. (c)
No comment needed.
-1-
\
, •
Response to
Comments Page/Item Number
31n
37/8
37/10
Discussion
While the referenced methcxl for pre<:licting a decay rate is understocxl to
be a screening-level analysis, the site-specific available data suggest that
the results from this methcxl over-predict the contaminant degradation
rate. (Analysis includccl in my memorandum; comment response to point
8, pages 37 and 38). The Buschek and Alcantar methcxl for estimating a
degradation rate is subject to specific limitations to the use of tl1c methcxl
(steady, one-dimensional transport). It is unclear from the data presented
in the Preliminary Design Report how closely the GE Subsitc matches the
site conditions assumed in the mcxlel. One may further observe from the
response to comment 25 (response on pages 8 and 9) that there is
apparently no source area for the relatively high PCE concentrations
observed at well MW-14. Given this information, one could readily make
a case for a much flatter slope to the line than that shown on Figure H.2-
14, if there is limitccl lateral plume dispersion and the centerline of the
plume is not appropriately represented by the data from MW-16.
I concur that tl1e mcthcxl used in my analysis of PCE degradation rates
presented a somewhat unrealistically high estimate of contaminant half
life, and a low estimate of the degradation rate. However, the 0.6y·1
contaminant degradation rate over-prcclicts the loss of contaminant mass,
based on site-specific concentration data. llms, based on site-specific
concentration data, one would reasonably conclude that the loss of
contaminant mass is, for at least some parts of the site, not well predicted
by the 0.6y ·1.dcgradation rate estimatccl using the Buschck ancl Alcantar
mcthcxl, as presented in the Preliminary Design Report.
I agree that the BIOSCREEN mcxlel provides some useful insight into the
contaminant transport process, and concur that contaminant half-life
values I used for input to the BIOSCREEN mcxlel arc a likely over-
estimate of the half life generally applicable to the GE Subsitc. However,
the concern still remains that the complexities of this site arc such that
using the mcxlel as a predictive approach for defining tl1e maximum extent
of contaniination, or for predicting future concentrations of PCE at a
specific downgradient location and at a specific time, is inappropriate.
Probably the most significant concern (in addition to the possible, if not
probable smaller value for the degradation rate constant) is that the
mcxlcling documented in the Preliminary Design Report assumes that
source area contaminant mass is released into uncontaminated ground
water, whereas there is already a substantial contaminant mass present in
the aquifer downgradicnt of the source area. One of the useful insights
one may obtain from the BIOSCREEN mcxlel concerns the approximation
of observed ground-water PCE distribution using the BIOSCREEN
mcxlel, as an check on the degradation rate constant. There arc (as would
be the case witl1 any BIOSCREEN analysis of the GE Subsitc) numerous
uncertainties inherent in evaluations of the GE Subsite I have
-2-
• •
Response to
Commer.ts Page/Item Number Discussion
37 /I 0, continued performed using this model. However, those evaluations suggest that the
degradation rate constants used for the BIOSCREEN modeling
documented in the Preliminary Design Report are not good estimates,
because they indicate that PCE would not be detected· at the
concentrations that have been observed in the MW-14 location.
37/11 Revision of certain model input is acceptable; however, the more
significant concerns about the BIOSCREEN model (and the natural
attenuation analysis in general) relate to the degradation rate constant
calculated using the Buschek and Alcantar method. Also, for the
BIOSCREEN analysis, the presence of existing PCE contamination in
downgradient parts of the plume adds a complicating factor, as noted in
my memorandum, and discussed again in the 37 /IO discussion above. I
agree that some of the concl•1sions I made in my BIOSCREEN analysis
arc based on overly conservative assumptions, but consider the principal
conclusions in that analysis as valid: for the GE Subsite, BIOSCREEN
model results should not be considered as accurate predictors of future
site conditions and BIOSCREEN results should not be used for making
conclusions to support a remedial action, because there are problems with·
the application of the BIOSCREEN moclel to the specific site conditions
documented in the Preliminary Design Report. The model is, however,
probably appropriate for evaluating the potential range of degradation rate
constants applicable to the site, and thus setting a valid bound on the
effectiveness of biodcgradation as a remedial process. i
37 /14 One can conclude from the data in the Preliminary Design Report that
natural attenuation is limiting the downgraclient migration of some
contaminants to some extent, and is reducing the concentrations of
chlorinated VOCs in most sampled areas at tl1c GE Subsitc. However, it
is still in question how effective that process alone is in plume
remediation, and how effectively natural attenuation is arresting
downgradient migration of the plume front. The most conclusive data
from tl1e Preliminary Design Report indicate that natural attenuation has
not been totally effective in limiting plume growth in the vertical direction.
37/15 Concerns about the ·BIOSCREEN modeling arc addressed in previous
sections of this memorandum.
37/16 Regardless of natural attenuation, plume stability at the downgradient
plume margin would be influenced by discharge to the stream .. As
previously noted, there are site ground-water monitoring data which
indicate the absence of complete plume stability in terms of vertical
contaminant transport. At t11e point one would expect t11c plume stability
issue to be of great concern (i.e. downgradient of MW-14), tl1e plume is at
-3-
• •
Response to
Comments Pagentem Number Discussion
37 /16, continued the ground-water discharge point, and further downgradient plume
migration is thus arrested through that process. Thus, the influence of
natural attenuation on plume spreading downgradient of MW-14 cannot
be directly evaluated through ground-water monitoring. If one defines
stability as the absence of concentration increases at shallow ground-
water monitoring locations where contamination has been detected, then
the plume at the GE Subsite can be termed stable overall. However, this
definition of stability is only one measure of plume stability.
37 /18 The additional efforts necessary to evaluate this proposed remedial
component need to be understood as requiring revisions to the modeling
analysis in the Preliminary Design, will require a more sophisticated
modeling analysis and a comparison of the natural attenuation component
to the ROD-envisioned remedial action (i.e. basically an FS-type
comparative analysis of alternatives), and may require specific additional
data collection to support tJ1e natural attenuation proposed remedy (will
require additional data for the Shepherd Farm area).
The issue of additional site-related ground-water contaminants which have
ROD-specified performance standards (such as beryllium) would have to
be addressed in order for the EPA to select a revised ground-water
remedial action incorporating the natural attenuation remedy, as
envisioned in the Preliminary Design Report. For any COCs listed in the
ROD tliat ',\'0uld not be addressed by Ilic active ground-water remedial
actions in the most contaminated part of the GE Subsite plume, natural
attenuation would have to be considered in some detail.
Finally, the EPA must be concerned about the vertical migration of PCE
or oilier chlorinated VOC contaminants, which is indicated in the
Preliminary Design Report. This migration may not have practical
consequences in terms of potential threats to receptors, but docs have
implications in tenns of remedial time frames and selection of remedy.
The ability of Ilic natural attenuation component of the remedy to address
Ill.is vertical nligration in the downgradient parts of the plume, relative to
the ROD-specified ground-water remedial action, and the significance of
this vertical transport process, would have to be addressed in the natural
attenuation evaluation.
-4-
.
' •
Review oflntermediate Design Investigation Work Plan
In Section 3.1.1.2, in the last sentence, the proposal to not analyze samples collected below the point where total PCBs are less than 10 ppm is a possible concern. Since the sample collection will occur in a landfill, it is possible that contaminated material may be found at deeper locations, whereas if PCB sources are known to be at the surface, substantial PCB concentration declines with increasing sample depth would be anticipated. If there are existing sample data from the landfill which demonstrate the concentration of soil contamination or contamination sources are near the surface, this information should be referenced to support the proposal to limit deeper soil analyses at locations with shallower soil PCB concentrations
below 10 ppm. Also, if the visual evidence is believed to be a reliable indicator of PCB contamination in the 100 to 1000 ppm range, the reason for this conclusion needs to be presented in this section (the maximum total PCB concentrated in surface soils at landfill A was reported as 690 ppm; reference Table
C-3, Preliminary Design Report). Otherwise, it is advisable not to limit PCB analyses as proposed in Section 3.1.1.2.
Section 3.2.1.2 proposes construction and sampling of a new bedrock monitoring well near well MW-16. Although the explanation for the relatively low PCE concentration found in the MW-16 sample is plausible, the low concentration in the MW-16 sample may also be related to the location of MW-16 being off the centerline of a plume extending from the source area past MW-8 and MW-9 to MW-14. Therefore,
consideration should be given to sampling well MW-37 (according to Figure 3-5, this well was not sampled in 1994 or 1997), and, based on the results of that sampling, determine ifMW-16 or MW-37 is the better location for placement of the proposed bedrock well.
-5-
; i
,, .... .,,.
• ' • i • •
UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
REGION4
61 Forsyth Street, S.W.
Atlanta, Georgia 30303-3104 RECEIVED
May 6, 1998 MAY 151998 4WD-NSRB
Michael J. Bush, Manager
Environmental, Health, and Safety
GE Lighting Systems, Inc
. SUPERFUND SECTIO"'
3010 Spartanburg Highway
Hendersonville, NC 28792
SUBJ: GE/Shepherd Farm NPL Site
East Flat Rock, NC
Dear Mr. Bush:
Enclosed are the Agency's comments on the response to comments dated April
1998. Please provide.a written letter response to these comments to EPA no later than May 27, 1998. It is important that we reach resolution on these comments so that we can preclude these same comment<1,from being made on the upcoming 60% Remedial Design documents. Response to co'mments from one additional reviewer will be transmitted next week.
Please keep in mind that the only way to change the remedy from that presented
in the ROD is through either a ROD amendment or an Explanation of Significant
Differences (ESD). The RD cannot change the remedy.
If you have any questions, please give me a call at 404/562-8824.
cc: Dave Mattison, NC DENR
zelle S. Bennett
Remedial Project Manager
Todd Hagemeyer, HSI Geotrans
Lynn France, COM
•
RESPONSE TO COMMENTS -PRELIMINARY (30%) REMEDIAL DESIGN REPORT
General Comments
· 1. Response to General Comment 1 -Concerning natural attenuation, the response to EPA comments were generally complete and at least attempted to address EPA's questions. However, the overall technical support for intrinsic bioremediation is lacking. The authors demonstrate good support for apparent mass depletion over the three year monitoring period. However, the field site conditions do not support the known pathways for intrinsic biotransformation reactions required to get mass removal. The general lack of technical support of the conclusions comes from the reliance upon limited laboratory studies to explain the apparent loss of contaminants in the field.
Generally the site is aerobic with some limited areas of anoxic or anaerobic spaces. In addition, other field site characteiization data do not support the known conditions needed to obtain biotransformation and biodegradation of PCE to CO2• The aerobic nature of the site is largely limiting to the initial dechlorination of PCE > TCE>DCE. These reactions occur under anaerobic redox conditions approximat9ly equal to -100 to -300 mV. The PRPs state that redox conditions are indicative of conditions conducive to anaerobic dechlorination (at+ 132 to +300 mV). However, redox measurements in the field are difficult and should only be used as crude indicators of the types of reaction which might be expected. Electron acceptor and hydrogen data are better indicators of conditions for dechlorination reactions. Generally this site is high in oxygen and/or nitrate and apparently low in sulfate and non-detectable for methane. Low concentrations of ferrous iron is expected due to the high iron and nitrate. These are not conditions where one would expect to find decblorination even to DCE.
Initial dechlorination of chlorinate~:e.thenes normally occurs urider minimally iron reducing conditioris (Eh < -50 mV) and more likely under sulfate reducing conditions (Eh< -200 mV). Even under these low redox conditions, DCE would be expected to persist and has been found to do so at most sites around the world. DCE would be expected to dechlorinate to VC at an Eh < 250 mV which are usually methanogenic conditions. The GE/ Shepherd
Farm site shows no evidence of these redox conditions even in the lowest DO areas. There was no evidence of any significant iron reducing activity in any of the wells monitored. Given the general lack of site conditions conducive to dechlorination and the high concentration of PCE to daughter products TCE and DCE, plus the general lack of VC, the authors proposed that DCE and VC are aerobically degrading due to site conditions. They support this hypothesis with limited laboratory studies on VC and another more recent study on aerobic biodegradation of DCE. The former study indicates iron reducing bacteria are capable of oxidizing VC and other studies have shown VC to be aerobically degradable. The conditions for this process are generally known and somewhat accepted (there is still some reservation of how these processes act in the field). However, the one laboratory study cited (AFCEE Manual) by the authors to support the lack of DCE found at the site is weak because the reference is only an abstract of a laboratory study. The PRPs must develop stronger arguments for both the anaerobic and aerobic biodegradation than the disappearance of DCE.
-1-
• f • < .. \
\
2. Response to General Comment 5 -The hydraulic conductivity for the asphaltic layer used in the HELP model was determined through iteration. Please provide some backup information demonstrating that this iterative hydraulic conductivity is realistic and feasible for asphalt.
3. Response to General Comment 6 -The PRPs state that low DO in MW-9 (1.65 mg/L), MW-12 (0.66 mg/L) and MW-14(0.82 mg/L) are supportive of reductive dechlorination to PCE> DCE. However, one well is aerobic and all wells have significant nitrate indicating nitrate reducing conditions. Any dechlorination would be expected to be very limited and likely to generate only TCE.
The measured redox potential quoted from the AFCEE Technical Protocol is likely an oversight as it should be a negative number (-200 to -300 mV). We agree with the PRPs that field conditions would indicate biodegradation conditions to be sub-optimal.
4. Response to General Comment 7 -Provided the additional groundwater modeling and risk assessments proposed include a conservative evaluation of the present and future impacts of flushing of the contaminants into the creek on the ecological community in the creek at both subsites, this response is acceptable. The PRPs should note, however, that the tact that PCE was not measured in the stream samples collected does not mean that PCE is not discharging into the creek.
5. Response to General Comment 8 -This response is acceptable as long as the proposed additional groundwater modeling is conservative enough to reasonably ensure
full capture of the central plume with the revised extraction well system design.
Specific Comments
1. Responses to Specific Comme'nis 14 & 195 -The responses to specific comments 14 and 195 state that the laboratory data was generated using a Laboratory Information Management System (LIMS). However, the concern is that the data tables used in the report are not being generated with the use of the LIMS, therefore allowing tor the possibility of transcription errors. Furthermore, please provide a copy of the data validation reports in order to provide some assurance as to the quality and validity of the data being submitted.
2. Response to Specific Comment 16 -The response to specific comment 16 does not mention whether total polychlorinated biphenyls (PCBs) were added to the list of target parameters, as mentioned in the fifth sentence of the first paragraph of Section 2.1. Please correct this oversight.
3. Response to Specific Comment 18 -The response to specific comment 18 does not
provide the-depth at which soil samples were collected from Landfill A. -Please correct this oversight. ·
4. Response to Specific Comment 25 -The response to specific comment 25 included the submission of Figure 1 in Attachment C. Figure 1 attempts to demonstrate that the contaminant plume, as evidenced by tetrachloroethylene (PCE) concentrations, is
-2-
•
migrating laterally or slightly downward at the original source area and upward at monitoring well MW-14. Figure 1 depicts the portion of the plume that exhibits
concentrations of PCE grealer than 200 parts per billion (ppb) as arbitrarily migrating to a depth of 2,120 feet (ft), referenced to mean sea level (msl). However, Figure 1 also depicts the groundwater contamination as migrating to a depth of 2085 ft, msl, at recovery well RW-3. At this depth, it appears that the plume is migrating laterally towards MW-14, and has the potential to migrate beneath Bat Fork Creek. This section should be revised to more accurately reflect the actual subsurface conditions, including the true depth to which contamination exists and the actual groundwater flow characteristics.
5. Respo{lse to Specific Comment 27 -The response to specific comment 27 states that a revised Figure 2-17 was included in Attachment B of ·this submittal. However, the revised figure did not include the highlighted points where total PCBs in surficial soils exceed the performance goal highlighted. Please correct this oversight.
6. Response to Specific Comment 32 -The error in reported PCB concentrations at several points in the northeast corner of Figure 2-18 needs to be corrected. The revised Figure 2-17 has the corrected values. ·
7. Responses to Specific Comments 43 & 44 -The responses to specific comments 43 and 44 state that according to water-level data, ·groundwater does not flow across Spartanburg Highway or Tabor Road and no contamination was detected in any of the off-site wells that were sampled in July 1997 at the GE subsite. However, previous groundwater sampling events and contaminant flow modeling, as shown on Figure 2-26, that was conducted by GE indicates that contaminants have migrated west of Spartanburg Highway and north of Tabor Road. Furthermore, this figure indicates the potential for contaminant migration to the east and south of Bat Fork Creek. Although of/site wells did not exhibit contamination above detection limits during the July 1997 sampling event, the relative sparseness of data along';th·e perimeter raises many questions·. These issues must be addressed prior to implementation of the groundwater remediation system. Furthermore, a considerable effort must be made towards contaminant delineation before any regulatory consideration of monitored natural attenuation, even as a remedial supplement, will be made.
The fact remains that GE has assumed responsibility for the contamination that was found off-site. GE connected many homes to city water and is providing bottled water to other families. Now, this RD seems to contradict the previous findings and off-site data. Please correct this discrepancy.
8. Response to Specific Comment 50 -This comment concerns Section 2.7.4.2, not 2.7.4.3 The sentence that needs to be deleted is "As a result ... Shepherd Farm Subsite."
9. Response to Specific Comment 53 -The text needs to indicate that this statement is true for this round of sampling only, not historical.
10. Response to Specific Comment 60 -The response to specific comment 60 did not
include the revision of Figure 2-21 to clarify the purpose of this figure nor to provide for the results of vinyl chloride analysis. Please correct this oversight.
-3-
•
11. Responses to Specific Comments 65, 95 & 119 -The responses to specific comments 65, 95 and 119 state that Subsurface Detection Investigations, Inc. (SOIi) has determined, based on the ground penetrating radar (GPR) data, that the bottom of the debris was detected on all survey lines and that as a result, the depth of penetration was sufficient. The only GPR data included in the Geophysical Investigation report submitted by SOIi was collected from two transect lines conducted in a portion of Landfill A. Please provide the additional GPR data collected from Landfill B, as well as the remainder-of.the data from Landfill A, which was used to support SOIi's conclusions.
Furthermore, the response to specific comment 119 did not adequately address why Section 5.2 states that only one foot of soil is to be dug for the Landfill B excavation. Laboratory analytical data from subsurface sampling indicates that PCB contamination
exists to a minimum of 3 ft below land surface (bis) and GPR data indicates that the landfill has a total depth of 5-7 ft bis. Although it is recognized that General Electric (GE) has chosen to use the subsurface sampling results as the basis of their design, GE _should recognize that confirmatory sampling must be implemented to determine the adequacy of their excavation efforts. · However, even with the use of previously obtained subsurface sampling ·results, a minimum of 3 ft of soil should be used as the basis of the GE design for the Landfill B excavation. Please correct these oversights.
12. Responses to Specific Comments 72, 102, 103 & 224 -The response to specific comment 72 states that the text will be revised. However, this statement should remain in the text because PCE concentrations have increased in monitoring well MW-12B and monitoring well MW-14B. GE has offered very little information that can substantiate its claims that PCE i~ being degraded and is not migrating downward. GE has spent considerable effort in demonstrating that an upward vertical gradient exists at the GE subsite. However, GE has failed to address the downward migration of PCE contamination. GE has stated th.lit the operation of the Accelerated Groundwater Remediation System (AGRS) shall prevent the further downward contaminant migration
and shall actually create an upward gradient in which to extract the contaminated
groundwater from the bedrock. However, as stated in the response to specific comment 170, the operation of the AGRS has not generated a significant upward gradient. . Furthermore, the proposed extraction wells are to be _set at approximately the same depth as the current extraction wells, a depth which was cited as a possible source of the problem extracting contaminated groundwater from the bedrock aquifer. Please provide justification for setting the wells at this depth in light of the GE statement that the problem with contaminant extraction may lie in the fact that the current extraction wells are partially set in the bedrock. The resolution of these items, including how GE proposes to delineate and remediate the contaminated bedrock aquifer, must be addressed in the 60% remedial design report.
13. Response to Specific Comment 73 -The response to specific comment 73 states that
the text will be revised. Please add the actual detection limit used in determining the presence or absence of volatile organic compounds (VOCs) other than PCE.
14. Response to Specific Comment 74 -This comment response is a statement of opinion.
Please provide the necessary evidence in order to substantiate these claims.
-4-
•
15. Response to Specific Comment 81 -The use of the Buscheck and Alcantar method of approximating a first order biodegradation term.is acceptable for PCE. However, the biodegradation of PCE does not constitute "natural attenuation" or even intrinsic biodegradation. Conversion of PCE to TCE is strictly a biotransformation and the rate term approximates this conversion only. The method is not reliable for calculating metabolite degradation since they can be "formed and degraded simultaneously". However, if TCE and DCE are the end products this would not be the case.
It is agreed that the method can be used to assist in documenting the loss of the parent compound. It is also agreed that mineralization is not a requirement for the proper application of this term. However, the "loss of contamination" and loss of PCE should be viewed separately as they are distinctly different:
16. Response to Specific Comment 87 -The response to specific comment 87 addresses those reasons why PCE should not migrate downward. However, the response does not address why PCE contaminant levels have increased from 1994 to 1997 in monitoring well MW-12°8 and monitoring well MW-148. Please revise this response to account for this trend.
17. Response to Specific Comment 92 -The r,esponse to specific comment 92 refers the reader to general response no. 1. Specific comment 92 requests that the last sentence of the first paragraph of Section 4.1 be deleted. However, general response no. 1 does not address this section. Please delete the sentence as requested.
18. Response to Spe.cific Comment 93 -The response to specific comment 93 refers the reader to general response no. 1 .. Specific comment 93 requests that the second to last sentence of the last paragraph otSlelction 4.1 be deleted. However, general response no. 1 does riot address this section. Please delete the sentence as requested. ·
19. Response to Specific Comment 108 -The response to specific comment 108 indicates that the contingency plan was conceptual in nature and that if the plan is conceptually agreeable to the United States Environmental Protection Agency (US EPA), a more . detailed approach will be provided in the 60% report. However, as previously stated in the comment, there is insufficient detail to make any determination regarding the acceptability of this proposal. Please provide greater detail to resolve such questions as presented in specific comment 108.
20. Response to Specific Comment 117 -The. response to specific comment 117 indicates that the conclusions regarding the plume delineation were based on residential well and test well information. Although it is recognized that GE shall further investigate the subsurface conditions at the Shepherd Farm subsite, GE must recognize that the US EPA comments regarding the Shepherd Farm plume delineation shall remain until further substantiated evidence has been presented.
21. Response to Specific Comment 127 -Table 6-4 should be revised to correct the typographical error (a revised table was not submitted). The list of wells in Section 6.5.2 contains two discrepancies with the list provided with the response (MW-2.0A versus MW-
-5-
•
20B, and MW-42 versus MW54). The discrepancies should be corrected.
22. Response to Specific Comment 133 -There are big gaps between MW-3 and MW-27, and MW-3 and MW-25 where contaminated groundwater could discharge into Bat Fork Creek. COM Federal believes that if monitored natural attenuation of the "fringe area" is to be selected as the remedy, these gaps should be monitored with additional monitoring wells to evaluate the effectiveness of this remedy.
23. Responses to Specific Comments 133 & 134 -The responses to specific comments 133 and 134 are inadequate. The reason that the comments were made was to point out the lack of sufficient plume data at-the perimeter. Furthermore, since contamination currently exists at these wells, GE must propose a monitoring well network that will serve to provide adequate warning in the event that contaminants further migrate offsite. GE must address the issues raised in the above comments regarding plume delineation and then propose a much more stringent monitoring network.
24. Response to Specific Comment 135 -The response to specific comment 135 indicates
that GE wishes to conduct semivolatile organic compound (SVOC) analysis only on an annual basis. However, until further investigation and compliance monitoring further substantiate GE's claims, SVOC sampling shall be conducted in accordance with the remainder of the other parameters (quarterly for the first three years, semiannual for the following two years, and annually thereafter).
25. Response to Specific Comment 138 -The response to specific comment 138 provides a general description of the air emission analysis conducted to determine that an air emission permit will not be required. The calculations, including all formulae and methodologies, used to determine .this fact should be included as a separate appendix. In addition, the ROD calls for "grany,tated activated carbon adsorption to treat the vapor effluent or off-gas to remove the contaminants stripped from the groundwater prior to being discharged to the atmosphere."
26. Response to Specific Comment 140 -The response to specific comment 140 simply addresses GE's concerns regarding the need for building a pipeline to connect the GE and Shepherd Farm groundwater remediation systems. However·, the response fails to address the fact that the Record of Decision (ROD) requires the installation of an active groundwater remediation system at the Shepherd Farm subsite, that the system will not impact any downstream wetlands, that the system will have minimal impacts on Bat Fork Creek and that the system will have minimal impacts to the residents of the Spring Haven community. Furthermore, in light of the many modifications to the ROD that GE has proposed, the addition of an onsite groundwater remediation system to the Shepherd
Farm subsite, if warranted, may be a plausible request. However, GE must bear in mind that the system must be managed in such a manner as to have a minimal impact on the Spring Haven community.
Is not the Shepherd Farm subsite part of the site, so that a separate treatment system built on this subsite would still be considered "on-site"?
-6-
27. Response to Specific Comment 146 -The response to specific comment 146 does not address whether drinking water well SWW-10 will be added to the compljance well network or whether the groundwater sampling schedule for the Shepherd Farm subsite has been modified to be the same as that for the GE subsite. Please make the appropriate revisions.
28. Response to Specific Comment 156 -The response to specific comment 156 does not address whether monitoring well MW-42, as identified in the last sentence of Section 6.5.2, or monitoring well MW-54, as identified in Figure 2-25, is to be used as a performance monitoring well. Please clarify this discrepancy.
29. Response to Specific Comment 160 -The response to specific comment 160 states that although contaminants have migrated downward at monitoring well MW-12B and monitoring well MW-14B, a mixed and upward vertical gradient exists at monitoring well MW-12B and monitoring well MW-14B, respectively. However, the response fails to provide an explanation for these events. Please provide an explanation for the downward contaminant migration although mixed/upward gradients exist at the site.
30. Response to Specific Comment 161 -The response to specific comment 161 states that the ROD specifies that the "cap will be a composite liner and shall consist of clay, a flexible membrane liner, and if necessary, a drainage layer." However, GE should recognize that no liner system shall be approved which does not meet the minimum performance requirements of a Resource Conservation and Recovery Act (RCRA) type liner system.
31. Response to Spe.cific Comment 170 -The response to specific comment 170 states fails to address how GE plans to address the contaminated bedrock aquifer considering that the naturally occurring and AGR~Hriduced upward gradient has failed to prevent PCE contamination from migrating further downward. Please revise this response and the appropriate sections in the remedial design accordingly.
32. Response to Specific Comment 185 -The response to specific comment 185 indicates that the lowest possible quantification limits were used. The quantification limits used for the Shepherd Farm investigation are one order of magnitude higher than the minimum Contract Laboratory Program (CLP) limits. Although the soil samples near Drum 1 and Drum 2 exhibited some contamination, the levels were not extraordinarily high. In the case of the soil sample collected near monitoring well MW-62, no analytes were detected at all therefore not requiring any sample dilution whatsoever. Please provide justification for not employing lower quantification limits in these analyses.
33. Response to Specific Comment 191 -The response to specific comment 191 failed to present the results of the headspace analysis conducted on saprolite soils. Please correct this oversight.
34. Response to Specific Comment 198 -The response to specific comment 198 describes the installation of monitoring well MW-62A. It is recognized that at the time of the outer casing installation, the existence of the fracture at 50.5 ft bis was presumably not known. However, upon the discovery of the fracture zone 48 hours later when drilling the second,
-7-
•
smaller borehole, provisions should have been made for assessing the 3.5-ft thick fracture
zone. Provisions should have been made to drill another well in order to assess this
fracture zone. GE states that the fracture in question would not provide a more
representative groundwater sample from the bedrock aquifer than the screened interval.
However, GE should recognize that GE must assess all potential pathways for
· contaminant migration, whether the migration occurs in the "bedrock aquifer" or in bedrock fractures, which dominate groundwater flow conditions in bedrock. GE should include
provisions for attempting to monitor this, and au other, fracture zones, in their efforts to delineate the plume and the subsurface conditions that exist at the GE and Shepherd Farm subsites.
35. Response to Specific Comment 222 -The figure should be corrected. While some degradation pathways can be drawn for either abiotic or biotic reactions, the effectiveness
of these mechanisms is dramatically different. For example, 1, 1-DCE is a major abiotic
product of 1, 1, 1-TCA and is considered the product of this reaction when found in the
field. However, it has been reported as a very minor product of biological dechlorination of TCE and is usually not depicted in this pathway. The other pathways involving aerobic
versus anaerobic reactions also result in different products.
36. Response to Specific Comment 225 -There is limited evidence for the aerobic
biodegradation of c-DCE (two lab studies and no field evidence known• to this reviewer).
Given the limited information for this important pathway, and the importance the PRPs are placing on this (unreferenced) paper, it would seem appropriate to support the hypothesis with other known conditions which cause, stimulate, induce or in some way control the aerobic biodegradation of DCE at this or any other site where DCE is used aerobically as a primary substrate. It is not known to occur widely at many sites. In addition, anaerobic
degradation is noi likely to proceed to carbon dioxide as stated, but to VC which was not
reported. '/ ·.
37. Response to Specific Comment 226 -Bradely and Chapelle, 1997, refers to the iron-reducing capability for degrading VC and DCE. The reference documents the anaerobic biodegradation of DCE under methanogenic and iron-reducing conditions. The intermediates of both conditions are VC, ethene, and ethane, and finally results in carbon
dioxide. However, iron reducing conditions will catalyze VC degradation directly to carbon dioxide. The best reference for the process described by the PRPs is Bradley and Chapelle, Env. Science & Technology, 1998. These authors document the aerobic
biodegradation of DCE to carbon dioxide in stream-bed microcosms. Although no organic
carbon ·analyses of the sediments were conducted, the sediments were aerobic (2 mg/L
DO) probably indicating insignificant biodegradable organic carbon. This study does support the hypothesis of the authors although there has been no actual field demonstration of this technology.
The PRPs should support their hypothesis by documenting the aerobic biodegradation of
DCE potential as a primary substrate from soils and groundwater at the GE site.
The response to specific comment 226 defends the statement that the overall decrease in PCE and trichloroethylene (TCE) mass is 20% or greater. However, the accuracy of this statement still remains given the overall precision of these estimates, especially
considering that the estimates place the increase in dichloroethene (DCE) mass
-8-
• \
somewhere between 25-60%. Please amend this statement to include a reference to the overall precision and accuracy of these estimates.
38. Specific Comment 228 -The AFCEE Technical Protocol for Natural Attenuation of Chlorinated Aliphatic Hydrocarbons in Groundwater (1997) indicates that negative Eh would be required for reductive dechlorination.
-9-
JAMES B. HUNT JR.
GoVE.RNOR
'J'.':',~E MCOEVI.TT,
, • SECRETARY ,'., .• : .
i;'tiilt:;f
.. WILLIAM l!.'MEYER ;--~-•.:_ -.
i;.•
•
Ms. Giezelle Bennett
Superfund Branch
Waste Management Division
@
NORTH CAROLINA DEPARTMENT OF
ENVIRONMENT AND NATURAL RESOURCES
DIVISION OF WASTE MANAGEMENT
May4, 1998
United States Environmental Protection Agency
Region IV
61 Forsyth Street, 11 lh Floor
Atlanta, Georgia 30303
RE: Response to Comments -
Preliminary (30%) Remedial Design Report
General Electric/Shepherd Farm NPL Site
East Flat Rock, Henderson County
Dear Ms. Bennett:
The Response to Comments -Preliminary (30%) Remedial Design Report for the
General Electric/Shepherd Farm National Priorities List (NPL) Site was received by
the Superfund Section of the North Carolina Department of Environment and Natural
Resources (NC DENR). The Superfund Section has reviewed this document and
offers the following attached comments.
We appreciate the opportunity to comment on this document. If you have any
questions, please feel free to call me at [phone], extension 349.
Attachment
Sincerely,·
'7)J ;&;/~ #,~.~
David B. Mattison, CHMM
Environmental Engineer
Superfund Section
401 OBERLIN ROA.O, SUITE 150, RALEIGH, NC 27605
PHONE [phone] FAX 919-71 S-3605
AN EQUAL OPPORTUNITY/ AFFIRMATIVE ACTION EMPLOYER· 50% RECYCLE0/10% POST-CONSUMER PAPER
•
Ms. Giezelle Bennett
May 4, 1998
Page 1
RESPONSE TO COMMENTS -PRELIMINARY (30%) REMEDIAL DESIGN REPORT
Specific Comments and Responses
Responses to Specific Comments 14 & 195
1. The responses to specific comments 14 and 195 state that the laboratory data was generated
using a Laboratory Information Management System (LIMS). However, the concern is that
the data tables used in the report are not being generated with the use of the LIMS, therefore
allowing for the possibility of transcription errors. Furthermore, please provide a copy of the
data validation reports in order to provide some assurance as to the quality and validity of the
data being submitted.
Response to Specific Comment 16
2. The response to specific comment 16 does not mention whether total polychlorinated
biphenyls (PCBs) were added to the list of target parameters, as mentioned in the fifth
sentence of the first paragraph of Section 2.1. Please correct this oversight.
Response to Specific Comment 18
3. The response to specific comment 18 does not provide the depth at which soil samples were
collected from Landfill A. Please correct this oversight.
Response to Specific Comment 25
I
4. The response to specific comment 25 included the submission of Figure 1 in Attachment C.
Figure 1 attempts to demonstrate that the contaminant plume, as evidenced by
tetrachloroethylene (PCE) concentrations, is migrating laterally or slightly downward at the
original source area and upward at monitoring well MW-14. Figure 1 depicts the portion of
the plume that exhibits concentrations of PCE greater than 200 parts per billion (ppb) as
arbitrarily migrating to a depth of 2,120 feet (ft), referenced to mean sea level (msl).
However, Figure 1 also depicts the groundwater contamination as migrating to a depth of
2085 ft, ms!, at recovery well RW-3. At this depth, it appears that the plume is migrating
laterally towards MW-14, and has the potential to migrate beneath Bat Fork Creek. This
section should be revised to more accurately reflect the actual subsurface conditions,
including the true depth to which contamination exists and the actual groundwater flow
characteristics.
Ms. Giezelle Bennett
May 4, 1998
Page 2
•
Response to Specific Comment 27
5. The response to specific comment 27 states that a revised Figure 2-17 was included in
Attachment B of this submittal. However, the revised figure did not include the highlighted
points where total PCBs in surficial soils exceed the performance goal highlighted. Please
correct this oversight.
Responses to Specific Comments 43 & 44
6. The responses to specific comments 43 and 44 state that according to water-level data,
groundwater does not flow across Spartanburg Highway or Tabor Road and no
contamination was detected in any of the off-site wells that were sampled in July 1997 at the
GE subsite. However, previous groundwater sampling events and contaminant flow
modeling, as shown on Figure 2-26, that was conducted by GE indicates that contaminants
have migrated west of Spartanburg Highway and north of Tabor Road. Furthermore, this
figure indicates the potential for contaminant migration to the east and south of Bat Fork
Creek. Although offsite wells did not exhibit contamination above detection limits during the
July 1997 sampling event, the relative sparseness of data along the perimeter raises many
questions. These issues must be addressed prior to implementation of the groundwater
remediation system. Furthermore, a considerable effort must be made towards contaminant
delineation before any regulatory consideration of monitored natural attenuation, even as a
remedial supplement, will be made.
Response to Specific Comment 60
7. The response to specific comment 60 did not include the revision of Figure 2-21 to clarify the
purpose of this figure nor to provide for the results of vinyl chloride analysis. Please correct
this oversight.
Responses to Specific Comments 65, 95 & 119
8. The responses to specific comments 65, 95 and I I 9 state that Subsurface Detection
Investigations, Inc. (SDII) has determined, based on the ground penetrating radar (GPR)
data, that the bottom of the debris was detected on all survey lines and that as a result, the
depth of penetration was sufficient. The only GPR data included in the Geophysical
Investigation report submitted by SDII was collected from two transect lines conducted in
a portion of Landfill A. Please provide the additional GPR data collected from Landfill B, as
well as the remainder of the data from Landfill A. which was used to support SDII's
conclusions.
•
Ms. Giezelle Bennett
May 4, 1998
Page 3
Furthermore, the response to specific comment 119 did not adequately address why Section
5.2 states that only one foot of soil is to be dug for the Landfill B excavation. Laboratory
analytical data from subsurface sampling indicates that PCB contamination exists to a
minimum of 3 ft below land surface (bis) and GPR data indicates that the landfill has a total
depth of 5-7 ft bis. Although it is recognized that General Electric ( GE) has chosen to use
the subsurface sampling results as the basis of their design, GE should recognize that
confirmatory sampling must be implemented to determine the adequacy of their excavation
efforts. However, even with the use of previously obtained subsurface sampling results, a
minimum of 3 ft of soil should be used as the basis of the GE design for the Landfill B
excavation. Please correct these oversights.
Responses to Specific Comments 72, 102, 103 & 224
9. The response to specific comment 72 states that the text will be revised. However, this
statement should remain in the text because PCE concentrations have increased in monitoring
well MW-12B and monitoring well MW-14B. GE has offered very little information that can
substantiate its claims that PCE is being degraded and is not migrating downward. GE has
spent considerable effort in demonstrating that an upward vertical gradient exists at the GE
subsite. However, GE has failed to address the downward migration of PCE contamination.
GE has stated that the operation of the Accelerated Groundwater Remediation System
(AGRS) shall prevent the further downward contaminant migration and shall actually create
an upward gradient in which to extract the contaminated groundwater from the bedrock.
However, as stated in the response to specific comment 170, the operation of the AGRS has
not generated a significant upward gradient. Furthermore, the proposed extraction wells are
to be set at approximately the same depth as the current extraction wells, a depth which was
cited as a possible source of the problem extracting contaminated groundwater from the
bedrock aquifer. Please provide justification for setting the wells at this depth in light of the
GE statement that the problem with contaminant extraction may lie in the fact that the current
extraction wells are partially set in the bedrock. The resolution of these items, including how
GE proposes to delineate and remediate the contaminated bedrock aquifer, must be addressed
in the 60% remedial design report.
Response to Specific Comment 73
I 0. The response to specific comment 73 states that the text will be revised. Please add the actual
detection limit used in determining the presence or absence of volatile organic compounds
(VOCs) other than PCE.
Ms. Giezelle Bennett
May4, 1998
Page4
•
Response to Specific Comment 74
11. The response to specific comment 74 is a statement of opinion. Please provide the necessary
evidence in order to substantiate these claims.
Response to Specific Comment 87
12. The response to specific comment 87 addresses those reasons why PCE should not migrate
downward. However, the response does not address why PCE contaminant levels have
increased from 1994 to 1997 in monitoring well MW-12B and monitoring well MW-14B.
Please revise this response to account for this trend.
Response to Specific Comment 92
13. The response to specific comment 92 refers the reader to general response no. I. Specific
comment 92 requests that the last sentence of the first paragraph of Section 4.1 be deleted.
However, general response no. 1 does not address this section. Please delete the sentence as
requested.
Response to Specific Comment 93
14. The response to specific comment 93 refers the reader to general response no. I. Specific
comment 93 requests that the second to last sentence of the last paragraph of Section 4. 1 be
deleted. However, general response no. 1 does not address this section. Please delete the
sentence as requested.
Response to Specific Comment 108
15. The response to specific comment 108 indicates that the contingency plan was conceptual in
nature and that if the plan is conceptually agreeable to the United States Environmental
Protection Agency (US EPA), a more detailed approach will be provided in the 60% report.
However, as previously stated in the comment, there is insufficient detail to make any
determination regarding the acceptability of this proposal. Please provide greater detail to
resolve such questions as presented in specific comment 108.
Response to Specific Comment 117
16. The response to specific comment 117 indicates that the conclusions regarding the plume
delineation were based on residential well and test well information. Although it is recognized
that GE shall further investigate the subsurface conditions at the Shepherd Farm subsite, GE
must recognize that the US EPA comments regarding the Shepherd Farm plume delineation
shall remain until further substantiated evidence has been presented.
Ms. Giezelle Bennett
May 4, 1998
Page 5
•
Responses to Specific Comments 133 & 134
17. The responses to specific comments 133 and 134 are inadequate. The reason that the
comments were made was to point out the lack of sufficient plume data at the perimeter.
Furthermore, since contamination currently exists at these wells, GE must propose a
monitoring well network that will serve to provide adequate warning in the event that
contaminants further migrate offsite. GE must address the issues raised in the above
comments regarding plume delineation and then propose a much more stringent monitoring
network.
Response to Specific Comment 135
18. The response to specific comment 135 indicates that GE wishes to conduct semivolatile
organic compound (SVOC) analysis only on an annual basis. However, until further
investigation and compliance monitoring further substantiate GE's claims, SVOC sampling
shall be conducted in accordance with the remainder of the other parameters ( quarterly for
the first three years, semiannual for the following two years, and annually thereafter).
Response to Specific Comment 138
19. The response to specific comment 138 provides a general description of the air emission
analysis conducted to determine that an air emission permit will not be required. The
calculations, including all formulae and methodologies, used to determine this fact should be
included as a separate appendix.
Response to Specific Comment 140
20. The response to specific comment I 40 simply addresses GE's concerns regarding the need for
building a pipeline to connect the GE and Shepherd Farm groundwater remediation systems.
However, the response fails to address the fact that the Record of Decision (ROD) requires
the installation of an active groundwater remediation system at the Shepherd Farm subsite,
that the system will not impact any downstream wetlands, that the system will have minimal
impacts on Bat Fork Creek and that the system will have minimal impacts to the residents of
the Spring Haven community. Furthermore, in light of the many modifications to the ROD
that GE has proposed, the addition of an onsite groundwater remediation system to the
Shepherd Farm subsite, if warranted, may be a plausible request However, GE must bear in
mind that the system must be managed in such a manner as to have a minimal impact on the
Spring Haven community.
Ms. Giezelle Bennett
May 4, 1998
Page 6
•
Response to Specific Comment 146
21. The response to specific comment 146 does not address whether drinking water well SWW-
10 will be added to the compliance well network or whether the groundwater sampling
schedule for the Shepherd Farm subsite has been modified to be the same as that for the GE
subsite. Please make the appropriate revisions.
Response to Specific Comment 156
22. The response to specific comment 156 does not address whether monitoring well MW-42,
as identified in the last sentence of Section 6.5.2, or monitoring well MW-54, as identified in
Figure 2-25, is to be used as a performance monitoring well. Please clarify this discrepancy.
Response to Specific Comment 160
23. The response to specific comment 160 states that although contaminants have migrated
downward at monitoring well MW-12B and monitoring well MW-14B, a mixed and upward
vertical gradient exists at monitoring well MW-12B and monitoring well MW-14B,
respectively. However, the response fails to provide an explanation for these events. Please
provide an explanation for the downward contaminant migration although mixed/upward
gradients exist at the site.
Response to Specific Comment 161
24. The response to specific comment 161 states that the ROD specifies that the "cap will be a
composite liner and shall consist of clay, a flexible membrane liner, and if necessary, a
drainage layer." However, GE should recognize that no liner system shall be approved which
does not meet the minimum performance requirements of a Resource Conservation and
Recovery Act (RCRA) type liner system.
Response to Specific Comment 170
25. The response to specific comment 170 states fails to address how GE plans to address the
contaminated bedrock aquifer considering that the naturally occurring and AGRS-induced
upward gradient has failed to prevent PCE contamination from migrating further downward.
Please revise this response and the appropriate sections in the remedial design accordingly.
•
Ms. Giezelle Bennett
May 4, 1998
Page 7
Response to Specific Comment 185
26. The response to specific comment 185 indicates that the lowest possible quantification limits
were used. The quantification limits used for the Shepherd Farm investigation are one order
of magnitude higher than the minimum Contract Laboratory Program (CLP) limits. Although
the soil samples near Drum 1 and Drum 2 exhibited some contamination, the levels were not
extraordinarily high. In the case of the soil sample collected near monitoring well MW-62,
no analytes were detected at all therefore not requiring any sample dilution whatsoever.
Please provide justification for not employing lower quantification limits in these analyses.
Response to Specific Comment 191
27. The response to specific comment 191 failed to present the results of the headspace analysis
conducted on saprolite soils. Please correct this oversight.
Response to Specific Comment 198
28. The response to specific comment 198 describes the installation of monitoring v1ell MW-62A.
It is recognized that at the time of the outer casing installation, the existence of the fracture
at 50.5 ft bis was presumably not known. However, upon the discovery of the fracture zone
48 hours later when drilling the second, smaller borehole, provisions should have been made
for assessing the 3. 5-ft thick fracture zone. Provisions should have been made to drill another
well in order to assess this fracture zone. GE states that the fracture in question would not
provide a more representative groundwater sample from the bedrock aquifer than the
screened interval. However, GE should recognize that GE must assess all potential pathways
for contaminant migration, whether the migration occurs in the "bedrock aquifer" or in
bedrock fractures, which dominate groundwater flow conditions in bedrock. GE should
include provisions for attempting to monitor this, and all other, fracture zones, in their efforts
to delineate the plume and the subsurface conditions that exist at the GE and Shepherd Farm
subsites.
Response to Specific Comment 226
29. The response to specific comment 226 defends the statement that the overall decrease in PCE
and trichloroethylene (TCE) mass is 20% or greater. However, the accuracy of this statement
still remains given the overall precision of these estimates, especially considering that the
estimates place the increase in dichloroethene (DCE) mass somewhere between 25-60%.
Please amend this statement to include a reference to the overall precision and accuracy of
these estimates.
tt!I
GEOTRANS
A Tl:TRA Tl:Cl-t COMPANY
Ms. Giezcllc Bennett
Remedial Project Manager
U.S. EPA Region 4
I 00 Alabama Street, S. W.
Atlanta, Georgia 30303-3104
April 10, 1998
41,ouo Holcomb Bridge Road
Building 200, Suite 305
Roswell, Georgia
:rnon,
770-642-1000 FAX 770-(,42-BB0B
APR 13 1998 ' -~
Reference: Response to Comments on Preliminary Design (30%) Report
GE/Shepherd Farm Sile
HSI GcoTrans Project No. N754-007
Dear Ms. Bennett:
Enclosed please find responses to your comments dated March 11, 1998 and March 19,
1998 on the above rel'crcnccd report (dated February 7, 1998). A copy of these El'A's comments
is provided in Attachment/\. Attachment 13 contains revised ligurcs and tables. New ligurcs
and tab ks generated lo respond lo El' A comments arc provided in Attachment C. Attachment D
contains the correspondence from the U.S Army Corps of Engineers regarding the site wetlands.
The requested landfill and groundwater model information is provided digitally on lhe enclosed
CD-ROM.
EPA has requested all laboratory analysis reports, chain of custody documentation,
QA/QC documentation, data validation reports and lield notes for all sampling events. This is a
request that we arc willing lo fullill, however, with some clarification described below, EPA may
determine this activity unnecessary. EPA has performed field oversight on a majority of the
sampling events, including oversight for every type of sampling event (e.g., soil gas survey, . .
geophysics survey, soil borings, well installations, soil sampling, surface water sampling,
sediment sampling, and groundwater sampling). As pointed out in specific comment no. I 4,
submittal of the lab and validator's reports is unnecessary since the results were generated using
LIMS. The requested information consists of nearly three feet of paper (or 4,000 to 5,000
pages). The requested documents arc in our files and available for review.
There arc three major issues addressed in the enclosed responses lo comments. First, the
intermediate design will be submitted with the ROD-selected remedy, as requested by EPA. ll is
proposed lo additionally provide an allcrnalivc remedy in a separate section of the intermediate
design report. This alternative remedy must provide performance that is equivalent lo the ROD-
selected remedy. As described in the enclosed responses, this alternative remedy includes
•
monitored natural attenuation (MNA) as a supplemental remedy to pump-and-treat. The natural
attenuation portion of the Preliminary Design Report will be included in the Trcatabilily Study
which will also incorporate the responses herein to the comments on natural attenuation.
Scientific and policy advancements have been made in the last few years since the ROD and
Consent Decree were issued that make MNA a viable new remedial method to evaluate. Further
evaluation of this potential remedy is planned, including additional groundwater modeling and a
risk assessment.
Second, a schedule modilication is requested to collect new data requested by El' A and to
evaluate a potential remedy change at Landfill A. Several new field investigations arc proposed.
These investigations include exploratory trenching at Landfill A, a Gcoprobc, or equivalent,
investigation at Shepherd Farm, the installation of two new saprolitc monitor wells at Shepherd
Farm and one new saprolite well on the cast side ofl3at Fork Creek along Tabor Road at the GE
Subsitc, and a replacement soil boring at Shepherd Farm. These new data should be incorporated
into the intermediate design. It will require at least two months to complete these activities. It is
requested that the submittal of the intermediate design be moved to June 26, 1998. We believe
this proposed schedule modification to be an aggressive and optimistic schedule considering the
length of time that will be necessary to receive EP A's approval of site characterization activities,
obtain off-site access, install wells, and allow wells to stabilize before sampling.
Third, no additional sources of groundwater contamination arc suspected. This
conclusion is based on soil gas sampling, soil sampling, and groundwater sampling that was
performed in accordance with the Statement of Work (SOW) SOW and Remedial Design Work
Plan (RDWP). The observed levels of PCE at wells MW-14 and MW-25 arc consistent with Ilic
conceptual model of the hydrogcologic system, as described in the enclosed responses. No
additional field investigations arc scheduled lo search for unknown sources of groundwater
contamination, other than the replacement soil boring at Shepherd Farm.
Please call me to discuss the enclosure. We would be happy to schedule a meeting to
discuss these issues.
cc: Michael Bush (GELS)
Charles Faust (HSI GcoTrans)
Lynne France (COM Federal)
David Mattison (NC DENR)
~-~4-t)~ F4-_
Todd Hagemeyer, P.G.
Senior Hydrogcologist
HSIGEOTRANS
•
UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
REGION 4
61 Forsyth Street, S.W.
Atlanta, Georgia 30303-3104
4WD-NSRB
Michael J. Bush, Manager
Environmental, Health, and Safety
GE Lighting Systems, Inc
3010 Spartanburg Highway
Hendersonville, NC 28792
SUBJ: GE/Shepherd Farm NPL Site
East Flat Rock, NC
Dear Mr. Bush:
March 19, 1998
RECEIVED
MAR 20 1998
SUPERPUND SECTION
Attached are additional comments on the 30% RD Report. Please include
them with the March 11, 1998 comments and respond to them accordingly.
If you have any questions, please give me a call at 404/562-8824.
cc: Dave Mattison, NC DENR
· elle S. Bennett
Remedial Project Manager
Todd Hagemeyer, HSI Geotrans
Lynn France, COM
,
I, •
UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
4WD-OTS
MEMORANDUM
REGION 4
6 I Forsyth Street
Atlanta, Georgia 30303-3104
March I 9, 1998
SUBJECT: GE/Shepherd Farm Site, East Flat Rock, North Carolina
RECEIVED
MAR 2 0 1998
SUPER FUND SECTl0''
FROM: William N, O'Steen, Environmental Scientist~
Office of Technical Services, Waste Management Division
TO: Giezelle Bennett, Remedial Project Manager
North Site Management Branch
This memorandum responds to your request for a review of the Preliminary Design (30 % Design) Report for the-GE/Shepherd Farm Site, East Flat Rock, North Carolina. This review focused on the pro"i>osed natural attenuation remedial action component to address ground-water contamination, as well as the proposed performance monitoring of the ground-
water remedial action. For your convenience, comments are referenced to specific sections or pages of this report, as applicable. A summary of my review is presented at the beginning of this memorandum.
Summary of Review
The 30% Remedial Design Report for the GE/Shepherd Farm Site proposes ground-water remedial actions for these two areas which are substantially different than the ground-water
remedial action selected in the Record of Decision for this site. The proposed remedial actions are reviewed in this memorandum, and are either found to be unacceptable as a remedial action from a purely technical perspective, or unacceptable without further data analysis and detailed comparison to the ROD-specified remedial action for ground water, again from a purely technical perspective.
At the Shepherd Farm area, the 30% Design Report proposes to use monitored natural attenuation as the ground-water remedial action. Not only is this remedial action inconsistent
•
-2-
with the Record of Decision, the available data do not indicate that natural attenuation
(biodegradation of the principal site contaminant, PCE) is occurring. Instead, available ground-
water data suggest that PCE is being transported by the ground water away from the core area of
ground-water contamination in the Shepherd Farm area.
At the GE area, monitored natural attenuation of the less contaminated part of the ground-water
plume is proposed. The 30% Design Report defines this as the fringe plume. There are several
problems with this proposal, although the possibility that monitored natural attenuation is a viable
remedial alternative for either all or part of this fringe plume area is not discounted in this
memorandum. Problems with the monitored natural attenuation proposal are summarized as
follows:
1. Analysis of PCE biodegradation rates in the 30% Design Report is overly simplistic and is
probably inaccurate. Data indicate variable PCE degradation rates apply in the GE area.
In the "core plume", an apparent PCE biodegradation rate constant of approximately
0.086, equivalent to a PCE half life of 8.06 years, is calculated. This value assumes that
the only source of PCE mass loss is from biodegradation, while other factors, including
vertical PCE migration and potential mass removal via recovery well operation, may
account for some of the apparent mass loss in the core plume area. Geochemical
conditions in this core plume area appear to be favorable for a relatively low rate of PCE
biodegradation. In the less contaminated part of the plume, apparent PCE mass loss
constants indicate an apparent PCE half life of between 3 and 4 years, as an average. This
half life is approximately three times greater than that considered in the 30% Design
Report, and indicates a greater potential for chlorinated VOC plume migration than that
assumed in the report. ·
2. The monitored natural attenuation analysis does not consider the presence of contaminants
other than PCE and its degradation products.
3. The monitored natural attenuation analysis defines a core plume and fringe plume area on
the basis of PCE, TCE, and nitrate concentrations. It is possible that this definition is
inappropriate, because of the presence of other contaminants for which concentrations
around the identified core plume-fringe plume boundary are unknown. Otherwise, the
boundaries of the core plume-fringe plume area are ill defined because of limited
monitoring data in the interior part of the fringe plume area.
4. A monitored natural attenuation remedial action cannot be accepted by the EPA without a
detailed comparison to an active remedial action (in this case, the ROD-specified remedial
action). This detailed analysis is not included in the 30% Design Report.
5. Geochemical factors in the fringe plume area generally appear favorable for
biodegradation of the leS's chlorinated organic compounds, but generally appear
unfavorable for biodegradation of more chlorinated organic compounds such as PCE.
Chlorinated VOC concentrations in the fringe plume area indicate that although PCE
biodegradation has occurred, PCE remains the primary VOC present.
The 30% Design Report includes a remedial action performance monitoring plan which is
•
-3-
inadequate. Concerns about this plan include the following:
I. It is based on the proposed combination of core plume active remedial action and fringe
plume monitored natural attenuation. This plan cannot be considered appropriate for the
GE area at this time.
2. The performance monitoring plan includes a contingency remedial action option which is
unacceptable. Specific concerns about this contingency proposal are included in this
memorandum.
3. The plan does not include adequate baseline ground-water monitoring in certain parts of
the GE area for non-VOC substances with ROD-specified performance standards.
4. The plan does not propose adequate ground-water monitoring across the area of ground-
water contamination to evaluate the remedial action's performance in attaining ROD-
specified performance standards.
30% Design Review Comments and Data Analysis
Section 1.2 of the report needs to state the Record of Decision (ROD)-specified remedial action
for ground water.
Section 1.3 of the report indicates that a monitored natural attenuation program will be designed
to address remediation of a dilute ground-water plume. This remedial action was neither specified
in the ROD, nor was it evaluated in the site Feasibility Study. As discussed in detail in this
memorandum, even if this monitored natural attenuation program was acceptable from a policy
perspective, numerous technical deficiencies in the 30% Design _Report's evaluation of monitored
natural attenuation would still need to be addressed in order for the EPA to accept this
component to the ground-water remedial action.
On page 2-14, in Section 2.7.4.2, the last two sentences imply that because the residential wells
are completed in the saprolite, contamination has not migrated into the bedrock. This statement
requires correction.
On page 2-17, the second paragraph under the heading Model Calibration needs to specify the
correct transrnissivity units.
At the bottom of page 3-8, the discussion of a nearly steady-state plume configuration makes the
assumption that this apparent steady-state plume configuration is due to natural attenuation. Data
presented in this report clearly demonstrate that natural attenuation of the plume at the GE part of
the site is occurring, including an intrinsic biodegradation component to natural attenuation of
chlorinated volatile organic compounds (VOCs). However, the observed ground-water
concentration trends at the GE part of the site may be related to other factors some of which do
not indicate steady-state plume conditions. Most specifically, the role of present recovery wells at
the facility in affecting plume dimensions and mass calculations is not defined in the report.. The
•
-4-
report does not adequately discuss the apparent increase in contaminant concentrations in deeper
monitoring intervals with respect to the presumption of a steady-state plume . Additionally, the
influence of source remedial actions on ground-water concentrations would have to be more
thoroughly evaluated in order to better define the role that ground-water natural attenuation
processes have had in reducing contaminant concentrations.
The Section 3.3.3 evaluations of natural attenuation are reviewed in some detail in the following
comments on report Appendix H.
The PCE concentration versus distance comparison (Section 2.2.2, Appendix H; Figure H.2-14)
presumes that PCE concentration changes from the source area to the downgradient discharge
zone, along the presumed center of the PCE contaminant plume can be accurately defined by the
monitoring wells shown on Figure H.2-14. This conclusion is not likely to be the case, for the
following reasons:
► Ground-flow in the saprolite may be largely along or associated with relict fractures,
which may essentially be missed by several of the monitoring wells shown on this figure.
Particularly, the MW-16 sample data are inconsistent with the remaining sample data.
► The Figure H.2-14 data include one water-supply well on the opposite side of Bat Fork
Creek from the GE area, when data presented elsewhere in the report make a clear case
that under ambient conditions, the stream functions as a ground-water discharge area.
► MW-42 defines the maximum observed source area concentration, not the overall source
area concentration. A more reasonable source area concentration for the modeling
analysis associated with Figure H.2-14 would be to define the source area concentration as
an average PCE concentration for the wells in the area generally upgradient of MW-8,
MW-9 and MW-14. These would be MW-42, RW-2 and RW-3 (reference Figure H.2-2,
yielding an average source area concentration of 1600 ug/L.
Section 2.2.2 of Appendix H discusses use of the Buschek and Alcantar method for analysis of
degradation rate. This method was used to predict a first-order tetrachloroethene (PCE)
degradation rate constant of 0.6 y"1• A first-order rate constant of 0.6y"1, represents a contaminant
half life of 1.155 y. Thus, if valid, a 3-year period (1994 to 1997; the dates when ground-water
concentrations were measured) represents more than two PCE half lives. Over this period, the
half-life determined by the Buschek and Alcantar method as presented in the 30% Design Report
would predict PCE concentrations would be reduced to less than 25% of the initial concentration
at each monitoring point. Reviewing data from Figure H.2-2 and Figure H.2-6, Table I of this
memorandum is constructed:
( •
-5-
Table I. PCE Degradation Analysis
1997 PCE predicted
Monitoring 1994 1997 with A= 0.6y' Ratio, 1997 Observed to
Point PCE observed PCE (assumed 2+ half lives) 1997 Predicted
MW-I! 1600 830 <400 >2.08
MW-12 1200 530 <300 >1.77
MW-9 410 240 <103 >2.33
MW-38 370 150 <93 >l.61
MW-22A 73 55 <18 >3.06
MW-14 380 260 <95 >3.22
RW-1 1600· 1200 · <400 >3
RW-2 1900• 1600 <475 >3.37
RW-3 280• 1200 <70 >17.1
RW-4 1500• 890 <375 >2.37
If these data are further evaluated, the following ,determinations arc made:
Average Average
1994 1997
PCE (Col PCE estimated A• estimated t½0
all daL1 931.3 695.5 0.0973 7.12 y
less contaminated
area (C0 < 1000)* 308.3 176.3 0.186 3.72 y
more contmninated
area (C0 > 1000,+RW-3) 1347 1042 0.086 8.06 y"Y
notes
*excludes the RW-3 observation, where the PCE concentration increased from 1994 to 1997
• calculated using C(t) = C0e·~. where C(t) is concentration at I= 3 years, C0 is initial concentration (1994
concentration) and k = A = degradation rate constant.
01½ = half life = 0.693/A
• 1990 data
~ very conservative estimate, since the earlier recovery well data are from 1990, which would indicate a smaller
degradation rate constant than if a 1994 initial concentration was used.
Based on the analysis in Table 1, the apparent loss of PCE is generally less than the value
presented in Section 2.2.2 of Appendix H. Additionally, the apparent loss of PCE is generally
lowest in the more highly contaminated parts of the plume, around the suspected source area.. As
the Buschek and Alcantar method is a simplistic, one-dimensional model, subject to considerable
uncertainty, its use is considered inappropriate for the estimation of PCE degradation rates at this
site. However, if the Buschek and Alcantar method is used with (I) an assumption of a source
area concentration of 1600 ug/L and (2) an assumption that the MW-16 PCE concentration of 12
ug/L does not represent the true centerline of the plume, an apparent rate constant of 0.2295, with
an apparent half life of 3.02 years is determined. These values compare reasonably well with the
half-life and rate constant values for the less contaminated part of the plume, as presented in
Table 1. As an aside, if the Appendix H PCE mass estimates are used to estimate a degradation
•
-6-
rate (i.e. if the 1994 PCE mass is substituted for C0 and the 1997 PCE mass is substituted for
C(t)), a site-wide average half life of7.42 y is estimated. This value compares closes to the 7.12 y
average half life presented in Table I.
An unknown factor which may relate to the apparent loss of total VOC mass from the ground
water (reference the data analysis of Section 2.1.3 in Appendix H) is the possible operation of the
recovery well system between the 1994 and 1997 ground-water sampling events. Use of the
recovery wells through this time period will need to be discussed in this report before the
estimated PCE mass loss that could largely be attributed to bioremediation can be determined.
Appendix H, Section 2.2.3 presents the results of Bio screen modeling. Bio screen modeling of
contaminant transport must be considered as a very rough estimation procedure, not suitable for
making conclusions applicable to selection of a remedial action for this site. This statement is
made for the following reasons:
► Bioscreen assumes simple ground-water flow conditions.
► Bio screen is designed as a screening model to determine if remediation through natural
attenuation should be further investigated.
► Input parameters to the model are not adequately site-specific. For example, the
assumption of a 0.1 % fraction of organic carbon in the aquifer may be inaccurate. A
review of data in the National Soil Characterization Database (Internet address
www.statlab.iastate.edu:80/soils/ssVnatch data.html indicates that the organic carbon in
soils of the Hayesville series (found at the site) in nearby Buncombe County, North
Carolina and Clay County, North Carolina decreases from approximately I% in near-
surface soil to between 0.0 I to 0.02% (fraction of organic carbon, or foe = 0.000 I to
0.0002) at depths of approximately 2 meters or greater. It would be reasonable to assume
these very low organic carbon values are representative of the soil in the GE/Shepherd
Farm area. If an foe value of 0.000 I was placed into the Bioscreen model, the results
would be considerably different than for an foe of 0.00 I. The material included in
Appendix A to this memorandum documents the difference which might be observed.
► Bioscreen assumes that one· contaminant half-life applies to the entire modeled area. Data
from the GE area suggest that variable half lives apply. This type of complexity is not
handled by the model.
► Bio screen does not account for superimposition of a modeled contaminant source on top
of an already present contaminant plume, yet that procedure was done, as a source mass
of35 kg was assumed (approximately 10% of the 1997 estimated total PCE mass). This
process fails to account for the larger PCE mass (and PCE degradation products) already
in the ground water.
As a review of the 30% Design Report Bio screen modeling, Bio screen was run using the Table
H.2-2 data (except for the following revisions):
(a) for dispersivity, the default value calculated by the Bioscreen model (aL = 28.9 ft; ay = 2.9 ft),
•
-7-
based upon a flow-path length of 1500 ft) was used.
(b) a Koc for PCE = 364 was used (U.S. EPA, 1990).
(c) a source zone 200 ft wide and 30 ft deep was assumed.
(d) an initial source zone concentration= 1500 ug!L (approximate estimate based on data from
Table I of this memorandum) was used.
(e) contaminant half life= 7.12 years, or 3.72 years (see Table I of this memorandum) was used.
(f) soil organic carbon was assumed to be 0.00 I in one series of model runs and 0.000 I in a
second series of model runs.
Observed concentrations for comparison to the model were considered to be upper-aquifer 1997
PCE concentrations along the A-A' cross section (reference report Figure 3-6). These
concentrations are 530 ug/L at MW-12; 240 ug/L at MW-9, and 260 ug/L at MW-14.
Results of this modeling are presented in Appendix A of this memorandum These result indicate
the following:
I. The modeling does not adequately explain the relatively high concentration of PCE at
MW-14 at t = 3 years.
2. The modeling does not predict a plume extent as far downgradient as MW-9 at t = 3
years, unless the fraction of organic carbon is less than 0.00 I.
3. An organic carbon content of 0.00 I would under-predict the plume downgradient extent
at a time of3 years (1997, with 1994 as a starting time), regardless of the contaminant
degradation rate.
4. At an organic carbon content of0.0001, the plume is predicted to extend to the
downgradient margin of the modeled area (i.e. a 1500-foot distance), with a concentration
of IOI ug/L, assuming the 7.12-year half life is valid. For a half life of3.72 years, the
concentration at 1500 feet is predicted to be 47 ug/L.
These observations indicate the inappropriateness of Bio screen for a modeling analysis of the GE
area. However, one must also note that the modeling results of Appendix A suggest that ground-
water contamination may migrate further downgradient than is assumed in the 30% Design
Report. This discrepancy is due to a longer contaminant half life likely applying to at least a part
of the site, and the potentially smaller contaminant retardation factor than that assumed in the
30% Design Report. ·
On page H-28, the report states that with respect to the "fringe" plume, " ... CAH
concentrations ... are at low to non-detect levels and physical attenuation processes ... coupled with
limited aerobic biodegradation are sufficient to severely retard any contaminant migration." This
statement is unsubstantiated in the report. The phrase "severely retard" is undefined, and does
not define adequate plume attenuation·for purposes of evaluating a natural attenuation component
to the remedial action.
If one uses the criteria on page H-28 to define the core plume area, it may be concluded that data
•
-8-
for concentration comparison outside the core area are limited to 5 observations (Table H.2-4;
MW-2, MW-12B: MW-14B, MW-21, MW-22A). MW-32 is not considered to be a valid point
for comparison since neither PeE nor TeE have been detected in that location. Of these five
observations, three observations, defined as having a stable concentration trend, actually show
small to significant increases in PeE concentrations from 1994 to 1997 (MW-2, MW-12B, MW-
14B). MW-12B and MW-14B are deeper monitoring wells associated with shallower "core
plume" contamination. These deeper wells are associated with relatively high dissolved oxygen
concentrations, and would thus not be expected to show appreciable PeE degradation. They are
however, useful for an assessment of PeE contamination migrating into deeper parts of the aquifer
over time. MW-2 is outside the area of the most significant PeE contamination, and is probably
invalid for discussing trends in PeE concentration over time in the fringe plume area. Of the
remaining two wells which show decreases in PeE concentration outside the core plume area,
MW-21 had only 1.5 ug/L PeE when initially sampled in 1994. Data from this well are therefore
useless in assessing trends in fringe plume voe concentrations. MW-22A data indicate decreasing
PeE concentrations from 1994 to 1997. However, the MW-22A data also show non-detect
concentrations of any other voes in the 1997 sampling (Reference Table e-13) , and very low
concentrations of PeE degradation products when sampled in 1994 (reference Figure H.2-7 and
Figure H.2-8). This distribution of chlorinated voes at MW-22A suggests either extremely
efficient biodegradation of the less chlorinated voes, or limited biodegradation of those
constituents. If the latter conclusion is correct, then the observed reduction in PeE at this
monitoring location would have to be ascribed to other processes, of which natural attenuation
(processes other than biodegradation) is likely only one component. In any event, the conclusion
at the bottom of page H-28 and top of H-29 that natural attenuation processes will likely be
sufficient for containment and eventual removal of contaminants of concern outside the central
plume is not supported by these fringe-plume data and the data analysis in Appendix H to the
report.
Appendix H, Section 2.3 applies the data analysis techniques of Section 2.2 to data for the
Shepherd Farm area. The problems observed for the GE area with respect to identifying
contaminant degradation rate are also identified for this analysis. Moreover, the available data for
the Shepherd Farm area are so limited as to make degradation rate analysis almost meaningless.
One may observe that in the area of highest PeE concentration, the concentration has decreased
from 1990 to 1997 (compare Figure H.2-17, SF-PW-01 PeE at 140 ug/L to Figure H.2-16, PeE
at 93 ug/L at SWW-9). However, those same figures suggest that PeE in the area most directly
downgradient from that part of the Shepherd Farm area has increased over the last 3 years (Figure
H.2-17 shows PeE at non-detect at TW-54 in I 994; Figure H.2-16 shows PeE at 48 ug/L in 1997
in the approximate same location). These observations indicate probable transfer of mass from
upgradient to downgradient locations, rather than PeE mass loss due to biodegradation. Data
from Table e-13 indicate that the only chlorinated voe presently at Shepherd Farm is PeE. This
observation leads to a conclusion that biodegradation of PeE is not likely a significant process at
the Shepherd Farm area. This conclusion is not an argument that overall, PeE concentrations are
not generally decreasing over time. Rather, the available data suggest that concentrations are
decreasing due to dispersion, advective transport of the plume away from the source area(s), and
•
-9-
dilution of the plume through recharge. The modeling exercises to predict PCE degradation or
model PCE transport with Bioscreen must be considered invalid, because the available data do not
support a conclusion that a biodegradation process is significant in this area. As an aside, for the
one area where PCE concentrations are observed to decline from 140 ug/L in _1990 to 93 ug/L in
I 997 (SF-PW-01 to SWW-9 comparison), if biodegradation was responsible for that concentration
decrease, and these data were valid for comparison, the apparent degradation rate would be
estimated to be 0.0584, with a PCE half-life of 11.86 years.
With regard to the Section 2.4 conclusions in Appendix H, the following comments are made:
The fourth point on page H-32 regarding the core plume stability in the GE area is probably
correct overall. However, the effects of any operation of existing recovery wells on contaminant
concentrations and plume extent are not discussed in the report. Additionally, source-control
actions (reference report Section 1.1, page 1-3) have contributed to the limitation of plume
migration and are not mentioned in this section of the report. With regard to the plume periphery,
the presence of the ground-water discharge area (Bat Fork Creek) limits plume migration. In the
absence of these factors, it is questionable how effective the natural attenuation processes would
be in limiting plume migration. That is, the effectiveness of natural attenuation in limiting plume
growth and remediating contamination at this site is not adequately demonstrated in the report.
The second point on page H-33 references contaminant transport simulations which have several
deficiencies, as discussed in this memorandum. The validity of these simulations is therefore
questionable.
Data are inadequate to conclude that the third point on page H-33 is correct. Continued plume
migration will be limited by the proximity of the ground-water discharge area'. yet some
concentration increases in some of the more downgradient parts of the plume cannot be
discounted, considering the information presented in this report.
Section 4.5, page 4-4 proposes to exclusively use monitored natural attenuation at the Shepherd
Farm area. The data and analyses presented in this report do not demonstrate that natural
attenuation is an effective remedial alternative to either limit plume migration or to reduce ground-
water contaminant concentrations within an acceptable time frame in the Shepherd Farm area.
Additionally, monitored natural attenuation as an exclusive remedial process is not specified in the
ROD.
The Section 4.5.1 proposal for remedial action at the GE part of the site calls for a pump-and-treat
remedial action to address the core plume area, with monitored natural attenuation to address the
fringe p!u\ne area. Presumably, these areas are as defined in Figure H.2-15 of the report.
Otherwise, the information on page H-28 can be used to define the "fringe plume" area as having:
• PCE concentrations below 200 ug/L
• TCE concentrations below 20 ug/L
•
-10-
• "Low" nitrate levels
One may question this Section proposal on the basis of several factors:
1. Is this plan consistent with the ROD?
2. Is the definition of the core and fringe plumes appropriate, given the nature of
contamination at the site?
3. Does this remedial action result in attainment of remedial objectives (plume remediation to
ROD-specified performance standards) in a reasonable time frame?
4. Does this remedial action stop ( or adequately limit) further plume front migration?
The Record of Decision clearly considers enhanced in-situ bioremediation as a process that can be
used at the GE part of the site, if appropriate, to improve the removal of contaminants from the
subsurface across the area of ground-water contamination. Specifically, page 79 of the Record of
Decision implies that enhancement of bioremediation is designed to speed remedial action through
increased degradation of sorbed or trapped VOC contamination which could otherwise require a
considerable amount of time to adequately migrate via ground water to a recovery well for ex-situ
treatment. This conceptualization of the bioremediation process is clearly different from the
procedure envisioned in the 30% Design Report. Thus, the plan in Section 4.5.1 is inconsistent
with the Record of Decision.
The definition of the core and fringe plumes considers PCE, TCE and nitrate. Nitrate is not
included in the performance standards specified in the ROD. In addition to PCE and TCE, there
are eleven other ground-water constituents with performance standards specified in the Record of
Decision. Fringe-plume monitoring wells with 1997 water-quality data (reference Figure H.2-15
and Table C-13) include MW-2, MW-22A, and MW-38. These fringe-plume wells can be
compared to core plume wells MW-20/MW-20B, MW-11, MW-42, MW12/MW-12A/MW-12B,
MW-18, MW-23, MW-8, MW-9, MW-16, and MW14/MW-14A/MW-14B. Table 2a defines the
average core plume well concentrations of other contaminants with performance standards. Data
were taken from report Table C-13. Table 2b provides a comparison of Table 2a average to
concentrations observed in Remedial Design samples from fringe-plume wells (reference Remedial
Design data, report Table C-13).
Constituent
barium (Ba)
beryllium (Be)
nickel (Ni)
lead (Pb)
manganese (Mn)
vinyl chloride (VC)
Table 2a. Core-Plume Wells-Average Concentrations
Average Concentration. ug/L *
109
not detected in core plume area (ND)
33.7
20.9
285
not detected in core plume area
•
-11-
Table 2a, continued
Constituent Average Concentration, ug/L*
total 1,2-dichloroethene (tl,2 DCE) 82.8
benzene (B) not detected in core plume area
nitrobenzene (NB) not detected in core plume area
chloroform (CF) not detected in core plume area
1,2-dichloroethane (1,2 DCA) 179
• average of observations with detectable concentrations
Table 2b. Comparison of Core-Plume Averages to Fringe Plume Well Concentrations
Fringe
Plume
Well
MW-2
Ba
MW-22A
MW-38 917
109
Be Ni Pb
17.4 54.2
ND 33.7
Concentration Comparison•
Mn< vc
434
285
7380
285
tl,2 DCE B NB CF
4J
ND
1,2 DCA
•made where the fringe plume concentration at a specific monitoring location exceeds the core plume average
concentration
The data in Table 2b are too limited to fully evaluate if the core plume and fringe plume definitions
are appropriate. One may observed that at MW-38, several inorganic constituents exceed average
concentrations in the core plume. This monitoring location is slightly outside the core plume-
fringe plume boundary shown on Figure H.2-15. Of the constituents in the MW-38 sample which
are present in concentrations above core-plume averages, beryllium and manganese are well above
performance standards. The natural attenuation of inorganic compounds is not specifically
addressed by the 30% Remedial Design Document.
Further insight into the core and fringe plume definition may be obtained by reviewing Table C-13
water-quality data for monitoring wells located just outside of the fringe plume area shown on
Figure H.2-15. These wells include MW-21, MW-58, and MW-59. Data from these wells
indicates that manganese exceeds the core plume average in samples from wells MW-58 and
MW-59 (total manganese 481 ug/L and 360 ug/L, respectively). MW-57 is the only monitoring
location which is unequivocally upgradient from any potential facility-related ground-water
contaminant sources. Manganese in the sample from this well was 102 ug/L.
Based on the preceding evaluations and discussions regarding the core and fringe plumes, it is
reasonable to conclude from the ground-water quality data presented in the 30% Design Report
that the fringe plume may contain above-background (and above performance standard)
•
-12-
. concentrations of inorganic constituents which are not specifically considered in the natural
attenuation proposal. One may further conclude that the Figure H.2-15 definition of the core and
fringe plumes is potentially inappropriate, due to either a lack of monitoring data which completely
define the core plume-fringe plume boundary on the basis of PCE, TCE and nitrate concentrations
measured in the 1997 ground-water sampling, or the presence of other ground-water contaminants
of concern in relatively high concentrations which may cross the core plume-fringe plume
boundary defined by the PCE, TCE, and nitrate criteria.
The third question to be asked is whether or not the ground-water remedial action described in the
30% Design Report results in attainment of ground-water remedial objectives in a reasonable time
frame. The key concern here is whether or not the fringe plume area would be adequately
addressed by the proposed natural attenuation remedy. This question cannot be answered without
additional assessment of the available data, and perhaps additional data collection. That
assessment would involve a comparison of the natural attenuation fringe-plume remedy to the
ROD-envisioned remedy where active remediation (capture of the entire plume by recovery wells)
in terms of anticipated remedial time frame to attain performance standards, and the costs of each
option. For the active remediation alternative to address the fringe-plume contamination, costs
would include more recovery wells than those proposed in the 30% Design Report. In a more
indirect manner, any potential adverse environmental impacts (as mentioned in report Section
3.1.3.3) could be factored in as a sort of cost for the ROD-envisioned remedial action. For the
natural attenuation fringe-plume action, costs would likely include additional ground-water
monitoring wells to assess the progress of ground-water remediation in deeper parts of the aquifer
within the fringe-plume area. Additionally, a more frequent "compliance" well monitoring
program than that proposed in report Section 6.5.5 would be required, for a natural attenuation
remedy, at least for a duration sufficient to field-confirm the efficiency of natural attenuation
process for ground-water remediation and for limiting plume expansion.
The fourth question to be asked is whether or not the natural attenuation remedy for the fringe-
plume adequately stops or limits further plume-front migration. This question also cannot be
answered without additional data analysis and perhaps additional data collection. Such an analysis
would assume that the recovery well system effectively cuts off contaminant mass-flux away from
the core-plume or source area, at the downgradient margin of the composite capture zone for the
recovery wells. Next, the ambient ground-water velocities downgradient of that zone ·would have
to be adjusted to account for the change in hydraulic gradient imparted by the recovery well
operation. For the organic contaminants of concern which are principal to the site (i.e. the
chlorinated solvents) a more complex bioremediation model than either Bioscreen, or the one-
dimensional model for determination of apparent biodegradation rate (Buschek and Alcantar),
would be required in order to evaluate bioremediation capability and sustainability in this setting.
As a part of this evaluation, biodegradation indicator parameter data (reference Table C-13) would
be more thoroughly evaluated than is done in the 30% Remedial Design Report. Finally, the
natural attenuation of inorganic contaminants of concern would have to be addressed.
•
-13-
With respect to monitoring wells, whether or not the monitored natural attenuation option was
accepted by the EPA as a part of the overall ground-water remedial strategy, additional ground-
water monitoring would be required between the "compliance" wells as defined in Section 6.5.3
and the plume core wells (Section 6.5.2). This monitoring would be necessary in order to
document the progress of the remedial action at restoration of ground-water quality.
Requirements for such monitoring would likely include the additional installation of monitoring
wells for the monitored natural attenuation remedy. Later versions of the Remedial Design Report
will have to address this inonitoring deficiency.
Some subset of monitoring wells at the GE site will require monitoring for Record of Decision-
specified performance standard constituents other than those listed in report Table 6-5. Most
obviously from this report, MW-38, with a reported sample beryllium concentration above the
performance standard would need to be monitored for this metal, in order to assure that the
remedial objectives were being met in that part of the plume. Monitoring wells between the core
plume wells and "compliance " wells would have to be monitored for a suite of contaminants
which could not be specified until either existing wells in that area were sampled for all
performance standard constituents, or new wells.in that part of the plume were drilled and then
sampled for the performance standard constituents. At the "compliance" monitoring wells, a
baseline analysis of all performance standard constituents is needed in order to establish that only
the VOCs need to be monitored during the remedial action. The 1997 sampling data could
probably be used for this purpose; however, most proposed "compliance" wells were not sampled
then. The acceptable analytical parameter selection may be location-specific, such that, for
example, some "compliance" wells would be monitored for one or more metals.
Section 6.5.3 proposes a contingency plan based on target ground-water concentrations of three
specific volatile organic compounds being exceeded. Specific points where the monitoring
associated with this contingency plan would be considered are the "compliance" monitoring wells.
The EPA has expectations that the ground-water remedial action for this site will both effectively
arrest downgradient plume migration, and virtually eliminate expansion of ground-water
contamination above performance standards into areas which are currently less contaminated. As
such, any such contingency triggering mechanism must be generic, tied to either a relative increase
in contamination of any location-specific contaminant of concern or a minor absolute increase in
concentration (e.g. 20% of the ROD-specified performance standard), rather than an increase
above some arbitrary standard. The contingency mechanism would be applied in both the plume
margin and plume interior. Ideally, such a triggering mechanism would be statistically based,
although other types of contingency triggers would be considered by the EPA. As an additional
comment on this proposal, the nature of the contingency anticipated for such an event should also
be noted in this section of the report.
Section 6.8.2 envisions a possible contingency measure as hydraulic containment of the plume
fringe. This discussion suggests that this contingency remedy would be implemented such that
total containment is achieved. It is more likely that any perceived failure of the natural attenuation
remedy (or a ground-water remedial action as envisioned in the Record of Decision) will apply to
•
-14-
one part of the plume fringe, and adjustments to the ground-water remedial action will only need
to be made in that area. However, apparent failure of the remedial action in one area should also
trigger a complete reassessment of the site ground-water conceptual model, and the design for the
entirety of the ground-water remedial system.
Section 7, the discussion of monitored natural attenuation as the sole ground-water remedial action
for the Shepherd Farm area, cannot be accepted as written, for the following reasons:
I. Monitored natural attenuation conflicts with the ROD-Specified Remedial Action for the
site.
2. Available ground-water data are inadequate to assess the nature of natural attenuation at
this area, or to even demonstrate that natural attenuation is occurring.
If you have any questions concerning this memorandum, or need additional technical assistance,
please contact me at x28645.
References
U.S. Environmental Protection Agency, 1990, Basics of Pump-and-Treat Ground-Water
Remediation Technology, EPN600/8-90/003.
RUN
CENTERLINE RUN ARRAY
Restore Formulas for Vs,
Dispersivities, R, lambda, other
0.596 0.033 0.000 0.000 0.000 0.000
0.725 0.511 0.027 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000
0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000
0.530 0.240 0.260
0.133 0.299 0.569 0.625 0.182 0.025 0.002 0.000 0.000 0.000 0.000
0.133 0.222 0.330 0.310 0.083 0.011 0.001 0.000 0.000 0.000 0.000
0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000
RUN
CENTERLINE
Restore Formulas for Vs,
Dispersivities, R, lambda, other
0.725
0.725
0.000
0.530
0.596
0.442
0.000
0.033 0.000
0.022 0.000
0.000 0.000
0.240
0.000 0.000
0.000 0.000
0.000 0.000
0.000 0.000 0.000 0.000 0.000
0.000 0.000 0.000 0.000 0.000
0.000 0.000 0.000 0.000 0.000
0.260
0.625 0.182 0.025 0.002
0.133 0.171 0.204 0.165 0.040 0.005 0.000 0.000 0.000 0.000 0.000
0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000
:.:=: .::,. : ...... =-._ . ..,:_ .. -... '= ...
\;1111£·,~Q,
i'-t ':5''.0.300:
;, ._=:_-~1-k Next Timestep
10 Years
•
Restore Formulas for Vs,
Dispersivities, R, lambda, other
Replay
Animation
0.725
0.000
0.530
0.906
0.821
0.000
0.902 0.480 0.105
0.759 0.386 0.083
0.000 0.000 0.000
0.240
0.010 0.000 0.000 0.000 0.000 0.000
0.008 0.000 0.000 0.000 0.000 0.000
0.000 0.000 0.000 0.000 0.000 0.000
0.260
0.555 0.620 0.614
0.133 0.157 0.184 0.211 0.239 0.268 0.293 0.303 0.282 0.101
0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000
tiE:!o~·oct
1111;m1;~;it,,wi1::rim1r-iiji~::j;,11:i~as1t
Replay Next Timestep
'
RUN
CENTERLINE
Shepherd Fa
RUN ARRAY
Paste Example Dataset
Restore Formulas for Vs,
Dispersivities, R, lambda, other
0.725 0.906 0.902 0.480 0.105 0.010 0.000 0.000 0.000 0.000 0.000
0.725 0.751 0.649 0.317 0.066 0.006 0.000 0.000 0.000 0.000 0.000
0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000
0.530 0.240 0.260
0.230 0.614 0.460 0.238
0.133 0.143 0.151 0.162 0.166 0.166 0.160 0.140 0.096 0.047
0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000
[_~1;~~1f\~
=--
i
~~?~
.. ~~~if; .
~ ~ii!~;~~~i·
•
Ms. Giezelle Bennett
Superfund Branch
Waste Management Division
• NORTH CAROLINA DEPARTMENT OF
ENVIRONMENT AND NATURAL RESOURCES
DIVISION OF WASTE MANAGEMENT
March 16, 1998
United States Environmental Protection Agency
Region IV
6 I Forsyth Street, 11th Floor
Atlanta, Georgia 30303
RE: Preliminary (30%) Remedial Design Report
General Electric/Shepherd Farm NPL Site
East Flat Rock, Henderson County
Dear Ms. Bennett:
The Preliminary (30%) Remedial Design Report for the General Electric/Shepherd
Farm National Priorities List (NPL) Site was received by the Superfund Section of the
North Carolina Department of Environment and Natural Resources (NC DENR).
The Superfund Section has reviewed this document and offers the following attached
comments.
We appreciate the opportunity to comment on this document. If you have any
questions, please feel free to call me at [phone], extension 349.
Attachment
Sincerely,
j)J 731/(fe-,e#;t~
David B. Mattison, CHMM
Environmental Engineer
Superfund Section
401 OBERLIN ROAD, SUITE 150, RALEIGH, NC 27605
PHONE [phone] FAX [phone]
AN EQUAL. OPPORTUNITY/ AFFIRMATIVE ACTION EMPLOYER· SO% RECYCLED/I 0% POST-CONSUMER PAPER
Ms. Giczelle Bennett
30% RD Comments
March 16, 1998
Page I
•
PRELIMINARY (30%) REMEDIAL DESIGN REPORT
List of Figures
I. The title for Figure 6-6 should state "Schematic diagram of Recovery Wells RW-1 through
RW-18." Furthermore, the word "diagram" is misspelled in the title for Figure 6-8. Please
correct these oversights.
List of Tables
2. The List of Tables states that the title for Table 2-4 is "General residential well information."
However, the title for Table 2-4 is "Remediation goal exceedances in groundwater for
baseline conditions (July 1997)." Please clarify this discrepancy.
Section 1.3 Purpose and Objectives of Preliminary Design
3. The last sentence of the second paragraph of this section states that one portion of the
preliminary design focuses on " ... design of a monitored natural attenuation program at the
GE and Shepherd Farms Subsites to address remediation of the dilute groundwater plume."
Please provide a definition of"dilute groundwater plume" and describe how this definition is
appropriate within the scope of this remedial design.
4. The second sentence of the last paragraph of this section states that "the findings from these
studies, as well as operation of the AGRS, have contributed refinement or modification of the
original remedial approach described in the ROD (e.g., the application of monitored natural
attenuation)." The modification of the remedial approach will require revisiting the Record
of Decision (ROD) and may require a ROD Amendment or an Explanation of Significant
Difference (ESD). Further discussion of these issues as well as a thorough justification must
be made before proceeding with the changes to the remedial design. Please make the
appropriate changes to this section.
Section 1.4 Report Organization
5. The last sentence states that original lab reports and validator's reports were not included with
this report. Unless General Electric (GE) can demonstrate that the analytical results as
provided in the tables were generated using an approved Laboratory Information
Management System (LIMS), this documentation must be provided as an appendix to the
remedial design. Please correct this oversight.
Ms. Giezelle Bennett
30% RD Comments
March 16, 1998
Page 2
Section 2.1 Overview
•
6. The fifth sentence of the first paragraph of this section states that seven volatile organic
compounds (VOCs), two semi-volatile organic compounds (SVOCs) and five metals were
additionally added to the list of target parameters. Please revise this sentence to include the
addition of total polychlorinated biphenyls (PCBs) to the list of target parameters.
Furthermore, this sentence should state " ... because these constituents had to be monitored
at the AGRS discharge to Bat Fork Creek as part of the NPDES Permit." Please correct this
oversight.
Section 2.2 Landfill A
7. The fifth paragraph of this sections states that, at the boring placed at soil gas sample location
LFA-10, 1,2-dichloroethane (1,2 DCA) was detected at a concentration of 1.9 milligrams per
kilogram (mg/kg) at a depth of three to four feet (ft) below land surface (bis). However, the
sampling data, summarized in Table C-2 in Appendix C, indicates that 1,2 DCA was detected
at 19 mg/kg and that benzene was detected at a concentration 2. 7 mg/kg at a depth of six to
eight ft bis. Please clarify these discrepancies.
Section 2.4 Dry Sludge Impoundment
8. The last two sentences of the first paragraph do not agree with the total PCB laboratory
analysis results included as Table C-5 in Appendix C. The total PCB laboratory analysis
results indicate that surficial soil sample DSI-1 I is located outside of the fenced berm and
contains greater than IO mg/kg total PCBs. Please clarify this discrepancy. Furthermore, GE
should recognize that further westwardly delineation of PCB contamination at the dry sludge
impoundment (DSI) is required.
9. The second paragraph of this section describes the collection of soil samples from within the
DSI for PCB analysis using toxicity characteristic leachate procedure (TCLP). The first
sample was collected from the impoundment fill material and the second sample was collected
from the upper-most occurrence of native soil. However, it is unclear if the samples were
collected from the areas where PCB-contaminated fill material from the Shepherd Farm
Subsite was placed within the DSI. Please clarify this discrepancy.
Section 2.5.1 Soils Around Selected Existing Wells at the GE Subsite
10. The last sentence of the first paragraph of this section should state" ... refer to Appendix B
for details on soil gas results)."
Ms. Giezelle Bennett
30% RD Comments
March 16, 1998
Page 3
•
11. The last sentence of the first paragraph of this section states that "no chlorinated or. petroleum
hydrocarbons were detected in the soil gas near either monitoring well ... " The first sentence
of the second paragraph of this section reiterates the above conclusions. However, Table I
of Appendix B indicates that total volatiles were detected by a flame ionization detector (FID)
in surficial soil samples MWl4-3, MW25-2, MW25-3 and MW25-4. Furthermore, Table I
of Appendix B indicates that vinyl chloride was detected in surficial soil sample MW25-2.
Please clarify these discrepancies.
12. The first sentence of the second paragraph of this section should state" ... a confirmation
soil boring was completed near the wells." Please correct this oversight.
13. The fifth sentence of the second paragraph ofthis section should state "Table C-6 summarizes
the results of the subsurface sampling." Please correct this oversight.
14. The sixth and seventh sentences of the second paragraph of this section states that _"no voes
were detected in any of the samples" and that "no secondary sources ofVOe contamination
at these locations are suspected." However, in addition to the soil gas data discussed in
comment # I 0, Table e-6 of Appendix e indicates that cis-1,2-dichloroethene ( cis-1,2-DeE)
was detected in the soil sample collected at a depth of six to eight ft bis in the vicinity ofMW-
25. Please clarify this discrepancy.
Section 2.5.2 Shepherd Farm
15. The third sentence of the first paragraph of this section references Figure 2-15. The figure
referenced should be Figure 2-16. Please correct this oversight.
16. The fourth paragraph of this section states that the only chlorinated hydrocarbon detected was
methylene chloride and that no petroleum hydrocarbons were not detected. However, Figure
2-21 and Table e-8 indicate that vinyl chloride was detected at surficial soil sampling location
SF-29 and voes were detected at surficial soil sampling locations SF-10, SF-12 and SF-29.
Please clarify these discrepancies.
17. The fifth paragraph of this section describes the collection and laboratory loss of all but one
soil sample collected from a soil boring that was installed at soil gas sample location SF-10.
The last three sentences of this paragraph state that since voes were not detected in the one
soil sample collected from an unknown depth or from groundwater samples collected from
the subsequent well installation, one may conclude that voes are not present in soil at MW-
62. This conclusion cannot be made without resampling the soils at MW-62. Please delete
this statement and provide a timeframe for completing a new soil boring at MW-62.
Ms. Giezelle Bennett
30% RD Comments
March I 6, 1998
Pnge4
•
I 8. The fifth and sixth sentences of the sixth paragraph of this section provide the results of
surficial soil samples collected from beneath remnants of two 55-gallon drums. In an effort
to keep the analytical results parallel, the sixth sentence should read as "TCE was detected
at 0.048 mg/kg and 0.029 mg/kg and CIS-1,2-DCE was detected at 0.015 mg/kg and 0.009
mg/kg."
Section 2.6.4 Fish Tissue Sampling
19. The last sentence of this section states that the maximum concentration detected during the
1997 sampling event is 0.58 mg/kg. However, Table C-12 indicates that the maximum
concentration detected during the 1997 sampling event is 0.287 mg/kg. Please clarify this
discrepancy.
Section 2.7.3 Water-Level Survey
20. The last sentence of this section is misleading. The last sentence states that, based on the July
1997 water level survey, groundwater does not flow from the GE Subsite offsite north of
Tabor Road or offsite west of Spartanburg Highway. Although generalized flow conditions
indicate eastwardly flow towards Bat Fork Creek, the last sentence of this section implies that
groundwater contamination is limited to the GE Subsite. However, groundwater monitoring
has demonstrated that groundwater contamination is not limited to the GE Subsite. Please
either delete this sentence or restate this sentence to indicate that the groundwater level
survey indicates generally eastwardly flow conditions.
Section 2. 7.4.1 GE Subsite
21. The conclusions regarding the plume delineation in the second paragraph of this section are
unsubstantiated. Although this configuration of plume delineation may be correct, the lack
of data points along the perimeter of the plume places some degree of uncertainty into these
conclusions. Please revise these conclusions and provide justification for not conducting
additional sampling from the existing monitoring wells that exist at the GE perimeter.
Section 2. 7.4.2 Shepherd Farm Subsite
22. Please revise the conclusion in this section regarding the plume delineation. The plume
delineation provided is only an estimate and all conclusions regarding the plume delineation
should be made with the appropriate caveats. Furthermore, the presence of methylene
chloride at monitoring weU MW-61 and the presence of lead at monitoring wells MW-61 and
MW-63 should be addressed in this section.
Ms. Giezclle Bennett
30% RD Comments
March I 6, 1998
Page 5
•
23. The last two sentences of this section do not make any sense. The two sentences imply that
the dug wells, which are open to the saprolite, will prevent PCE and any other VOC from
being present in the bedrock. Please revise these two sentences.
Section 2. 7.4.3 Residential Wells
24. The word "aesthetic" is misspelled in the sixth sentence of the second paragraph of this
section. Please correct this oversight.
Section 3.1.3.1 Hydrogeologic Setting
25. The sixth sentence of the second paragraph of this section is misleading. This sentence states
that groundwater does not flow from the GE Subsite offsite north of Tabor Road or offsite
west of Spartanburg Highway. Although generalized flow conditions indicate eastwardly
flow towards Bat Fork Creek, this statement implies that groundwater contamination is
limited to the GE Subsite. However, groundwater monitoring has demonstrated that
groundwater contamination is not limited to the GE Subsite. Please either delete this sentence
or restate this sentence to indicate that the groundwater level survey indicates generally
eastwardly flow conditions.
26. · The last paragraph of this section states that vertical head differences were measured at the
GE and Shepherd Farm Subsites. However the only summary of the results of the vertical
head testing is given in the last sentence; groundwater flow is upward near Bat Fork Creek.
Please provide a more detailed summary of the results of the vertical head analysis.
Section 3.2.1 GE Subsite
27. The last sentence of the second paragraph of this section indicates that, based on the PCB
data collection activities conducted in 1994, the depth of Landfill B is one to three feet below
land surface. However, Section 3.2.2 of the Geophysical Investigation Report (SDII, 1997)
included as Appendix A indicates that the depth of burial of suspected debris ranged from
near surface to five to seven ft bis. Please clarify this discrepancy. Furthermore, Section
3.2.2 of Appendix A indicates that the GPR survey is limited to a depth of five to seven ft bis
due to conditional limits posed by the soil. Please provide additional information or
clarification as to how to compensate for this limitation.
Ms. Giczelle Bennett
30% RD Comments
March I 6, 1998
Pagc6
•
28. The last sentence of the third paragraph of this section states that the DSI is disconnected to
the water table. Section 3.1.3.1 stated that groundwater flow generally follows the
topography and that recharge is primarily due to the infiltration of precipitation. Please revise
the last sentence of the third paragraph of this section in accordance with Section 3 .1.3. I
unless a demonstration can be made for the prevention of infiltration at the DSI.
29. The last two sentences of this section are incorrect. Although other significant sources of
groundwater have not been discovered to date, surficial and subsurface investigations have
discovered other areas that exhibit chlorinated and petroleum hydrocarbon contamination.
Please revise these last two sentences accordingly.
Section 3.2.2 Shepherd Farm Subsite
3 0. The second sentence of the second paragraph should state " ... were detected in surficial
soils ... " Please correct this oversight.
Section 3.3.1 Mass Estimates ofVOCs
31. Please provide copies of all input and output generated during the voe mass estimate
analysis.
Section 3.3.2 Extent ofVOCs
32. The last sentence of the first paragraph of this section indicates that voes contamination
increased from I 994 to 1997 at two bedrock groundwater monitoring wells (MW-12 and
MW-14). Furthermore, Figures 3-3 through 3-8 indicate that voes may be pooling at the
saprolite/bedrock interface. This potential source of groundwater contamination needs to be
thoroughly investigated, including the possibilities for fractured flow. Please provide a
workplan and timeframe for investigating this issue. Please revise all other portions of this
submittal accordingly, including Section 3 .2.1 as stated in comment 29.
Ms. Giezelle Bennett
30% RD Comments
March 16, 1998
Page 7
Section 3.3.3.1
•
Overview
33. The second and third paragraphs of this section provide an adequate overview analysis of the
biodegradation parameters and various trends noticed at the GE Subsite. However, in order
to thoroughly demonstrate the natural attenuation potential of the GE Subsite (and the
Shepherd Farm Subsite), adherence should be made to the protocol as set forth in the
Technical Protocol for Natural Attenuation of Chlorinated Aliphatic Hydrocarbons in
Ground Water (United States Air Force Center for Environmental Excellence, 1997), Use
of Monitored Natural Attenuation at Superfund, RCRA Corrective Action, and Underground
Storage Tank Sites (US EPA, 1997) and Draft EPA Region IV Suggested Practices for
Evaluation of a Site for Natural Attenuation (Biological Degradation) of Chlorinated
Solvents (US EPA, 1997). Please make the appropriate revisions to this document.
Section 3.3.3.2 Estimation of Mass Removal
34. Please provide all input and output generated by the SiteGIS analysis, TinMass,
BIOSCREEN, Buschek and Alcantar analysis, etc. as an appendix to this section or to
Appendix H. Please include all data and figures used or generated during the course of these
analyses.
Section 3.6.3.6 Summary of Natural Attenuation Study
3 5. The second bullet item of this section states that the lack of oxygen within the central plume
indicates the presence of biological activity. Although the groundwater's low DO content
indicates the lack of oxygen, this does not necessarily indicate the presence of biological
activity. Please revise this sentence accordingly.
36. The seventh bullet item of this section states that no off-site contamination is present.
However, Figures H.2-2, H.2-5, H.2-6 and H.2-9 of Appendix H -Natural Attenuation
Report actually depict offsite contamination crossing these very boundaries. This section, as
well as all associated modeling, should be revised to reflect this offsite contamination. As
stated previously in both these comments and in the Preliminary Remedial Design, the
inadequacy of the data concerning the horizontal extent of contamination must be addressed
in order to effectively design the groundwater remediation system.
Ms. Gieze!le Bennett
30% RD Comments
March I 6, 1998
Page 8
•
Section 4.2 Landfill B
3 7. The fourth sentence of this section states that subsurface sampling has indicated that PCB
contamination exists only I to 3 ft bis. However, Section 3 .1.2 of Appendix A indicates that
the suspected buried debris/disturbed soils ranged from near land surface to 5 to 7 ft bis.
Please clarify this discrepancy.
Section 4.5 Groundwater Remediation
3 8. The third sentence of the fifth sentence should state " ... which specifies requirements which
must be met to implement natural attenuation."
Section 4.5.1 GE Su bsite
3 9. The first paragraph ohhis section indicates two areas of the natural attenuation study which
need to be addressed. Although this report states repeatedly that the contamination has not
migrated offsite, the need for further investigation exists at the perimeter of the GE Subsite.
The second item that must be addressed is the potential existence ofVOC pooling along the
saprolite/bedrock interface at monitoring wells MW-12 and MW-14. Please revise this
section and provide a workplan and timeframe for completion of these items.
40. Although the work conducted to date may indicate the viability of natural attenuation at this
site, the justifications for the proposed remedial action may be premature. Many of the
previous comments regarding the delineation of the plume, the stability of the plume, the
spatial distribution of the contaminants and the biodegradation parameters, etc must be.
resolved prior to making any firm conclusions. Please revise this section accordingly.
Section 5.1 Landfill A Cap
41. This section provides a general background of the asphalt cover. Please provide detailed
information including, but not limited to, the material specifications, material testing
requirements, the installation procedures, quality assurance/quality control measures to be
implemented, durability, specific operation and maintenance requirements, durability and load
bearing capacity. Please use available US EPA guidance as an aid in developing the cap
design.
Ms. Giezelle Bennett
30% RD Comments
March 16, 1998
Page 9
•
Section 5.2 Landfill B
•
42. This section states that only one foot of soil is to be dug for the Landfill B excavation.
However, Section 4.2 stated that subsurface sampling indicated PCB contamination existed
I to 3 ft bis. Furthermore, Section 3 .1.2 of Appendix A indicates that the suspected buried
debris/disturbed soils ranged from near land surface to 5 to 7 ft bis. Please clarify these
discrepancies.
43. This section does not address how the potential wetland to the south of Landfill B is to be
addressed. Please correct this oversight.
Section 5.3 Dry Sludge Impoundment
44. This section provides a general background of the cover design. Please provide detailed
information including, but not limited to, the material specifications, material testing
requirements, the installation procedures, quality assurance/quality control measures to be
implemented, durability, specific operation and maintenance requirements, durability and load
bearing capacity. Please use available US EPA guidance as an aid in developing the cap
design.
Section 6.7.2 Air Permit and Emission Controls
45. This section states that an air permit has not been required for the AGRS system and is not
anticipated for the full scale system. Please provide details as to how GE plans to estimate
the emissions and determine compliance with federal, state and local air emission standards.
Section 7.1 Monitored Natural Attenuation
46. The modeling conducted to date has been an estimate of the pumping required to contain the
contaminant plume. Please bear in mind that this hydraulic containment will require
refinement in the Intermediate Remedial Design. Furthermore, the issue of wetlands impact
needs to be addressed. This section states that negligible impact may be felt by the wetlands.
Please provide further details as to how the wetlands issues will be addressed.
Ms. Giezelle Bennett
30% RD Comments
March I 6, 1998
Page IO
Section 8 Schedule
• •
4 7. Thie fifth paragraph of this section states that no Intennediated Design submittal will be made
for the containment portion of the remedial design. However, this is in contrast to Section
5.6 which stated that the design specifications were to be submitted in the Intermediate
Design submittal. Due to the number and type of comments being submitted at this junction
of the project, GE should submit all design specifications with the Intennediate Design
submittal.
Figure 2-5 Soil Gas Survey Results at Landfill A for Petroleum Hydrocarbons
48. This figure does not agree with the data and figures provided in Appendix B. In addition to
the data provided in Figure 2-5, the data and tables included in Appendix B states that: Soil
gas sample LFA-8 contained 16.9 micrograms per liter (µg/L) total VOCs; soil gas sample
LFA-11 contained 1.63 µg/L xylenes; soil gas sample LFA-13 contained 1.29 µg/L vinyl
chloride and 11.3 µg/L total VOCs; and, soil gas sample LFA-29 contained 10.7 µg/L total
VOCs. Please correct these oversights.
Figure 2-7 Surficial Soil Sample Locations at Landfill A
49. This figure lists the PCB sample results rather than the sampling location for surficial soil
samples LFA-3 and LFA-9. Please correct this oversight.
Figure 2-13 Surficial Soil Sample Results at Landfill B
50. The PCB results for the surficial soil samples collected at Landfill B, as indicated on Figure
2-13, do not agree with the PCB results as provided on Table C-4 in Appendix C. Please
correct this oversight.
Figure 2-14 Surficial Soil Sample Locations and Results at the DSI
51. The PCB results for the surficial soil samples collected at the DSI, as indicated on Figure 2-
14, do not agree with the PCB results as provided on Table C-5 in Appendix C. Please
correct this oversight..
Ms. Giezelle Bennett
30% RD Comments
March 16, 1998
Page 11
Figure 2-17 Surficial Soil Sample Results at Shepherd Farm
•
52. The PCB results for the surficial soil samples collected at the Shepherd Farm Subsite, as
indicated on Figure 2-17, do not agree with the PCB results as provided on Table C-7 in
Appendix C. Please correct this oversight.
Figure 2-26 Total VOC Concentrations (ppb)
53. The values for total VOCs are incorrect for several groundwater monitoring wells and
groundwater extraction wells. Groundwater monitoring well MW-18 should be 943 parts per
billion (ppb) instead of 928 ppb as indicated on Figure 2-26. Furthermore, the presence of
methylene chloride should be included in this figure for those wells that exhibited methylene
chloride. The Superfund Section recognizes that methylene chloride was not identified in the
ROD as a contaminant of concern. However, GE must recognize that all contaminants must
be remediated in accordance with Title ISA, Chapter 2L of the North Carolina Administrative
Code (I SA NCAC 2L). Please revise this figure accordingly.
Figure 2-29 PCE Concentration in Groundwater (as measured in July 1997)
54. Since no figure was provided for the total VOC concentration in the groundwater, please
place a note in this figure denoting the presence of methylene chloride at groundwater
monitoring well MW-61.
Table 2-2 Target Parameters for Remedial Design Characterization
55. This table indicates that total PCBs is a remedial design target parameter, that cis-1,2-DCE
and zinc are both remedial design target parameters and National Pollutant Discharge
Elimination System (NPDES) permit target parameters, and that 1,2-dichloropropane and
chromium are neither remedial design target parameters nor NPDES permit target
parameters. However, this information appears to be in conflict with Table 1-1. Please
clarify these discrepancies.
Table 2-4 Remediation Goal Exceedances in Groundwater for Baseline Conditions (July
1997)
56. The values in Table 2-4 do not agree with the results provided in Table C-13 in Appendix C.
Please correct this oversight.
Ms. Giezcllc Bennett
30% RD Comments
March I 6, 1998
Page 12
•
Figure 6-6 Schematic Diagram of Recovery Wells RW-1 through RW-18
5 7. This figure should also indicate the type and placement of the surface seal, the type of backfill
around the solid riser pipe, the type and location of the lower seal, the screen slot size, etc.
Please revise this figure accordingly.
Appendix A Geophysical Investigation Report (SDII, 1997)
58. The legend for Figures 2 through 7 indicate that some visual distinction exists between the
lines used to denote: the approximate location of the electromagnetic (EM) transect lines; the
lines used to denote the approximate location of the ground penetrating radar (GPR) transect
lines; and, the lines used to denote the approximate location of GPR and EM transect lines
combined. However, the differences between these lines are not readily apparent in Figures
2 through 7. Please clarify this discrepancy.
Appendix B Soil Gas Survey Report (Target Environmental Services, Inc., 1997)
Sampling Procedures
59. This section describes the procedures used to collect the soil gas samples. Please reference
the appropriate United States Environmental Protection Agency (US EPA) field method of
sample collection and provide the appropriate documentation of compliance with the US EPA
field method as well as the quality assurance/quality controls implemented during sample
collection.
Analytical Procedures
60. The first paragraph of this section lists the specific analytes considered for the US EPA
Method 8010 analysis. Please provide justification for not including 1,2-dichloroethane (1,2-
DCA) and vinyl chloride, both remedial design target parameters, in the US EPA Method
80 IO analysis. The Superfund Section recognizes that screening for vinyl chloride was
conducted during US EPA Method 8020 FID analysis. However, although the similar US
EPA Method 8021 is used for vinyl chloride analysis, US EPA Method 8020 is currently not
an acceptable method of vinyl chloride analysis. Please provide the validation study results
supporting the implementation of US EPA Method 8020 for vinyl chloride analysis.
Additionally, please provide justification for simply screening for vinyl chloride content
according to US EPA Method 8020 and not providing more accurate results by analyzing for
vinyl chloride content according to the US EPA Method 80 IO.
Ms. Giezelle Bennett
30% RD Comments
March 16, 1998
Page 13
•
61. The last paragraph of this section indicates that one sample, MW-25-2, was submitted for off-1
site gas chromatograph/mass spectrometer (GC/MS) analysis for vinyl chloride according to
EPA 8260m procedures. Please provide all supporting documentation of this analysis
including chain of custody documentation and laboratory analytical results documentation.
In addition to this supporting documentation, please provide justification for conducting vinyl
chloride analysis on a soil gas sample using a wet chemistry method. Furthermore, please
discuss the accuracy, reliability and implications of the results of the vinyl chloride analyses
conducted on sample MW-25-2.
QA/QC Table 1
62. This table indicates that field control samples 101, 102 and 103 were collected on May 2,
1997. However, all supporting documentation indicates that field control samples 101, 102
and 103 were collected on May 27, 1997. Furthermore, this table indicates that field control
sample 108 was collected on May 30, 1997. However, all supporting documentation
indicates that field control sample 108 was collected on May 29, 1997. Please clarify these
discrepancies.
QA/QC Table 2
63. This table indicates that field control samples 10 I, 102 and I 03 were collected on May 2,
1997. However, all supporting documentation indicates that field control samples IO 1, I 02
and 103 were collected on May 27, 1997. Furthermore, this table indicates that field control
sample 108 was collected on May 30, 1997. However, all supporting documentation
indicates that field control sample 108 was collected on May 29, 1997. Please clarify these
discrepancies.
Figure 3 GC/FID Results -Petroleum Hydrocarbons
64. This figure indicates that soil gas sample LFA-11 contained no detectable concentrations of
petroleum hydrocarbons. However, Table 1 indicates that soil gas sample LF A-11 contained
1.63 µg/L xylenes. Please clarify this discrepancy.
65. This figure indicates that soil gas sample LFA-13 contained 1.29 µg/L vinyl chloride.
However, Table 1 indicates that soil gas sample LFA-13 contained no detectable
concentrations of vinyl chloride. Please clarify this discrepancy.
Ms. Giczelle Bennett
30% RD Comments
March I 6, 1998
Page 14
Target Standard Operating Procedures (SOP)
•
66. The Target Standard Operating Procedures (SOPs) include the methodology for conducting
the US EPA Method 8010 and US EPA Method 8020 analyses for the given constituents.
However, supplemental analyses for chloroform were conducted in accordance with EPA
Method 8010. Although the Superfund Section recognizes that this method is acceptable for
the analyses of chloroform, please provide the addendum to the SOP which contain the
additional information and procedures in order to conduct such analyses.
67. The SOP for conducting the methane analyses by GC/fCD was inadvertently omitted. Please
correct this oversight.
Chain of Custody Forms
68. The chain of custody forms indicate that custody seals were not applicable for this project and
that the samples were received in good condition and "cold." Please provide justification for
not implementing proper sample handling procedures, including the use of custody seals and
the measurement of sample temperature upon laboratory receipt.
Appendix C Analytical Results
Table C-2 Results of Soil VOC Investigation at Landfill A
69. The quantitation limits used in the analysis of the soil samples collected from Landfill A are
inadequate. In order to demonstrate the presence or absence of contaminants within the soil
of Landfill A, laboratory analysis must be conducted at quantitative limits as specified by the
US EPA Contract Laboratory Program (CLP). Separate laboratory runs for the various
compounds may be required so that CLP detection limits are maintained for all analytes.
Table C-2 implies that, with the exception of 1,2-DCA and benzene, further VOC
· contamination does not exist in Landfill A However, this demonstration cannot be made due
to the elevated laboratory detection limits. Please make sure that all future analyses are
conducted with the appropriate detection limits.
Ms. Giezelle Bennett
30% RD Comments
March 16, 1998
Page 15
Table e-6 Results of Soil VOe Investigation
•
70. The quantitation limits used in the analysis of the soil samples collected in the vicinity of
monitoring wells MW-14 and MW-25 are inadequate. In order to demonstrate the presence
or absence of contaminants in the vicinity of monitoring wells MW-14 and MW-25,
laboratory analysis must be conducted at quantitative limits as specified by the US EPA CLP.
Separate laboratory runs for the various compounds may be required so that CLP detection
limits are maintained for all analytes. Table C-6 implies that, with the exception of cis-1, 2-
DCA, further VOC contamination does not exist in the vicinity of monitoring wells MW-14
and MW-25. However, this demonstration cannot be made due to the elevated laboratory
detection limits. Please make sure that all future analyses are conducted with the appropriate
detection limits.
Table e-9 Results of Soil voe Investigation at Shepherd Farm Subsite
71. The quantitation limits used in the analysis of the soil samples collected at the Shepherd Farm
Subsite are inadequate. In order to demonstrate the presence or absence of contaminants at
the Shepherd Farm Subsite, laboratory analysis must be conducted at quantitative limits as
specified by the US EPA CLP. Separate laboratory runs for the various compounds may be
required so that CLP detection limits are maintained for all analytes. Table C-9 implies that
VOC contamination does not exist in the vicinity of monitoring wells MW-62. However, in
addition to not knowing the depth of sample collection, the elevated laboratory detection
limits make it impossible to conclude as to the actual presence or absence of contaminants
near monitoring well MW-62. Please make sure that all future analyses are conducted with
the appropriate detection limits.
Table e-I 1 Results of Stream Sediment Samples
72. The table that was included as Table C-11 is actually Table C-10 (Results of Stream Water
Samples). Please insert the correct table into this page.
Table e-13 Results of Initial Groundwater Sampling Investigation
73. This table indicates that groundwater samples were not collected for SVOC analysis for nine
of the groundwater monitoring wells. Please provide a justification for not collecting these
groundwater samples.
74. This table indicates that water samples were not collected for VOC analysis for residential
drinking water well WW-82. Please provide a justification for not collecting these samples.
Ms. Giczclle Bennett
30% RD CommcnLs
March 16, 1998
Page 16
Appendix D Groundwater Investigation Report
Section I Introduction
75. The last sentence should state " ... except where noted."
Section 2.2 Well Installation
•
76. The third paragraph of this section states that the monitoring wells were constructed of
stainless steel screens and polyvinyl chloride (PVC) casing. The US EPA SOP states that
stainless steel materials should be used if the monitoring program is to be designed for organic
compound analysis. Please provide justification for the use of PVC casing instead of stainless
steel casing.
Section 2.2.l Saprolite Wells
77. The first paragraph of this section indicates that saprolite soil samples were collected for
headspace analysis using a split spoon sampler. Please provide a description of the
methodology, description of the equipment used, and results of the headspace analysis.
78. The third paragraph of this section states that the stainless steel well screen was
decontaminated according to Method I in Section 3.1. However, Section 3.1 states that
Method 2 should be used for all materials used to collect samples undergoing volatile organic
compound or trace inorganic constituent analysis. The Superfund Section recognizes that
although the well screen is not a piece of sampling equipment, the well screen should be
handled with the same level of decontamination as the sampling equipment. Please provide
justification for only providing minimal decontamination of the well screen.
Section 2.2.2 Bedrock Wells
79. The sixth paragraph of this section states that the stainless steel well screen was
decontaminated according to Method I in Section 3. I. However, Section 3. I states that
Method 2 should be used for all materials used to collect samples undergoing volatile organic
compound or trace inorganic constituent analysis. The Superfund Section recognizes that
although the well screen is not a piece of sampling equipment, the well screen should be
handled with the same level of decontamination as the sampling equipment. Please provide
justification for only providing minimal decontamination of the well screen.
Ms. Giezelle Bennett
30% RD Comments
March 16, 1998
Page 17
Section 2.4.3 Residential Wells
•
80. The last sentence of the third paragraph of this section should be stated " . . . secondary
standards are established for taste and aesthetic purposes and do not signify an adverse health effect."
Section 2.5 Quality Assurance
81. This section states that all of the data, including the quality assurance/quality control samples
and procedures, have been reviewed and validated by a third party. Please include all
laboratory analysis reports, chain of custody documentation, quality assurance/quality control
documentation, data validation reports and field notes for all sampling events.
Table D.2-1 Summary of Well Construction Details
82. This table indicates that monitoring well MW-62, although intended to monitor the saprolite,
was set and screened five feet into the upper bedrock. Please provide further discussion of
why this well was set partially into the bedrock as well as a discussion of what purpose this
will serve the groundwater monitoring program.
Figure D.2-3 Location of Site Wells Sampled in July 1997 at the GE Subsite
83. This figure indicates a number of perimeter groundwater monitoring wells located at the GE
Subsite that were not sampled during the July 1997 sampling event. In light of the degree of
uncertainty of the contaminant plume boundary, please provide justification for not
conducting additional sampling.
Attachment 1 Well Construction Reports
84. The well construction report for groundwater monitoring well MW-62A indicates that a
substantial fracture exists at a depth of 50.5 feet to 54 feet below land surface. However, the
bentonite seal was placed at a depth of 50 feet to 54 feet below land surface. Please provide
a justification for placing the bentonite seal at this location and not monitoring this fracture.
Appendix E Groundwater Modeling Report
85. Please provide all data reports generated from the use ofMODFLOW and MODPATH.
These data reports should include all input and output values and should be included for
all iterations of each program. Any figures used or developed as a result of these
modeling efforts should also be included· in this report.
Ms. Giczclle Bennett
30% RD Comments
March I 6, 1998
Page 18
•
Section E.1 Conceptual Model, Model Geometry, and Boundaries
86. The second sentence of the first paragraph of this section references the use of 12 inches per
year (in/yr) net recharge for precipitation. Please provide a reference for this value.
87. The fourth paragraph of this section discusses the development and use of the topographic
drainage basin. Please provide a figure that graphically describes the topographic drainage
basin and briefly describe the development of the topographic drainage basin.
88. The fourth paragraph ofthis section indcates that drains were assigned to layer 1 in areas of
potential wetlands. Figure E.E-4 indicates the presence of only one drain. However, the
wetland delineation provided in Appendix F indicates the presence of three wetlands on or
adjacent to the GE Subsite. Please correct this oversight.
89. The fourth paragraph of this section states that residential wells were not simulated. The
Superfund Section recognizes that residential wells remove a small amount of water in
relation to the aquifer as a whole. However, since the topographic drainage basin is rather
small and the residential well information in the area is known, this information should be used
in the final remedial design.
Section E.2 Model Calibration
90. The fifth paragraph of this section discusses the pumping rates used to reproduce the
observed drawdowns experienced during the December 1997 aquifer test. Please
demonstrate that the pumping rates used for the model calibration correspond with the actual
pumping rates used during the aquifer test.
9 I. The first two bullet items in the fifth paragraph of this section (page E-5) should be stated as
follows:
"Saprolite hydraulic conductivity of approximately 5 ft/d provides a good match with
drawdown data, and provides an improved match in the eastern portion of the GE
Subsite. Hydraulic conductivity of less than approximately 5 ft/d provide an
improved match in the western portion of the GE Subsite.
Bedrock transmissivity of50 ft2/d (corresponding to hydraulic conductivity of0.5 ft/d
and and aquifer thickness of 100 ft) provides a good match with drawdown data, and
lower values provide an improved match in the western prtion of the GE Subsite."
Ms, Giezelle Bennett
30% RD Comments
March I 6, I 998
Page 19
•
92, The last sentence in the fifth paragraph of this section states that the output values obtained
are all very reasonable with respect to values determined from slug tests and pumping tests
reported from previous studies. Please provide, at a minimum, a summary of the previous
tests and demonstrate the "reasonableness" of the output data.
Section E.3Predictions of Groundwater Capture, GE Subsite
93. The third sentence of this section indicates that a detailed optimization of locations and rates
was not performed. Although the Superfund Section recognizes that a detailed optimization
may not be prudent at this junction, potential locations for the recovery wells have already
been used for the tratment system design in Appendix I. Please incorporate the locations of
the extraction wells into these model iterations.
94. The first sentence of the third paragraph of this section indicates the I 5 in/yr is the likely
upper bound of net recharge. Please provide a refeerence and justification for the use of this
value as an upper bound.
95. The last sentence of the third paragraph of this section indicates that a 25 percent safety factor
is appropriate for the design calculations. Please provide a brief justification for the use of
this factor of safety.
Figure E.E-5 Reduced hydraulic Conductivity, Layer 1
96, Please provide a legend which denotes the area of reduced hydraulic conductivity.
Figure E.E-6 Simulated Water Levels without Pumping, Layer 1, Modelwide
97. This figure indicates the location of the recovery wells although the figure depicts the
simulated water levels without pumping. Please delete the recovery wells from this figure,
as well as subsequent figures, which depict static conditions without pumping.
Appendix F Wetland Delineation Report (ESC, Ltd., 1998)
Site Reconnaissance
98. This section indicates that the jurisdictional boundaries of the wetlands are to be verified by
the United States Army Corps of Engineers (USACE). Please provide a timeframe for this
task to be completed. All reports generated by this verification inspection should be
submitted to the North Carolina Department of Environment and Natural Resources (NC
DENR).
Ms. Giezelle Bennett
30% RD Comments
March I 6, I 998
Page 20
Discussion
•
99. This section indicates that GE should perform a metes and bounds survey following the
USACE inspection. Please provide details of GE's plans to perform the metes and bounds
survey.
100. This section provides a discussion of the steps and requirements that GE must follow prior
to impacting wetland areas. Please provide a thorough discussion as to how GE plans to
anticipate the effects of the groundwater remediation system and what steps GE plans to take
to comply with the wetland impact permitting requirements.
Appendix G Landfill Modeling Report
IO I. This report provides an overview of the Hydrologic Evaluation of Landfill Performance
(HELP) modeling. Please provide the actual input and output generated by the HELP model.
Appendix H Natural Attenuation Report
102. All input and output results of the SiteGIS analysis, TinMass, BIOSCREEN, Buschek and
Alcantar analysis, etc. should be provided as an appendix to this section. Please include all
data and figures used or generated during the course of these analyses.
List of Figures
103. The title for Figure H.2-17 should state "PCE and TVOC concentrations in groundwater (as
measured in 1990 and 1994)." Please correct this oversight.
Section 1.4 Overview of Site Conditions
104. The sixth paragraph of this section states that the hydraulic conductivity of 4.0 ft/day was
used in the natural attenuation modeling of both subsites. However, the results of the
groundwater modeling indicated that a hydraulic conductivity of 5.0 ft/day best
represented the site. These values are in contrast to the actual values that have been
exhibited during previous studies. The results of the previous studies should be compared
in much greater detail to these estimates being used for modeling purposes. Furthermore,
the estimates used for modeling purposes should be used consistently between each type
of model. Please revise this section, as well as other affected sections/modeling iterations,
accordingly.
Ms. Giczelle Bennett
30% RD Comments
March 16, 1998
Page 21
•
Section 1.5 Mathematical Models for Natural Attenuation
•
105. The justification for using the simpler analytical models that is provided in this section is
incorrect. The Superfund Section recognizes that, for screening purposes, a simpler analytical
model may be employed in order to obtain an approximation of the attenuation rate at a site.
However, each model must be evaluated on the model's merits, not simply on the lack of
available data in which to operate a more complex, yet possibly more accurate, analytical
model. Please revise this section accordingly.
Section 2.1.1 Spatial Distribution of Contaminants
106. The first sentence in this section should state "in the groundwater at the GE Subsite were
visualized using SiteGIS." Please correct this oversight.
107. The second paragraph of this section describes the creation of three sets of control points.
These control points in effect create the contaminant boundary conditions. However, there
has been no evidence to support the creation of these zero boundary conditons that have been
placed at the two roads and one creek that surround the GE property. Figures H.2-2, H.2-5,
H.2-6 and H.2-9 actually depict offsite contamination crossing these very boundaries. This
section, as well as all associated modeling, should be revised to reflect this offsite
contamination. As stated previously in both these comments and in the Preliminary Remedial
Design, the inadequacy of the data concerning the horizontal extent of contamination must
be addressed in order to effectively design the groundwater remediation system.
Section 2.2.1 Preliminary Analysis
108. The preliminary analysis conducted in this section is a general discussion of the attenuation
potential of the GE Subsite. However, in order to thoroughly demonstrate the natural
attenuation potential of the GE Subsite (and the Shepherd Farm Subsite), adherence should
be made to the protocol as set forth in the Technical Protocol for Natural Attenuation of
Chlorinated Aliphatic Hydrocarbons in Ground Water (United States Air Force Center for
Environmental Excellence, 1997), Use of Monitored Natural Attenuation at Superfund,
RCRA Corrective Action, and Underground Staorage Tank Sites (US EPA, 1997) and Draft
EPA Region IV Suggessted Practices for Evaluation of a Site for Natural Attenuation
(Biological Degradation) of Chlorinated Solvents (US EPA, 1997). Please make the
appropriate revisions to this document.
Ms. Giezclle Bennett
30% RD Comments
March I 6, I 998
Page 22
Section 2.2.4 Identifying the Central Plume
•
I 09. The fifth sentence of the third paragraph of this section states that the low dissolved oxygen
(DO) levels indicate the depletion of oxygen and the existence of biological activity.
Although the groundwater's low DO content indicates the lack of oxygen, this does not
necessarily indicate the presence of biological activity. Please revise this sentence
accordingly.
Section 2.3.1 Areal Distribution of PCE at the Shepherd Farm Subsite
110. Figure H.2-16 and Figure H.2-17 depict the tetrachloroethene (PCE) concentrations located
at the Shepherd Farm Subsite. Please provide justification for maually drawing the plume
isoconcentration contours. The first paragraph of this section indicates that there is
insufficient data to characterize this subsite. Please provide a workplan and timeframe for
properly characterizing this subsite. Furthermore, please provide a justification for the
isoconcentration contours crossing the Unnamed Creek after much discussion of the creeks
being a source of groundwater discharge to the surface as well as a hydraulic boundary for
each subsite.
Section 2.4 Summary of Natural Attenuation Evaluation
111. The second bullet item of this section states that the lack of oxygen within the central plume
indicates the presence of biological activity. Although the groundwater's low DO content
indicates the lack of oxygen, this does not necessarily indicate the presence of biological
activity. Please revise this sentence accordingly.
. . . ' \ •
UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
REGION 4
61 Forsyth Street, S.W.
Atlanta, Georgia 30303-3104
March 11, 1998
4WD-NSRB
Michael J. Bush, Manager
Environmental, Health, and Safety
GE Lighting Systems, Inc
3010 Spartanburg Highway
Hendersonville, NC 28792
SUBJ: GE/Shepherd Farm NPL Site
East Flat Rock, NC
Dear Mr. Bush:
RECEIVEc,
MAR 12 1998
SUPERFUND SEC; ,,
Enclosed are the Agency's comments on the draft 30% Remedial Design dated
February 1998. Please provide a written letter response to these comments to EPA no
later than April 10, 1998. This document will not be finalized. Comments should be
incorporated into the upcoming 60% Remedial Design document due to the Agency on
May 4, 1998. Comments from one additional reviewer will be transmitted next week.
If you have any questions, please give me a call at 404/562-8824.
cc: Dave Mattison, NC DENR
Sine
G' zelle S. Bennett
Remedial Project Manager
Todd Hagemeyer, HSI Geotrans
Lynn France, CDM
, ' • COMMENTS ON THE PRELIMINARY DESIGN REPORT
GE/SHEPHERD FARM SITE
GENERAL COMMENTS
1. The Report has proposed changing the remedy at the Shepherd Farm Subsite from pump
and treat to natural attenuation. The report also has proposed designating a "fringe" plume
at the GE Site that contains concentrations over 200 times the remediation goals, and
using natural attenuation as the remedy for this portion of the plume. Neither is acceptable
to EPA or NC DENR and neither of these proposed remedies is in accordance with the
ROD, CD, or SOW.
2. The PRPs are proposing to excavate contaminated soils at Landfill Band relocating these
soils in the Dry Sludge lmpoundment (OSI) (along with the contaminated soils from
Shepherd Farm). This change to the remedy seems reasonable given the small quantity of
contaminated soils at Landfill B. It may be advantageous to also do the same with
contaminated soils at Landfill A. The PRPs claim that the OSI needs to be backfilled with
soils before it can be capped. Why not use the contaminated soils from Landfill A for this
purpose? Consolidating all the contaminated soils at the OSI simplifies the remedy as well
as monitoring and maintenance of the remedy. In addition, the PRPs can then use the
Landfill A area without concern as to any impacts on a landfill cap.
3. The design does not include infiltration or leachate monitoring for either Landfill A or the
OSI. Some type of monitoring is recommended to verify the integrity of the cap over the 30
year life of the remedy.
4. It is not clear that the asphalt layer design accounted for heavy traffic loads. This
parameter should be listed in the document as part of the design basis for the cap.
5. The report does not contain all the information needed to evaluate if the proposed alternate
cap for Landfill A is equivalent to the ROD selected cap. The specific data used to
determine the hydraulic conductivity of the asphalt layer, and the assumptions used in the
HELP model should be included. The additional data are needed because the cap has a
duel purpose: cover and storage. The use as storage may result in cracking, and the cover
design should include this possibility.
6. The PRPs correctly state that high nitrate concentrations limit biological dechlorination, but
continue to also state that PCE and TCE are dechlorinating to DCE and VC, and that VC is
biodegrading under mildly reducing to aerobic conditions. There are several problems with
these assertions. Redox conditions at the site are not supportive of PCE/TCE reducing to
VC. Oxygen concentrations are generally too high (0.7 to 5 mg/L), nitrate is too high and
total organic carbon is insufficient. There is only limited'detection of DCE and no detection
of VC. If VC were being biodegraded aerobically, it would still be detected in the
anaerobic zones. The low concentration of DCE relative to PCE/TCE does not suggest
dechlorination reactions are occurring. Redox measurements are not supportive of
conditions for dechlorination.
7. The PRPs are proposing natural attenuation as the remedy for the Shepherd Farm subsite
as well as for the "fringe area" of the GE subsite, claiming that this remedy is "fully
protective of human health and downgradient receptors". The PRPs have not
substantiated this claim, particularly for the Shepherd Farm subsite. At the Shepherd Farm
-1 -
• subsite, the only voe found in groundwater has been PCE. The lack of any degradation
products indicates that biological attenuation is not significant. Therefore, the only form of
attenuation apparently occurring at the Shepherd Farm subsite is flushing of the
contaminants into Bat Fork Creek. The PRPs, however, have not presented any kind of
analysis demonstrating that this flushing of VOCs into Bat Fork Creek is not and will not be
harmful to the ecological community in Bat Fork Creek. The same should be done for the
"fringe area" at the GE subsite since, biological activity in this area is apparently a minor
component of natural attenuation.
8. The proposed groundwater extraction well system is not conservative enough such that
capture of the central plume will undoubtedly be ensured. There is a big gap
(approximately 800 feet) between proposed extraction wells RW-6 and RW-7, and because
of the uncertainty associated with groundwater modeling in such a complex aquifer system,
there is concern that significant contamination could escape through this big gap. To
alleviate this concern, an extraction well to the system should be added just downgradient
of the OSI.
9. The natural attenuation study was done as part of the treatability study. Reporting it in this
document was completely out of context and was misunderstood. Please keep all
discussions of natural attenuation with the treatability study.
10. There are numerous typos, incomplete sentences, etc in this report. Please QA/QC the
next submission.
SPECIFIC COMMENTS
1. List of Figures -The title for Figure 6-6 should state "Schematic diagram of Recovery Wells
RW-1 through RW-18." Furthermore, the word "diagram" is misspelled in the title for
Figure 6-8. Please correct these oversights.
2. List of Tables -The List of Tables states that the title for Table 2-4 is "General residential
well information." However, the title for Table 2-4 is "Remediation goal exceedances in
groundwater for baseline conditions (July 1997)." Please clarify this discrepancy.
3. Page 1-2, 2nd complete paragraph, -Delete "and possibly" from the second sentence.
4. Page 1-2, Section 1.1.2, second paragraph, first sentence: This is the first time Clark's
property is mentioned. Is it pertinent to include it in this document? Especially, in light of
the fact that the last sentence in the first paragraph on page 1-3 states, "No
activities .. .lor. .. Clark property".
5. Page 1-3, 2nd line -Change "Tetrachloroethane" to "Tetrachloroethene".
6. Page 1-3, Section 1.2: Pertinent portions from the ROD, particularly the "Declaration for
the Record of Decision" and the "Description of the Selected Remedy" section, need to be
incorporated into this section of the RD. Without this information, it is not feasible to
accurately review this document as the reader does not know what should be embodied
into the design. Consequently, it is not feasible to determine if this 30% RD is in
compliance with the requirements of the ROD.
-2-
• • 7. Page 1-4, 3rd bullet-Add to the end of the 3rd bullet, "by pump and treat".
8. Page 1-5, Section 1.3 -The last sentence of the second paragraph of this section states
that one portion of the preliminary design focuses on " ... design of a monitored natural
attenuation program at the GE and Shepherd Farms Subsites to address remediation of
the dilute groundwater plume." Please provide a definition of "dilute groundwater plume"
and describe how this definition is appropriate within the scope of this remedial design.
9. Page 1-5, Section 1.3 -The second sentence of the last paragraph of this section states
that "the findings from these studies, as well as operation of the AGRS, have contributed
refinement or modification of the original remedial approach described in the ROD (e.g.,
the application of monitored natural attenuation)." The modification of the remedial
approach will require revisiting the Record of Decision (ROD) and may require a ROD
Amendment or an Explanation of Significant Difference (ESD). Further discussion of these
issues as well as a thorough justification must be made before proceeding with the
changes to the remedial design. Please make the appropriate changes to this section.
10. Page 1-5, Section 1.3, first paragraph: This paragraph refers to "refinement or modification
of the original remedial approach", will this require a ROD amendment or an ESD? The
language in this section is very vague. The objective of the RA is to achieve the
performance standards specified in the ROD using the remedy delineated in the ROD.
11. Page 1-4, 1-5, Section 1.3 -This section discusses the purpose and objectives of the
Preliminary Design and at the end of the first paragraph rightly states "for implementation
of the ROD remedies." Therefore, in the 2nd paragraph of this section, delete the word
"central" in front of the words "groundwater plume" and delete the entire (3). This is not per
the ROD. In addition, the 3rd paragraph should be deleted after the first sentence.
12. Page 1-5, Section 1.4 -Add the words "Work Plan" after "Treatability Study" in the second
sentence. Delete the 4th sentence.
13. Page 1-6, last sentence -Delete the words "upon request by EPA".
14. Page 1-6, Section 1.4 -The last sentence states that original lab reports and validator's
reports were not included with this report. Unless General Electric (GE) can demonstrate
that the analytical results as provided in the tables were generated using ari approved
Laboratory Information Management System (LIMS), this documentation must be provided
as an appendix to the remedial design. Please correct this oversight.
15. Page 2-1, first paragraph -Two typos to note: "the" in the 10th line, and "NPDES" in the
11th line.
16. Page 2-1, Section 2.1 -The fifth sentence of the first paragraph of this section states that
seven volatile organic compounds (VOCs), two semi-volatile organic compounds (SVOCs)
and five metals were additionally added to the list of target parameters. Please revise this
sentence to include the addition of total polychlorinated biphenyls (PCBs) to the list of
target parameters. Furthermore, this sentence should state " ... because these
constituents had to be monitored at the AGRS discharge to Bat Fork Creek as part of the
NPDES Permit." Please correct this oversight.
17. Page 2-3, 2nd complete paragraph -Add a note to the end of this paragraph that states
that the trench was not located in an area of suspected debris.
- 3 -
• •
18. Page 2-3, Section 2.2, last paragraph: Since the landfill is covered with asphalt, it would
be helpful if the depth that these soil samples were collected at was specified (i.e., top six
inches, top twelve inches, just below the asphalt etc.). Add a sentence that states that
the samples ranged from 0.009 ppm to 690 ppm.
19. Page 2-3, Section 2.2 -The fifth paragraph of this section states that, at the boring placed
at soil gas sample location LFA-10, 1,2-dichloroethane (1,2 DCA) was detected at a
concentration of 1.9 milligrams per kilogram (mg/kg) at a depth of three to four feet (ft)
below land surface (bis). However, the sampling data, summarized in Table C-2 in
Appendix C, indicates that 1,2 DCA was detected at 19 mg/kg and that benzene was
detected at a concentration 2.7 mg/kg at a depth of six to eight ft bis. Please clarify these
discrepancies.
20. Page 2-4, top paragraph -Please provide an interpretation of the methane results. What
does 5.83% indicate?
21. Page 2-5, last paragraph -Delete the 2.nd sentence.
22. Page 2-5, Section 2.4 -The last two sentences of the first paragraph do not agree with the
total PCB laboratory analysis results included as Table C-5 in Appendix C. The total PCB
laboratory analysis results indicate that surficial soil sample DSl-11 is located outside of the
fenced berm and contains greater than 1 0 mg/kg total PCBs. Please clarify this
discrepancy. Furthermore, GE should recognize that further westwardly delineation of PCB
contamination at the dry sludge impoundment (DSI) is required.
23. Page 2-5, Section 2.4 -The second paragraph of this section describes the collection of
soil samples from within the DSI for PCB analysis using toxicity characteristic leachate
procedure (TCLP). The first sample was collected from the impoundment fill material and
the second sample was collected from the upper-most occurrence of native soil. However,
it is unclear if the samples were collected from the areas where PCB-contaminated fill
material from the Shepherd Farm Subsite was placed within the DSI. Please clarify this
discrepancy.
24. Page 2-6, Section 2.5.1 -The last sentence of the first paragraph of this section should
state " ... refer to Appendix B for details on soil gas results)."
25. Page 2-6, Section 2.5.1 -The last sentence of the first paragraph of this section states that
"no chlorinated or petroleum hydrocarbons were detected in the soil gas near either
monitoring well ... " The first sentence of the second paragraph of this section reiterates
the above conclusions. However, Table 1 of Appendix B indicates that total volatiles were
detected by a flame ionization detector (FID) in surficial soil samples MW14-3, MW25-2,
MW25-3 and MW25-4. Furthermore, Table 1 of Appendix B indicates that vinyl chloride
was detected in surficial soil sample MW25-2. Please clarify these discrepancies.
26. Page 2-6, Section 2.5.1, first paragraph: This paragraph relates to contaminants being
detected in monitoring wells MW-14 and MW-25. Were these wells resampled to confirm
or refute the presence of these contaminants? It is not clear from the discussion.
27. Page 2-6, Section 2.5.2 Shepherd Farm -The points in Figure 2-17 where total PCBs in
surficial soils exceed the performance goal should be highlighted. The figure notations
seem to be off. The 3rd sentence should reference Figure 2-16, the 4th sentence should
reference Figure 2-17, and the last sentence should reference Figure 2-18.
- 4 -
• •
28. Page 2-6, Section 2.5.1 -The first sentence of the second paragraph of this section should
state " ... a confirmation soil boring was completed near the wells." Please correct this
oversight.
29. Page 2-6, Section 2.5.1 -The fifth sentence of the second paragraph of this section should
state "Table C-6 summarizes the results of the subsurface sampling." Please correct this
oversight.
30. Page 2-6, Section 2.5.1 -The sixth and seventh sentences of the second paragraph of this
section state that "no voes were detected in any of the samples" and that "no secondary
sources of voe contamination at these locations are suspected." However, in addition to
the soil gas data discussed above, Table C-6 of Appendix C indicates that cis-1,2-
dichloroethene (cis-1,2-DCE) was detected in the soil sample collected at a depth of six to
eight ft bis in the vicinity of MW-25. Please clarify this discrepancy.
31. Page 2-6, Section 2.5.2 -The third sentence of the first paragraph of this section
references Figure 2-15. The figure referenced should be Figure 2-16. Please correct this
oversight.
32. Page 2-7, Section 2.5.2 Shepherd Farm -It appears there is a data presentation problem
in the northeast corner of the plot in Figure 2-18 as several points outside the excavation
area in this corner indicate a PCB concentration of 1000 mg/kg.
33. Page 2-7 - A conclusion should be included. It appears that there was VOC contamination
in the soil in the past, and therefore, the cause of the voe contamination seen in the
groundwater.
34. Page 2-7, 4th paragraph -Need to indicate why this sampling was not done again.
35. Page 2-7, 5th paragraph -Were the drums sampled? If so, what was detected?
36. Page 2-7, Section 2.5.2 -The fourth paragraph of this section states that the only
chlorinated hydrocarbon detected was methylene chloride and that no petroleum
hydrocarbons were not detected. However, Figure 2-21 and Table C-8 indicate that vinyl
chloride was detected at surficial soil sampling location SF-29 and voes were detected at
surficial soil sampling locations SF-10, SF-12 and SF-29. Please clarify these
discrepancies.
37. Page 2-7, Section 2.5.2 -The fifth paragraph of this section describes the collection and
laboratory loss of all but one soil sample collected from a soil boring that was installed at
soil gas sample location SF-10. The last three sentences of this paragraph state that since
VOCs were not detected in the one soil sample collected from an unknown depth or from
groundwater samples collected from the subsequent well installation, one may conclude
that voes are not present in soil at MW-62. This conclusion cannot be made without
resampling the soils at MW-62. Please delete this statement and provide a time frame for
completing a new soil boring at MW-62.
38. Page 2-8, Section 2.6.1 General Approach -The Stream-3 sampling location shown in
Figure 2-22 is located upstream of the Shepherd Farm subsite. Is this location correct? If
so, the sample results should be considered to represent background conditions.
- 5 -
• • 39. Page 2-8, Section 2.5.2 -The fifth and sixth sentences of the sixth paragraph of this
section provide the results of surficial soil samples collected from beneath remnants of two
55-gallon drums. In an effort to keep the analytical results parallel, the sixth sentence
should read as "TCE was detected at 0.048 mg/kg and 0.029 mg/kg and CIS-1,2-DCE was
detected at 0.015 mg/kg and 0.009 mg/kg."
40. Page 2-10, Section 2.6.4 -The last sentence of this section states that the maximum
concentration detected during the 1997 sampling event is 0.58 mg/kg. However, Table C-
12 indicates that the maximum concentration detected during the 1997 sampling event is
0.287 mg/kg. Please clarify this discrepancy.
41. Page 2-1 0 -Table C-11 was not provided. Two copies of Table C-1 0 were included.
42. Page 2-11 -The values given in Table C-12 do not agree with the values given in the text.
43. Page 2-13, Section 2.7.3 -The last sentence of this section is misleading. The last
sentence states that, based on the July 1997 water level survey, groundwater does not
flow from the GE Subsite offsite north of Tabor Road or ottsite west of Spartanburg
Highway. Although generalized flow conditions indicate eastwardly flow towards Bat Fork
Creek, the last sentence of this section implies that groundwater contamination is limited to
the GE Subsite. However, groundwater monitoring has demonstrated that groundwater
contamination is not limited to the GE Subsite. Please either delete this sentence or
restate this sentence to indicate that the groundwater level survey indicates generally
eastwardly flow conditions.
44. Page 2-13, Section 2.7.4.1 -The conclusions regarding the plume delineation in the
second paragraph of this section are unsubstantiated. Although this configuration of plume
delineation may be correct, the lack of data points along the perimeter of the plume places
some degree of uncertainty into these conclusions. Please revise these conclusions and
provide justification for not conducting additional sampling from the existing monitoring
wells that exist at the GE perimeter.
45. Page 2-14, Section 2.7.4.2 -Please revise the conclusion in this section regarding the
plume delineation. The plume delineation provided is only an estimate and all conclusions
regarding the plume delineation should be made with the appropriate caveats.
Furthermore, the presence of methylene chloride at monitoring well MW-61 and the
presence of lead at monitoring wells MW-61 and MW-63 should be addressed in this
section.
46. Page 2-14, Section 2.7.4.2 -The last two sentences of this section do not make any sense.
The two sentences imply that the dug wells, which are open to the saprolite, will prevent
PCE and any other voe from being present in the bedrock. Please revise these two
sentences.
47. Page 2-14, Section 2.7.4.3 -The word "aesthetic" is misspelled in the sixth sentence of the
second paragraph of this section. Please correct this oversight.
48. Page 2-14, Section 2.7.4.3 Residential Wells -Contrary to what the text states, Table 2-3
does not provide general background information about the residential wells. In tact, there
is no table which provides this data.
49. Page 2-14, Section 2.7.4.2, 1st sentence -There are 10 residential wells shown in Figure
2-28, not two as indicated here.
. 6 .
• • 50. Page 2-14, Section 2.7 .4.2 -Delete the last sentence.
51. Page 2-14, Section 2.7.4.3, 2nd paragraph -Need to note that the NC Drinking Water
Standards are also ARARs and also apply. Add the word "on" between "based" and
"Federal".
52. Page 2-15, 1st complete paragraph-In the 4th line, change the word "that'' to "than".
53. Page 2-15, Section 2.7.4, last paragraph, first sentence: This sentence states "VOCs were
not detected at any of the residential wells, except two ... ". This is a very misleading
sentence. Based on the total number of residential wells sampled (according to the data
presented in this document), this equates to 20% of the residential wells as being
adversely impacted.
54. Page 2-16, 4th line -Add the word "eventually" before the word "discharge". In the 2nd full
paragraph, in the eighth line, change the word "recover'' to "recovery".
55. Page 2-17, the first full sentence on this page is incomplete.
56. Page 2-17, Section 2.7.5 Groundwater Modeling -How was it determined that the average
net recharge at this site is 12 inches/year? Please justify this determination, as this is a
critical parameter for the flow model. Normally, unless there is sufficient data to reliably
estimate net recharge, net recharge is considered a calibration parameter. This was not
the case for the flow model developed for this site. In addition, why was the parameter
VCONT not varied during calibration to determine its sensitivity.
57. Page 2-18, Section 2.7.5 Groundwater Modeling -The PRPs statement that the final
calibration model parameter values are all "very reasonable with respect to values
determined from slug tests and pumping tests reported in previous studies" should be
justified with a brief summary of these test results.
58. Page 2-18, first bullet-In the 4th line, add the word "match" after "improved".
59. Page 2-18, Section 2.8 -This section should reference Figure 2-31, instead of Figure 2-27.
60. Figure 2-21 is highly confusing. What is it showing? It says Petroleum Hydrocarbons, yet
it mentions vinyl chloride which is not a petroleum product. In addition, no units are given.
61. Page 3-2, Section 3.1.3.1 -The sixth sentence of the second paragraph of this section is
misleading. This sentence states that groundwater does not flow from the GE Subsite
offsite north of Tabor Road or offsite west of Spartanburg Highway. Although generalized
flow conditions indicate eastwardly flow towards Bat Fork Creek, this statement implies that
groundwater contamination is limited to the GE Subsite. However, groundwater monitoring
has demonstrated that groundwater contamination is not limited to the GE Subsite. Please
either delete this sentence or restate this sentence to indicate that the groundwater level
survey indicates generally eastwardly flow conditions.
62. Page 3-2, Section 3.1.3.1 -The last paragraph of this section states that vertical head
differences were measured at the GE and Shepherd Farm Subsites. However the only
summary of the results of the vertical head testing is given in the last sentence;
groundwater flow is upward near Bat Fork Creek. Please provide a more detailed
summary of the results of the vertical head analysis.
- 7 -
• • 63. Page 3-3, last sentence -Change to "When a majority of the closest residents were
connected ... " In addition, this hypothesis has not taken into account that there are still
quite a few users of the groundwater on the other site of both Bat Fork Creek and Tabor
Rd. Even if the residents don't drink the water, some still use it for other reasons.
Therefore, underflow may still be occurring. Remember that these conclusions are based
on the results of only one well being placed across Tabor Rd and one on the other side of
Bat Fork Creek.
64. Page 3-4, Section 3.1.3.3, 2nd Paragraph -"Too much extraction" is vague. Does this
mean extraction at too high a rate?
65. Page 3-5, Section 3.2.1 -The last sentence of the second paragraph of this section
indicates that, based on the PCB data collection activities conducted in 1994, the depth of
Landfill B is one to three feet below land surface. However, Section 3.2.2 of the
Geophysical Investigation Report (SOIi, 1997) included as Appendix A indicates that the
depth of burial of suspected debris ranged from near surface to five to seven ft bis. Please
clarify this discrepancy. Furthermore, Section 3.2.2 of Appendix A indicates that the GPR
survey is limited to a depth of five to seven ft bis due to conditional limits posed by the soil.
Please provide additional information or clarification as to how to compensate for this
limitation.
66. Page 3-5, Section 3.2.1 -The last sentence of the third paragraph of this section states
that the DSI is disconnected to the water table. Section 3.1.3.1 stated that groundwater
flow generally follows the topography and that recharge is primarily due to the infiltration of
precipitation. Please revise the last sentence of the third paragraph of this section in
accordance with Section 3.1.3.1 unless a demonstration can be made for the prevention of
infiltration at the OSI.
67. Page 3-5, Section 3.2.1 -The last two sentences of this section are incorrect. Although
other significant sources of groundwater contamination have not been discovered to date,
surficial and subsurface investigations have discovered other areas that exhibit chlorinated
and petroleum hydrocarbon contamination. Please revise these last two sentences
accordingly.
68. Page 3-5, Section 3.2.1, third paragraph, third sentence: This sentence states, " ... OSI is
above the water table ... ". Were ground water level measurements collected during both
dry and wet seasonal conditions?
69. Page 3-6, Section 3.2.2 -The second sentence of the second paragraph should state " ...
were detected in surficial soils ... " Please correct this oversight.
70. Page 3-6, Section 3.2.2 -The wells in the 3rd bullet, SWW-9 and SWW-1 O were not
sampled in 1994. How can a comparison be made?
71. Page 3-6, Section 3.3.1 -Please provide copies of all input and output generated during
the voe mass estimate analysis.
72. Page 3-7, Section 3.3.2 -The last sentence of the first paragraph of this section indicates
that voes contamination increased from 1994 to 1997 at two bedrock groundwater
monitoring wells (MW-12 and MW-14). Furthermore, Figures 3-3 through 3-8 indicate that
VOCs may be pooling at the saprolite/bedrock interface. This potential source of
groundwater contamination needs to be thoroughly investigated, including the possibilities
for fractured flow. Please provide a workplan and time frame for investigating this issue.
- 8 -
• • Please revise all other portions of this submittal accordingly, including Section 3.2.1 as
stated above.
73. Page 3-7, Section 3.3.1 (top of page): This paragraph states, " ... PCE comprises 100% of
the voe mass." This statement strongly indicates that no biodegradation is occurring at
Shepard Farm subsite, otherwise a number of sister products would also be detected.
74. Page 3-7, Section 3.3.2, first paragraph: This paragraph refers to Figures 3-3 through 3-8.
Based on the data presented in these figures, it is clear that the vertical delineation of the
plume below the GE Subsite has not been defined. However, no work has been proposed
to address this data gap. The plume does not appear to have decreased in size.
75. Page 3-7, Section 3.3.3 -This natural attenuation study was done as part of the treatability
study and not a part of the RD and therefore should be included with it. Natural
attenuation was not the remedy chosen; analyzing it was merely a tool to determine if in-
situ bioremediation was feasible at this Site. To now make it the remedy is not per the
ROD, CD, or SOW.
76. Page 3-8, Section 3.3.3.1, paragraph at top of page: This paragraph refers to a
comparison between 1994 and 1997 data. Since the results of this comparison are being
referred to here, the comparison should be incorporated into the 30% RD?
77. Page 3-8, Section 3.3.3.1 -The second and third paragraphs of this section provide an
adequate overview analysis of the biodegradation parameters and various trends noticed
at the GE Subsite. However, in order to thoroughly demonstrate the natural attenuation
potential of the GE Subsite (and the Shepherd Farm Subsite), adherence should be made
to the protocol as set forth in the Technical Protocol for Natural Attenuation of Chlorinated
Aliphatic Hydrocarbons in Ground Water (United States Air Force Center for
Environmental Excellence, 1997), Use of Monitored Natural Attenuation at Superfund,
RCRA Corrective Action, and Underground Storage Tank Sites (US EPA, 1997) and Draft
EPA Region IV Suggested Practices for Evaluation of a Site for Natural Attenuation
(Biological Degradation) of Chlorinated Solvents (US EPA, 1997). Please make the
appropriate revisions to this document.
78. Page 3-9, Section 3.3.3.1, top of page: This paragraph refers to data from well clusters,
MW-12 and MW-14, however, the location of this data within the 30% RD is not provided.
79. Page 3-9, 2nd Line -Define "anoxic".
80. Page 3-9, Section 3.3.3.2 -Please provide all input and output generated by the SiteGIS
analysis, TinMass, BIOSCREEN, Buschek and Alcantar analysis, etc. as an appendix to
this section or to Appendix H. Please include all data and figures used or generated during
the course of these analyses.
81. Page 3-10, Section 3.3.3.3 Estimation of Natural Attenuation Rates -The Buscheck and
Alcantar model is typically used to calculate a decay rate for a compound that is completely
decaying (i.e., a compound that is biodegraded as a carbon and energy source). The
biological mechanisms that contribute to biodegradation of these compounds (e.g., BTEX
compounds) are completely different from those contributing to chlorinated solvent
biodegradation. The Buscheck and Alcantar model can not be used for PCE decay unless
the whole dechlorination process is considered. The model does not account for TCE,
DCE, VC and ethene production and decay.
- 9 -
• •
82. Page 3-10, Section 3.3.3.4: It would be very helpful if the findings/predictions of the
BIOSCREEN simulations were also presented in figures.
83. Pages 3-1 O & 3-11, Section 3.3.3.4: An explanation needs to be incorporated into this
section expounding why a concentration of 110 ppb of PCE (Table 2-4) was detected in
monitor well·MW-22A which is approximately 1,500 feet (Figure 2-27) from the source.
The BIOSCREEN predicted the concentration of PCE would be below 1 ppb approximately
1,260 feet from the source in 1 O years.
The prediction does not make sense. The source was removed before 1990. According to
this, the plume should be very close to Bat Fork Creek and almost gone.
84. Page 3-11, Section 3.3.3.5-This entire section should be deleted. To arbitrarily classify
portions of the plume as central and fringe is not per the ROD, especially since "fringe" can
be 200 times the remediation goal. In the 2nd paragraph, Figure 3-1 O did not consider
Well MW-57.
85. Page 3-12, Section 3.6.3.6-As previously mentioned, this natural attenuation study will
only be considered in the context of the treatability study. The 4th bullet should be deleted.
The seventh bullet is erroneous because off-site private wells are contaminated.
86. Page 3-12, Section 3.6.3.6, second bullet: This bullet should reference Table H.2-4, on
page H-55 in Appendix H for this data.
87. Page 3-12, Section 3.6.3.6: Another possible explanation for the decline in concentrations
in the shallow wells is that the contaminants are migrating down vertically as depicted in
the increase in concentrations in contaminants in wells 12-B and 14-B.
88. Page 3-12, Section 3.6.3.6 Summary of Natural Attenuation Study -With respect to the
fourth, sixth, and seventh bullets, it should be noted that another reason the plume has
reached a quasi-steady state is because Bat Fork Creek is a groundwater discharge
boundary for the plume which prevents the plume from migrating any further than it has.
Due to the close proximity of voe contaminated groundwater next to the creek, it is likely
that VOCs have and are currently discharging into this creek, and the lack of measurement
of these voes is due to both dilution and volatilization once they enter the creek. The
PRPs assertion that contaminants in the central plume will not reach the downgradient
boundary (i.e., Bat Fork Creek) must be considered speculative as there are many
uncertainties associated with the natural attenuation analysis performed, and the analysis
was not a conservative analysis. A more thorough and conservative analysis, including an
ecological risk assessment needs to be performed to demonstrate conclusively that the
remedy will be fully protective of the ecological community in Bat Fork Creek.
89. Page 3-12, Section 3.6.3.6 -The second bullet item of this section states that the lack of
oxygen within the central plume indicates the presence of biologica,I activity. Although the
groundwaters low DO content indicates the lack of oxygen, this does not necessarily
indicate the presence of biological activity. Please revise this sentence accordingly.
90. Page 3-12, Section 3.6.3.6 -The seventh bullet item of this section states that no off-site
contamination is present. However, Figures H.2-2, H.2-5, H.2-6 and H.2-9 of Appendix H - .
Natural Attenuation Report actually depict offsite contamination crossing these very
boundaries. This section, as well as all associated modeling, should be revised to reflect
this offsite contamination. As stated previously in both these comments and in the
Preliminary Remedial Design, the inadequacy of the data concerning the horizontal extent
• •
of contamination must be addressed in order to effectively design the groundwater
remediation system.
91. Page 4-1, Section 4: It is difficult to provide a thorough review of this 30% RD without
knowing the objectives specified in the ROD.
92. Page 4-1, Section 4.1 -Delete the last sentence of the first paragraph.
93. Page 4-2, Section 4.1 -Delete the second to the last sentence in the last paragraph.
94. Page 4-3, Section 4.2: Does the excavation of Landfill B trigger any AGRA regulations?
95. Page 4-3, Section 4.2 -The fourth sentence of this section states that subsurface sampling
has indicated that PCB contamination exists only 1 to 3 ft bis. However, Section 3.1.2 of
Appendix A indicates that the suspected buried debris/disturbed soils ranged from near
land surface to 5 to 7 ft bis. Please clarify this discrepancy.
96. Page 4-3, Section 4.5, first sentence: Why is only the SOW referenced here and not the
ROD?
97. Page 4-4, Section 4.5 -The third sentence of the fifth paragraph should state" ... which
specifies requirements which must be met to implement natural attenuation."
98. Page 4-4, first line -Delete "by active or passive natural attenuation methods" from the first
sentence.
99. Page 4-4, Section 4.5, third paragraph: Throughout this 30% RD document, natural
attenuation is referred to as part of the remedy. Is this paragraph in agreement with the
ROD?
100. Page 4-4, Section 4.5, fourth paragraph, second sentence: This sentence refers to a 1997
OSWER directive, however, insufficient information is provided to identify the referenced
directive.
101. Page 4-4, Section 4.5 -Delete all sentences that deal with natural attenuation being the
remedy at the Shepherd Farm Site.
102. Page 4-5, Section 4.5.1, first paragraph: This paragraph states that "The plume on the
whole is stable and no evidence of substantial contaminant migration away from the site is
present". The plume seems to be migrating vertically downward. In fact, the vertical
extent of the plume has not been delineated, refer to Figures 3-1 through 3-8 which are
summarized in Figure 3-9. The delineation of the entire plume needs to be done prior to
deciding the depth of the extraction wells.
103. Page 4-5, Section 4.5.1 -The first paragraph of this section indicates two areas of the
natural attenuation study which need to be addressed. Although this report states
repeatedly that the contamination has not migrated offsite, the need for further
investigation exists at the perimeter of the GE Subsite. The second item that must be
addressed is the potential existence of VOC pooling along the saprolite/bedrock interface
at monitoring wells MW-12 and MW-14. Please revise this section and provide a workplan
and time frame for completion of these items.
-11 -
• •
104. Pages 4-5 & 4-6, Section 4.5.1: The sentence that carries over from page 4-5 to page 4-6
makes no sense.
105. Page 4-6, Section 4.5.1, second paragraph, first sentence: This sentence states,
" ... another alternative ... ". Is this in accordance to the ROD (i.e., Contingency ROD)?
106. Page 4-6, Section 4.5.1 -Although the work conducted to date may indicate the viability of
natural attenuation at this site, the justifications for the proposed remedial action may be
premature. Many of the previous comments regarding the delineation of the plume, the
stability of the plume, the spatial distribution of the contaminants and the biodegradation
parameters, etc must be resolved prior to making any firm conclusions. Please revise this
section accordingly.
107. Page 4-6, Section 4.5.1, third paragraph, fourth sentence: This sentence refers to
monitoring natural attenuation regularly, however, it does not specify what
parameters/criteria.are going to be monitored.
108. Page 4-6, Section 4.5.1, third paragraph, fifth sentence: This sentence states that if voe
concentrations increase greater than 25% for two sampling events, a contingency remedy
will be evaluated. Insufficient detail is provided to make this remotely acceptable. What
factor(s) will be utilized in the evaluation? How long will the contingency remedy be
evaluated before a decision is rendered? What criteria/parameters will be used to foster
the decision? What happens if the two sampling events are 1 or 2 years apart, that is an
unacceptable passage of time before a decision is made.
109. Page 4-6, Section 4.5.1, first sentence -The gpm sustained at each well is missing.
110. Page 4-7, Section 4.5.1, third bullet: The Agency is not in agreement with the statement
that there are no "NAPL source" located on-site. Because the vertical extent of
groundwater contamination has not been defined at the Site, the statement
" ... downgradient receptors would not be adversely impacted" cannot be substantiated.
Without completely delineating the extent of groundwater contamination, the best this
document can assert is that based on data collected to date, "no downgradient receptors
have currently been adversely impacted". However, this statement contradicts the
statement referred to above.
111. Page 4-7, Section 4.5.1, fourth bullet: It could be argued that it is this vertical migration
that is resulting in the decline in concentration in the upper zones of the aquifer. This bullet
states that" ... 1997 data ... clearly demonstrates contaminant attenuation". This statement
cannot be substantiated. The data may indicate that some form of natural attenuation is
occurring, but the data does not confirm its occurrence.
112. Page 4-7, Section 4.5.1, fifth bullet: This bullet refers to daughter breakdown products.
What evidence has GE provided the Agency that these "daughter products" were not
actually used as agents/reagents at the facility. Was PCE the only CAH used at this
facility?
113. Page 4-7, Section 4.5.1, sixth bullet: This bullet does not belong in this list of bullets.
These bullets are to provide justification for the proposed remedial design. The sixth bullet
informs the reader of an ongoing treatability study that GE hopes will promote their position ·
to be allowed to pursue natural attenuation at their site. Therefore, this bullet is deemed
"For Your Information" and does not belong here.
-12 -
• •
114. Page 4-8, Section 4.5.2, paragraph at top of page, second full sentence: This sentence
refers to isopleths based on 1994 and 1997 data. The figures that illustrate these isopleths
need to be referenced.
115. Page 4-8, Section 4.5.2 -All the contaminated soils at Spring Haven were not removed.
Only the top foot of contaminated soil was removed. Please revise accordingly.
116. Page 4-8, Section 4.5.2, last paragraph, last sentence: How quickly would this contingency
plan be implemented?
117. Page 4-8, Section 4.5.2, last paragraph: When were monitor wells 62, 62A, 63, and 63A
installed? If they were installed as part of the 1997 sampling effort, then there is
insufficient data to support the postulation that the Shepard Farm plume is shrinking or that
the leading edge of this plume is not migrating downgradient.
118. Page 5-1, Section 5.1 -This section provides a general background of the asphalt cover.
Please provide detailed information including, but not limited to, the material specifications,
material testing requirements, the installation procedures, quality assurance/quality control
measures to be implemented, durability, specific operation and maintenance requirements,
durability and load bearing capacity. Please use available US EPA guidance as an aid in
developing the cap design.
119. Page 5-2, Section 5.2 -This section states that only one foot of soil is to be dug for the
Landfill B excavation. However, Section 4.2 stated that subsurface sampling indicated
PCB contamination existed 1 to 3 ft bis. Furthermore, Section 3.1.2 of Appendix A
indicates that the suspected buried debris/disturbed soils ranged from near land surface to
5 to 7 ft bis. Please clarify these discrepancies.
120. Page 5-2, Section 5.2 -This section does not address how the potential wetland to the
south of Landfill B is to be addressed. Please correct this oversight.
121. Page 5-2, Section 5.3 -This section provides a general background of the cover design.
Please provide detailed information including, but not limited to, the material specifications,
material testing requirements, the installation procedures, quality assurance/quality control
measures to be implemented, durability, specific operation and maintenance requirements,
durability and load bearing capacity. Please use available US EPA guidance as an aid in
developing the cap design.
122. Page 5-5, Section 5.5: Will any excavation and placement of these soils trigger RCRA? If
not, then a short statement should be incorporated into this section stating why RCRA or
any other regulation (e.g., TSCA) is not applicable. This way the reader knows that this
issue has been considered.
123. Page 6-1, Section 6.1, first paragraph: It would be helpful if a figure showing the capture
zone created by the operation of the AGRS was incorporated into this document.
In addition, it would be helpful if the figure showing the recovery wells contained the
overlay of the plume.
124. Page 6-3, Section 6.2.3 -Why are discharge standards "assumed". They should be either
specified by the NPDES discharge permit or should be the remediation goals.
-13 -
• • 125. Page 6-3, Section 6.2.3 -It is not clear from the description in this section if the treated
groundwater will be pumped to the out fall on Bat Fork Creek or whether it will be gravity
led.
126. Page 6-9, Section 6.3.3.5, AGRS Control Panel Modifications -This section should include
provisions for carbon breakthrough monitoring.
127. Page 6-10, Section 6.4 Performance Monitoring -The monitor wells listed in Table 6-4 for
monitoring "VOCs in Groundwater" are not the same as those presented in Sections 6.5.2
and 7.2. Please explain.
128. Page 6-11, Section 6.5.2, first paragraph, first sentence: Why isn't one of the purposes of
the monitor network to monitor the hydraulic containment of the contaminated
groundwater?
129. Page 6-11, Section 6.5.2, first paragraph: The reader should be referred to Figure 2-25 for
the location of the monitor wells listed in this paragraph.
130. Page 6-11, Section 6.5.2, first paragraph, last sentence: Typo, "MW-20A" should read
"MW-20B". I could not locate a well MW-20A on Figure 2-25.
131. Page 6-11, Section 6.5.2, first paragraph, last sentence: Why aren't wells RW-5 through
RW-8 included in this list of wells to be monitored?
132. Page 6-12, Section 6.5.3, first paragraph, last sentence: It would be helpful if a Figure was
included in this document that the reader was referred to that showed the location of these
monitor wells.
133. Page 6-12, Section 6.5.3 Compliance Well Selection -To provide better compliance
monitoring, it is recommended that MW-2 be added to the compliance well network to
monitor the big gap between MW-3 and MW-27. It is also recommended that a new well
be constructed next to Bat Fork Creek midway between MW-3 and MW-25, and added to
the compliance well network to monitor the big gap between these wells.
134. Page 6-12, Section 6.5.3 -How were these compliance wells determined? MW-22A
already contains 55 ppb PCE, MW-25 contains 120 ppb. No sampling data is available for
MW-27.
135. Page 6-13, Section 6.5.5 -Groundwater sampling will be conducted quarterly for the first
three years, semi-annually for the next two, and annually thereafter. A live-year review will
be conducted five years from the date that the RA starts. At that time, the monitoring
program will be reviewed.
136. Page 6-14, Section 6.6, sixth sentence: This sentence refers to the change out of the bag
filter based on head loss. Will the head loss be specified in future documents?
137. Page 6-14, Section 6.7.2: A short statement should be incorporated into this section
stating why an air permit is not required.
138. Page 6-14, Section 6.7.2-This section states that an air permit has not been required for
the AGRS system and is not anticipated for the full scale system. Please provide details
as to how GE plans to estimate the emissions and determine compliance with federal,
state and local air emission standards.
-14-
• •
139. Page 6-15, Section 6.8.1 Enhanced Bioremediation -If enhanced bioremediation is
determined tci be effective, it should be actively implemented, instead of being a
contingency remedy, to accelerate the remediation of groundwater, per the ROD.
140. Page 7-1, Section 7.1 Monitored Natural Attenuation -With regard to the third bullet, no
wetlands have been identified downgradient of the Shepherd Farm plume, so groundwater
extraction should not affect any wetlands at this subsite. Furthermore, if the extracted
water is discharged into Bat Fork Creek (after treatment) just upstream of the plume, the
impact of groundwater extraction on Bat Fork Creek should be minimal. With regard to the
fourth bullet, please explain why a separate pump-and-treat system could not be
constructed at the Shepherd Farm subsite thus eliminating the need for an extended
pipeline from the GE subsite.
141. Page 7-1, Section 7 .1 -The modeling conducted to date has been an estimate of the
pumping required to contain the contaminant plume. Please bear in mind that this
hydraulic containment will require refinement in the Intermediate Remedial Design.
Furthermore, the issue of wetlands impact needs to be addressed. This section states that
negligible impact may be felt by the wetlands. Please provide further details as to how the
wetlands issues will be addressed.
142. Page 7-1, Section 7 .1, last sentence on page: What is the basis of this statement? Where
is the supporting data documenting that the discharge of treated groundwater to Bat Fork
Creek could adversely impact the creek?
143. Page 7-2, Section 7.2: A Figure should be included in this document showing the location
of these monitoring wells proposed to monitor the plume at the Shepard Farm subsite.
144. Page 7-2, Section 7.2, first paragraph, fourth sentence: Suggest adding monitor well MW-
63 to this list of monitoring wells.
145. Page 7-2, Section 7.2, first paragraph, last sentence: A Figure needs to be included in this
document showing the location of the monitoring point(s) on Bat Fork Creek.
146. Page 7-2, Section 7.2 Monitoring Plan For The Shepherd Farm Subsite -To provide better
compliance monitoring, SWW-10 should be added to the compliance well network as well
as two new wells to be constructed along Bat Fork Creek downgradient of the plume, to
monitor the concentrations of COCs just before entering Bat Fork Creek. In addition, MW-
63 should also be added to the compliance well network to ensure that contamination does
not migrate eastward from the source area without being detected. MW-61 does not need
to be included in the compliance well network as this is a background well which was
previously found to be clean.
The groundwater sampling schedule for the Shepherd Farm subsite should be the same as
for the GE subsite.
147. Page 8-1, Section 8 -The fifth paragraph of this section states that no Intermediate Design
submittal will be made for the containment portion of the remedial design. However, this is
in contrast to Section 5.6 which stated that the design specifications were to be submitted
in the Intermediate Design submittal. Due to the number and type of comments being
submitted at this junction of the project, GE should submit all design specifications with the
Intermediate Design submittal.
-15 -
• •
148. Figure 2-5 -This figure does not agree with the data and figures provided in Appendix B.
In addition to the data provided in Figure 2-5, the data and _tables included in Appendix B
state that: Soil gas sample LFA-8 contained 16.9 micrograms per liter (µg/L) total VOCs;
soil gas sample LFA-11 contained 1.63 µg/L xylenes; soil gas sample LFA-13 contained
1.29 µg/L vinyl chloride and 11.3 µg/L total voes; and, soil gas sample LFA-29 contained
10.7 µg/L total voes. Please correct these oversights. ·
149. Figure 2-7 -This figure lists the PCB sample results rather than the sampling location for
surficial soil samples LFA-3 and LFA-9. Please correct this oversight.
150. Figure 2-8: The title of this figure should read, "Surficial Soil Sample results fgfgg~ at
Landfill A".
151. Figure 2-13 -The PCB results for the surficial soil samples collected at Landfill B, as
· indicated on Figure 2-13, do not agree with the PCB results as provided on Table C-4 in
Appendix C. Please correct this oversight.
152. Figure 2-14-The PCB results for the surficial soil samples collected at the OSI, as
indicated on Figure 2-14, do not agree with the PCB results as provided on Table C-5 in
Appendix C. Please correct this oversight.
153. Figure 2-17: Did not read an explanation in the text as to why there was a gap in the
middle of Shepherd Farm subsite where no samples were collected ..
154. Figure 2-17 -The PCB results for the surficial soil samples collected at the Shepherd Farm
Subsite, as indicated on Figure 2-17, do not agree with the PCB results as provided on
Table C-7 in Appendix C. Please correct this oversight.
155. Figure 2-20 -Replace CH2CL2 with the compound's name.
156. Figure 2-25: On page 6-11, Section 6.5.2, last sentence, well MW-42 is identified as a well
to be monitored for performance. However, in Figure 2-25, well MW-54 and not MW-42 is
shaded green which indicates it is part of performance monitoring scheme. Which is
correct?
157. Figure 2-26 -The values for total voes are incorrect for several groundwater monitoring
wells and groundwater extraction wells. Groundwater monitoring well MW-18 should be
943 parts per billion (ppb) instead of 928 ppb as indicated on Figure 2-26. Furthermore,
the presence of methylene chloride should be included in this figure for those wells that
exhibited methylene chloride. The NC Superfund Section recognizes that methylene
chloride was not identified in the ROD as a contaminant of concern. However, GE must
recognize that all contaminants must be remediated in accordance with Title 15A, Chapter
2L of the North Carolina Administrative Code (15A NCAC 2L). Please revise this figure
accordingly.
158. Figure 2-29 -Since no figure was provided for the total voe concentration in the
groundwater, please place a note in this figure denoting the presence of methylene
chloride at groundwater monitoring well MW-61.
159. Figures 3-3 through 3-8: The usefulness of these figures are compromised for two
reasons. First, in some portions of these figures it is difficult to determine which well is
which. This problem is compounded by the fact that a number of data points presented in
Table 2-4 do not match with the concentrations incorporated into these figures.
-16 -
• •
160. Figure 3-6 -The data for wells MW-12 and MW-14 indicate that contaminants are
migrating vertically downward.
161. Figure 4-1-Does the ROD specify the minimal permeability allowed for the cap? Figure A
(based on the ROD) has a permeability> K::. 10·12 CM/S where as in Figure B (proposed
alternate cap) the permeability is> K::. 1 o•s CM/S. This is a difference of 4 orders of
magnitude.
162. Figure 4-3 -The dashed line needs to be defined.
163. Figure 4-3 -The cones of depression around wells RW-1 through RW-4 should be
included in this figure. If these cones of depression were not confirmed during the AGRS
pumping, what validity is there in the dashed line incorporated in this figure.
164. Figure 5-3 -The thickness and permeability of the geomembrane should be included in
this figure. ·
165. Figure 6-6 -This figure should also indicate the type and placement of the surface seal, the
type of backfill around the solid riser pipe, the type and location of the lower seal, the
screen slot size, etc. Please revise this figure accordingly.
166. Table 2-2 -Typo, there is no "X" for nickel, however, zinc has two "X".
167. Table 2-2 -This table indicates that total PCBs is a remedial design target parameter, that
cis-1,2-DCE and zinc are both remedial design target parameters and National Pollutant
Discharge Elimination System (NPDES) permit target parameters, and that 1,2-
dichloropropane and chromium are neither remedial design target parameters nor NPDES
permit target parameters. However, this information appears to be in conflict with Table 1-
1 . Please clarify these discrepancies.
168. Table 2-4 -The values in this table are double the values in Table C-13. Please explain.
169. Table 6-1 -The ROD specified goal for PCE is 1, not NA. The value listed for lead is also
a NC GW Standard and thus an applicable ARAR.
170. Table 6-2 -This table lists the depths of extraction wells RW-1 through RW-4 and the
proposed depths for extraction wells RW-5 through RW-8. In scrutinizing Figures 3-3
through 3-8, it is apparent that Site contaminants have migrated deeper in the groundwater
than the bottom of the existing or proposed extraction wells. Why isn't the extraction
system addressing this deeper contamination? Refer to previous comments.
I
171. Tables 6-3 and 6-5 -Will the data generated in response to requirements of Table 6-3 be
comparable to the data generated in response to the requirements of Table 6-5. Different
analY1ical procedures are incorporated into these tables.
172. Appendix A -Geophysical Investigation Report (SDII, 1997) The legend for Figures 2
through 7 indicate that some visual distinction exists between the lines used to denote: the
approximate location of the electromagnetic (EM) transect lines; the lines used to denote
the approximate location of the ground penetrating radar (GPR) transect lines; and, the
lines used to denote the approximate location of GPR and EM transect lines combined.
However, the differences between these lines are not readily apparent in Figures 2 through
7. Please clarify this discrepancy.
-17 -
• •
173. Appendix B -Soil Gas Survey Report (Target Environmental Services, Inc., 1997)
Sampling Procedures -This section describes the procedures used to collect the soil gas
samples. Please reference the appropriate United States Environmental Protection
Agency (US EPA) field method of sample collection and provide the appropriate
documentation of compliance with the US EPA field method as well as the quality
assurance/quality controls implemented during sample collection.
174. Appendix B -Analytical Procedures -The first paragraph of this section lists the specific
analytes considered for the US EPA Method 801 O analysis. Please provide justification for
not including 1,2-dichloroethane (1,2-DCA) and vinyl chloride, both remedial design target
parameters, in the US EPA Method 8010 analysis. The NC Superfund Section recognizes
that screening for vinyl chloride was conducted during US EPA Method 8020 FID analysis.
However, although the similar US EPA Method 8021 is used for vinyl chloride analysis, US
EPA Method 8020 is currently not an acceptable method of vinyl chloride analysis. Please
provide the validation study results supporting the implementation of US EPA Method 8020
for vinyl chloride analysis. Additionally; please provide justification for simply screening for
vinyl chloride content according to US EPA Method 8020 and not providing more ·accurate
results by analyzing for vinyl chloride content according to the US EPA Method 8010.
175. Appendix B, Analytical Procedures -The last paragraph of this section indicates that one
sample, MW-25-2, was submitted for off-site gas chromatograph/mass spectrometer
(GC/MS) analysis for vinyl chloride according to EPA 8260m procedures. Please provide
all supporting documentation of this analysis including chain of custody documentation and
laboratory analytical results documentation. In addition to this supporting documentation,
please provide justification for conducting vinyl chloride analysis on a soil gas sample using
a wet chemistry method. Furthermore, please discuss the accuracy, reliability and
implications of the results of the vinyl chloride analyses conducted on sample MW-25-2.
176. Appendix B, QA/QC Table 1 -This table indicates that field control samples 101, 102 and
103 were collected on May 2, 1997. However, all supporting documentation indicates that
field control samples 101, 102 and 103 were collected on May 27, 1997. Furthermore, this
table indicates that field control sample 108 was collected on May 30, 1997. However, all
supporting documentation indicates that field control sample 108 was collected on May 29,
1997. Please clarify these discrepancies.
177. Appendix B, QA/QC Table 2 -This table indicates that field control samples 1 O 1, 102 and
103 were collected on May 2, 1997. However, all supporting documentation indicates that
field control samples 101,102 and 103 were collected on May 27, 1997. Furthermore, this
table indicates that field control sample 108 was collected on May 30, 1997. However, all
supporting documentation indicates that field control sample 108 was collected on May 29,
1997. Please clarify these discrepancies.
178. Appendix B, Figure 3 -GC/FID Results -Petroleum Hydrocarbons -This figure indicates
that soil gas sample LFA-11 contained no detectable concentrations of petroleum
hydrocarbons. However, Table 1 indicates that soil gas sample LFA-11 contained 1.63
µg/L xylenes. Please clarify this discrepancy.
179. Appendix B, Figure 3 -GC/FID Results -Petroleum Hydrocarbons -This figure indicates
that soil gas sample LFA-13 contained 1.29 µg/L vinyl chloride. However, Table 1
indicates that soil gas sample LFA-13 contained no detectable concentrations of vinyl
chloride. Please clarify this discrepancy.
-18 -
• •
180. Appendix B, Target Standard Operating Proced_ures (SOP) -The Target Standard
Operating Procedures (SOPs) include the methodology for conducting the US EPA Method
8010 and US EPA Method 8020 analyses for the given constituents. However,
supplemental analyses for chloroform were conducted in accordance with EPA Method
8010. Although the NC Superfund Section recognizes that this method is acceptable for
the analyses of chloroform, please provide the addendum to the SOP which contain the
additional information and procedures in order to conduct such analyses.
181. Appendix B, Target Standard Operating Procedures (SOP) -The SOP for conducting the
methane analyses by GCffCD was inadvertently omitted. Please correct this oversight.
182. Appendix B, Chain of Custody Forms -The chain of custody forms indicate that custody
seals were not applicable for this project and that the samples were received in good
condition and "cold." Please provide justification for not implementing proper sample
handling procedures, including the use of custody seals and the measurement of sample
temperature upon laboratory receipt.
183. Appendix C, Analytical Results -Table C-2 -Results of Soil voe Investigation at Landfill A
-The quantitation limits used in the analysis of the soil samples collected from Landfill A
are inadequate. In order to demonstrate the presence or absence of contaminants within
the soil of Landfill A, laboratory analysis must be conducted at quantitative limits as
specified by the US EPA Contract' Laboratory Program (CLP). Separate laboratory runs for
the various compounds may be required so that CLP detection limits are maintained for all
analytes. Table C-2 implies that, with the exception of 1,2-DCA and benzene, further voe
contamination does not exist in Landfill A. However, this demonstration cannot be made
due to the elevated laboratory detection limits. Please make sure that all future analyses
are conducted with the appropriate detection limits.
184. Table C-6, Results of Soil voe Investigation -The quantitation limits used in the analysis
of the soil samples collected in the vicinity of monitoring wells MW-14 and MW-25 are
inadequate. In order to demonstrate the presence or absence of contaminants in the
vicinity of monitoring wells MW-14 and MW-25, laboratory analysis must be conducted at
quantitative limits as specified by the US EPA CLP. Separate laboratory runs for the
various compounds may be required so that CLP detection limits are maintained for all
analytes. Table C-6 implies that, with the exception of cis-1, 2-DCA, further voe
contamination does not exist in the vicinity of monitoring wells MW-14 and MW-25.
However, this demonstration cannot be made due to the elevated laboratory detection
limits. Please make sure that all future analyses are conducted with the appropriate
detection limits.
185. Table C-9, Results of Soil VOC Investigation at Shepherd Farm Subsite -The quantitation
limits used in the analysis of the soil samples collected at the Shepherd Farm Subsite are
inadequate. In order to demonstrate the presence or absence of contaminants at the
Shepherd Farm Subsite, laboratory analysis must be conducted at quantitative limits as
specified by the US EPA CLP. Separate laboratory runs for the various compounds may
be required so that CLP detection limits are maintained for all analytes. Table C-9 implies
that voe contamination does not exist in the vicinity of monitoring wells MW-62. However,
in addition to not knowing the depth of sample collection, the elevated laboratory detection
limits make it impossible to conclude as to the actual presence or absence of contaminants
near monitoring well MW-62. Please make sure that all future analyses are conducted with ·
the appropriate detection limits.
-19 -
• •
186. Table C-11, Results of Stream Sediment Samples -The table that was included as Table
C-11 is actually Table C-1 O (Results of Stream Water Samples). Please insert the correct
table into this page.
187. Table C-13, Results of Initial Groundwater Sampling Investigation -This table indicates that
groundwater samples were not collected for SVOC analysis for nine of the groundwater
monitoring wells. Please provide a justification for not collecting these groundwater
samples.
188. Table C-13, Results of Initial Groundwater Sampling Investigation -This table indicates that
water samples were not collected for VOC analysis for residential drinking water well WW-
82. Please provide a justification for not collecting these samples.
189. Appendix D, Page D-1, Section 1 -The last sentence should state " ... except where
noted."
190. Page D-2, Section 2.2 -The third paragraph of this section states that the monitoring wells
were constructed of stainless steel screens and polyvinyl chloride (PVC) casing. The US
EPA SOP states that stainless steel materials should be used if the monitoring program is
to be designed for organic compound analysis. Please provide justification for the use of
PVC casing instead of stainless steel casing.
191. Page D-3, Section 2.2.1 -The first paragraph of this section indicates that saprolite soil
samples were collected for headspace analysis using a split spoon sampler. Please
provide a description of the methodology, description of the equipment used, and results of
the headspace analysis.
192. Page D-3, Section 2.2.1 -The third paragraph of this section states that the stainless steel
well screen was decontaminated according to Method 1 in Section 3.1. However, Section
3.1 states that Method 2 should be used for all materials used to collect samples
undergoing volatile organic compound or trace inorganic constituent analysis. The
Superfund Section recognizes that although the well screen is not a piece of sampling
equipment, the well screen should be handled with the same level of decontamination as
the sampling equipment. Please provide justification for only providing minimal
decontamination of the well screen.
193. Page D-5, Section 2.2.2 -The sixth paragraph of this section states that the stainless steel
well screen was decontaminated according to Method 1 in Section 3.1. However, Section
3.1 states that Method 2 should be used for all materials used to collect samples
undergoing volatile organic compound or trace inorganic constituent analysis. The NC
Superfund Section recognizes that although the well screen is not a piece of sampling
equipment, the well screen should be handled with the same level of decontamination as
the sampling equipment. Please provide justification for only providing minimal
decontamination of the well screen.
194. Page D-11, Section 2.4.3 -The last sentence of the third paragraph of this section should
be stated " ... secondary standards are established for taste and aesthetic purposes and
do not signify an adverse health effect."
195. Page D-11, Section 2.5 -This section states that all of the data, including the quality
assurance/quality control samples and procedures, have been reviewed and validated by a
third party. Please include all laboratory analysis reports, chain of custody documentation,
-20 -
• • quality assurance/quality control documentation, data validation reports and field notes for
all sampling events.
196. Table D.2·1 • This table indicates that monitoring well MW-62, although intended to monitor
the saprolite, was set and screened five feet into the upper bedrock. Please provide
further discussion of why this well was set partially into the bedrock as well as a discussion
of what purpose this will serve the groundwater monitoring program.
197. Figure D.2·3 • This figure indicates a number of perimeter groundwater monitoring wells
located at the GE Subsite that were not sampled during the July 1997 sampling event. In
light of the degree of uncertainty of the contaminant plume boundary, please provide
justification for not conducting additional sampling.
198. Attachment 1 • Well Construction Reports • The well construction report for groundwater
monitoring well MW-62A indicates that a substantial fracture exists at a depth of 50.5 feet
to 54 feet below land surface. However, the bentonite seal was placed at a depth of 50
feet to 54 feet below land surface. Please provide a justification for placing the bentonite
seal at this location and not monitoring this fracture.
199. Appendix E, Groundwater Modeling Report• Please provide all data reports generated
from the use of MODFLOW and MODPATH. These data reports should include all input
and output values and should be included for all iterations of each program. Any figures
used or developed as a result of these modeling efforts should also be included in this
re~~ .
200. Page E· 1, Section E.1 · The second sentence of the first paragraph of this section
references the use of 12 inches per year (in/yr) net recharge for precipitation. Please
provide a reference for this value.
201. Page E-2, Section E.1 · The fourth paragraph of this section discusses the development
and use of the topographic drainage basin. Please provide a figure that graphically
describes the topographic drainage basin and briefly describe the development of the
topographic drainage basin.
202. Page E-2, Section E.1 · The fourth paragraph of this section indicates that drains were
assigned to layer 1 in areas of potential wetlands. Figure E.E-4 indicates the presence of
only one drain. However, the wetland delineation provided in Appendix F indicates the
presence of three wetlands on or adjacent to the GE Subsite. Please correct this
oversight.
203. Page E-2, Section E.1 • The fourth paragraph of this section states that residential wells
were not simulated. The NC Superfund Section recognizes that residential wells remove a
small amount of water in relation to the aquifer as a whole. However, since the
topographic drainage basin is rather small and the reside.ntial well information in the area is
known, this information should be used in the final remedial design.
204. Page E-5, Section E.2 • The fifth paragraph of this section discusses the pumping rates
used to.reproduce the observed drawdowns experienced during the December 1997
aquifer test. Please demonstrate that the pumping rates used for the model calibration
correspond with the actual pumping rates used during the aquifer test.
· 21 ·
• •
205. Page E-5, Section E.2 -The first two bullet items in the fifth paragraph of this section
should be stated as follows:
"Saprolite hydraulic conductivity of approximately 5 ft/d provides a good match with
drawdown data, and provides an improved match in the eastern portion of the GE
Subsite. Hydraulic conductivity of less than approximately 5 ft/d provide an
improved match in the western portion of the GE Subsite.
Bedrock transmissivity of 50 112/d (corresponding to hydraulic conductivity of 0.5 ft/d
and aquifer thickness of 100 ft) provides a good match with drawdown data, and
lower values provide an improved match in the western portion of the GE Subsite."
206. Page E-5, Section E.2 -The last sentence in the fifth paragraph of this section states that
the output values obtained are all very reasonable with respect to values determined from
slug tests and pumping tests reported from previous studies. Please provide, at a
minimum, a summary of the previous tests and demonstrate the "reasonableness" of the
output data.
207. Page E-5, Section E.3 -The third sentence of this section indicates that a detailed
optimization of locations and rates was not performed. Although the NC Superfund Section
recognizes that a detailed optimization may not be prudent at this junction, potential
locations for the recovery wells have already been used for the treatment system design in
Appendix I. Please incorporate the locations of the extraction wells into these model
iterations.
208. Page E-6, Section E.3 -The first sentence of the third paragraph of this section indicates
that 15 in/yr is the likely upper bound of net recharge. Please provide a reference and
justification for the use of this value as an upper bound.
209. Page E-6, Section E.3 -The last sentence of the third paragraph of this section indicates
that a 25 percent safety factor is appropriate for the design calculations. Please provide a
brief justification for the use of this factor of safety.
210. Figure E.E-5 -Please provide a legend which denotes the area of reduced hydraulic
conductivity.
211. Figure E.E-6 -This figure indicates the location of the recovery wells although the figure
depicts the simulated water levels without pumping. Please delete the recovery wells from
this figure, as well as subsequent figures, which depict static conditions without pumping.
212. Appendix F, Wetland Delineation Report (ESC, Ltd., 1998) -Site Reconnaissance -This
section indicates that the jurisdictional boundaries of the wetlands are to be verified by the
United States Army Corps of Engineers (USACE). Please provide a time frame for this
task to be completed. All reports generated by this verification inspection should be
submitted to the North Carolina Department of Environment and Natural Resources (NC
DENR) and EPA.
213. Appendix F -Discussion-This section indicates that GE should perform a metes and
bounds survey following the USACE inspection. Please provide details of GE's plans to
perform the metes and bounds survey.
-22 -
• • 214. Appendix F -Discussion -This section provides a discussion of the steps and
requirements that GE must follow prior to impacting wetland areas. Please provide a
thorough discussion as to how GE plans to anticipate the effects of the groundwater
remediation system and what steps GE plans to take to comply with the wetland impact
permitting requirements.
215. Appendix G, Landfill Modeling Report -This report provides an overview of the Hydrologic
Evaluation of Landfill Performance (HELP) modeling. Please provide the actual input and
output generated by the H.ELP model.
216. Appendix G, Landfill Modeling Report -Differential settlement assumptions need to be
added to each cap design.
217. Appendix H, Natural Attenuation Report All input and output results of the SiteGIS analysis,
TinMass, BIOSCREEN, Buschek and Alcantar analysis, etc. should be provided as an
appendix to this section. Please include all data and figures used or generated during the
course of these analyses.
218. Appendix H, List of Figures -The title for Figure H.2-17 should state "PCE and TVOC
concentrations in groundwater (as measured in 1990 and 1994)." Please correct this
oversight.
219. Page H-9, Appendix H, Section 1.4, Overview of Site Conditions-Although the calculated
hydraulic conductivities of the site ranged from 0.0027 ft/day to 13 ft/day, the model used a
value of 4 It/day (values of 5 ft/day and 2 ft/day were used for sensitivity). Given the wide
ranges of values at the site, a wider range should be used for sensitivity analysis.
220. Page H-9, Section 1.4 -The sixth paragraph of this section states that the hydraulic
conductivity of 4.0 ft/day was used in the natural attenuation modeling of both subsites.
However, the results of the groundwater modeling indicated that a hydraulic conductivity of
5.0 ft/day best represented the site. These values are in contrast to the actual values that
have been exhibited during previous studies. The results of the previous studies should be
compared in much greater detail to these estimates being used for modeling purposes.
Furthermore, the estimates used for modeling purposes should be used consistently
between each type of model. Please revise this section, as well as other affected
sections/modeling iterations, accordingly.
221. Page H-9, Section 1.5 -The justification for using the simpler analytical models that is
provided in this section is incorrect. The NC Superfund Section recognizes that, for
screening purposes, a simpler analytical model may be employed in order to obtain an
approximation of the attenuation rate at a site. However, each model must be evaluated on
the model's merits, not simply on the lack of available data in which to operate a more
complex, yet possibly more accurate, analytical model. Please revise this section
accordingly.
222. Page H-16, Appendix H, Figure H.1-2 -This figure is misleading because it does not
differentiate between abiotic/biotic pathways, major/minor pathways, and anaerobic/aerobic
pathways. The figure should be revised.
223. Page H-19, Section 2.1.1 ~ The first sentence in this section should state "in the
groundwater at the GE Subsite were visualized using SiteGIS." Please correct this
oversight.
-23 -
• •
22~--Page H-19, Section 2.1.1 -The second paragraph of this section describes the creation of
three sets of control points. These control points in effect create the contaminant boundary
conditions. However, there has been no evidence to support the creation of these zero
boundary conditions that have been placed at the two roads and one creek that surround
the GE property. Figures H.2-2, H.2-5, H.2-6 and H.2-9 actually depict offsite
contamination crossing these very boundaries. This section, as well as all associated
modeling, should be revised to reflect this offsite contamination. As stated previously in
both these comments and in the Preliminary Remedial Design, the inadequacy of the data
concerning the horizontal extent of contamination must be addressed in order to effectively
design the groundwater remediation system.
225. Page H-21, Appendix H, second paragraph -The report states "Relatively high levels of
carbon dioxide were observed in monitor wells with high PCE concentration. Carbon
dioxide is the ultimate daughter product of PCE biodegradation." This statement implies
that the detection of CO2 indicates complete degradation of PCE. Although high CO2 is
expected in areas of PCE dechlorination, it would be caused by microbial utilization of
organic carbon and the concomitant use of electron acceptors (oxygen, nitrate, ferric iron,
sulfate). If PCE was degrading to CO2, then DCE concentrations would be higher and VC,
ethene, and methane would be detectable. This would only occur if oxygen was less than
0.7 mg/L, all nitrate and sulfate were utilized, and ferrous iron concentrations were
increased in the area.
226. Page H-22, Appendix H, Section 2.1.3, Estimation of Contaminant Mass in Place -The
report states that the plume has experienced "greater than 20% reductions in PCE and
TCE mass. This statement may be inaccurate for several reasons. The PCE and TCE
averaged concentrations over large grid areas, could alone account for a 20% differential.
The mass calculations did not account for DCE. DCE clearly accounts for some total mass
since it is being produced as part of limited anaerobic dechlorination via biological
mechanisms. In addition, the analyses are based only on dissolved concentrations and do
not account for sorption.
227. Page H-23, Section 2.2.1 -The preliminary analysis conducted in this section is a general
discussion of the attenuation potential of the GE Subsite. However, in order to thoroughly
demonstrate the natural attenuation potential of the GE Subsite (and the Shepherd Farm
Subsite), adherence should be made to the protocol as set forth in the Technical Protocol
for Natural Attenuation of Chlorinated Aliphatic Hydrocarbons in Ground Water (United
States Air Force Center for Environmental Excellence, 1997), Use of Monitored Natural
Attenuation at Superfund, AGRA Corrective Action, and Underground Storage Tank Sites
(US EPA, 1997) and Draft EPA Region IV Suggested Practices for Evaluation of a Site for
Natural Attenuation (Biological Degradation) of Chlorinated Solvents (US EPA, 1997).
Please make the appropriate revisions to this document.
228. Page H-24 , Appendix H, first sentence -Redox measurements reported here (200 to 300
mV) are not supportive of conditions for dechlorination reactions.
229. Page H-26, Appendix H, Predicting Contaminant Transport -The PRPs state that the
BIOSCREEN model predicts that in 10 years the plume will migrate 1260 ft. and will not
reach Bat Fork Creek which is 1400 to 1500 ft. away. The difference in these distances is
10% to 15%. Given that BIOSCREEN is a "screening level model" there could be 10% to
15% variation in results due to the assumptions. In particular, inaccuracies can be
expected in the source term concentrations from MW-11, MW-12, and MW-42 (may
actually be higher), and the hydraulic conductivities, a PCE decay rate that assumes that
-24-
• •
PCE decay is complete. In addition, the assumptions that the aquifer is homogeneous and
isotropic and that molecular diffusion is negligible can cause inaccuracies.
230. Page H-28, Section 2.2.4 -The fifth sentence of the third paragraph of this section states
that the low dissolved oxygen (DO) levels indicate the depletion of oxygen and the
existence of biological activity. Although the groundwater's low DO content indicates the
lack of oxygen, this does not necessarily indicate the presence of biological activity.
Please revise this sentence accordingly.
231. Figure H.2-16 and Figure H.2-17 depict the tetrachloroethene (PCE) concentrations located
at the Shepherd Farm Subsite. Please provide justification for manually drawing the plume
isoconcentration contours. The first paragraph of this section indicates that there is
insufficient data to characterize this subsite. Please provide a workplan and timeframe for
properly characterizing this subsite. Furthermore, please provide a justification for the
isoconcentration contours crossing the. Unnamed Creek after much discussion of the
creeks being a source of groundwater discharge to the surface as well as a hydraulic
boundary for each subsite.
232. Page H-32, Section 2.4 -The second bullet item of this section states that the lack of
oxygen within the central plume indicates the presence of biological activity. Although the
groundwater's low DO content indicates the lack of oxygen, this does not necessarily
indicate the presence of biological activity. Please revise this sentence accordingly.
-25 -