Consulting • Engineering • Remediation 
35 Nagog Park
Acton, MA 01 720
April 17, 1995 
(508)635-9500
FAX (508) 635-91 80
ENSR Ref. No: 
7218-001
ENSR Doc. No: 
55-DPG-522
Ms. Caroline Kwan
Emergency and Remedial Response Division
U.S. Environmental Protection Agency
26 Federal Plaza
New York, NY 10278
Re: 
Comments on Tutu Site FS
Dear Ms. Kwan:
This letter provides comments on the draft Feasibility Study (FS) report recently prepared by
Geraghty & Miller on the behalf of the Tutu Environmental Investigation Committee (TEIC). This
comment letter was prepared at the request of Western Auto and their attorneys. Our comments
consist of three general comments followed by a number of specific comments.
General Comments
• 
Issuance of the FS without finalization of the Rl may be premature.
The FS relies on information in the Rl, however the final Rl has not yet been made available. We
are very interested in reviewing the final Rl because the draft Rl that was submitted to EPA
clearly disregarded Western Auto as a source of groundwater contamination. If the 99FS intends
to include Western Auto as partially responsible for the groundwater remediation (see comment
below), then some of the information or data interpretation in the Rl must have apparently
changed between the draft and final versions.
• 
There is a lack of correlation between the information and analyses in the Risk
Assessment and the conclusions drawn in the FS.
There does not appear to be any discernable connection between the risk assessment that was
performed by CDM Federal Programs (discussed in FS Section 2.4) and the soil cleanup
alternatives presented in the FS. For example, Western Auto is targeted for soil cleanup in the
FS based on soil concentrations above the NYS Technical and Administrative Guidance
Memorandum (NYS TAGM) levels for xylenes, benzo(b)fluoranthene, chrysene, and napthalene.
(Sample SS-1 was also above NYS TAGM levels for benzene and toluene, however as explained
below, SS-1 is more representative of a product sample than a soil sample.) However, the Risk
Assessment has determined that none of these parameters are Chemicals of Potential Concern
for groundwater contamination (Risk Assessment Table 2-23). In addition, the Risk Assessment
does not even mention Western Auto even though, just as the FS, the information in the Phase
II Rl was apparently reviewed and evaluated in the development of the Risk Assessment. As
TUT 
006 
2148
*64931*
64931
/""" 
Ms. Caroline Kwan
April 17, 1995
Page 2
discussed further below, we believe that the risk assessment approach, and the use of site-
specific, risk-based cleanup criteria are appropriate for this site.
• 
Soil cleanup values should be based on a site-specific risk assessment.
Risk-based cleanup goals based on site-specific fate and transport modeling should be used for
the soil cleanup values. The FS directly applies the NYS TAGM values as cleanup levels for the
Tutu site. The NYS TAGM values may be appropriate as screening levels. However, the NYS
TAGM procedure is largely generic and not site specific, and therefore is not appropriate for
developing risk-based cleanup values. EPA's general comment #2 in the comment letter on the
draft Rl (Peterson to McBurney and Seibel, 12/28/94) agrees with this and specifically states that
"ERA will be conducting more detailed fate and transport modeling to determine the cleanup
goals".
Specific Comments
• 
Soil quality behind Western Auto is improperly characterized.
- 
Tne paragraph on page 2-17 describing soil quality behind Western Auto is not representative
of actual soil conditions. The paragraph contains numerous instances where data are either
incorrect or improperly used. Ethylbenzene, for example, is quoted as ranging from 16 to 1700
ug/kg in the vicinity of the diesel tank. In fact, no sample was found to contain more than 340
ug/kg. The value of 1700 ug/kg is apparently the same misquote of an Arthur D. Little report
that was made in Figure 5-9 of the draft Rl. The ADL investigation found only 16.9 ug/kg of
ethylbenzene in its North Tank sample.
At the waste oil tank, benzene and toluene are stated to range from 3 to 140 ug/kg and 2 to
3900 ug/kg, respectively. There are no soil samples which were found to have benzene at 3
ug/kg or toluene at 2 ug/kg. There are, however, liquid samples which were found to have
those values. It can only be assumed that the liquid samples were mistakenly used to
characterize soil conditions.
Some analytical findings are misrepresented in the paragraph by the omission of data qualifiers.
In so doing, the reader is lead to believe that all values are precise when, in fact, many were
originally marked by the laboratories as being estimated below detection limits. Examples are
the values given for benzene and toluene at the diesel tank which are not qualified in the FS
even though all the original data are "J" qualified by the analytical laboratory.
Many of the laboratory results listed in this paragraph, are from samples that were collected in
the gravel layer or near the leaking PVC pipe. These sources have no connection to Western
Auto and samples from the sources are not representative of the soil quality behind the Western
/"-^ 
Auto store.
TUT 
QO6 2149
' t
/""^ 
Ms. Caroline Kwan
April 17, 1995
Page 3
Finally, many of the maximum values in the referenced paragraph are associated with sample
SS-1. This sample is presented as a soil sample but should instead be considered a product
sample. Essentially, it is a sample of product that leaked out of the waste oil vent pipe after the
pipe was inadvertently ruptured during the tank removal. It is misleading to present it as
representative of soil conditions at the site, especially in light of the fact that all soil with which
the leaking product came in to contact was removed prior to backfilling the tank graves. It
should also be noted that sample SS-1 was a split sample of sample T2-AS and was incorrectly
placed on Figure 5-9 of the Rl. The actual location is shown on the enclosed figure that shows
the correct locations of the samples taken behind the Western Auto store.
A more complete listing of the data inaccuracies contained in the referenced paragraph are
presented in Table 1.
• 
Results of the Gore-Sorber study are not included in the FS, though the results are
critical to identification of source(s) of petroleum hydrocarbons encountered behind
Western Auto.
ENSR previously submitted reports on the Gore-Sorber methodology and results of two surveys
in its November 14,1994, Comments on Draft Phase II Investigations. Since this submission,
.^, 
w L Gore has prepared a combined map presenting data from both surveys (11/93 and 7/94).
These maps are attached. A clearer presentation of these combined data can now be used to
better identify potential source(s) under the Four Winds Plaza building. Also, hot-spots
upgradient of Western Auto and identified on the maps have been evaluated together with the
petroleum fingerprint analysis of samples (described below) taken during the tank closure. A 4-
inch pipe was encountered during the tank excavation that was found to contain petroleum
hydrocarbon contamination consisting of a mid-range distillate (e.g. diesel) and heavy weight oil
(e.g. motor oil) and numerous leaky joints. This pipe, which was determined to be associated
with a Four Winds Plaza cistern and not either of the USTs behind Western Auto, was traced
upgradient from Western Auto under the Four Winds Plaza building. Gore-Sorber screening
modules indicate hot spots of petroleum organic vapors at the terminus of the pipe tracing
(sorber module #101615) and even further upgradient (at sorber #104178) under the Cost-U-
Less store. Composition of the vapors detected by the sorber is consistent with the constituents
found in the 4-inch PVC pipe. It is recommended that the 4-inch pipe be traced further and
evaluation made as to the ultimate source(s) of the petroleum product. The product source(s)
will act to contaminate downgradient areas, including Western Auto, until it has been remediated.
The information previously presented on the Gore-Sorber studies at the Four Winds property
have been almost completely ignored in the FS. The only reference to the Gore-Sorber studies
(FS Page 2-16, Paragraph 1) states that"... potential impacts to soil cannot be ruled out due to
the detection of elevated soil gas concentrations*. Despite this, no soil remediation for Four
Winds (or other mention of the Gore-Sorber studies) is presented in the FS. Geraghty and Miller
and de maximis were both provided with copies of the Gore-Sorber study reports. In our
^ 
comments on the preliminary draft Rl that were provided to de maximis, we stated that the Gore-
Sorber results should be incorporated into the Rl. We repeated this verbally at a PRP meeting
in September 1994 and at the PRP/EPA meeting in St. Thomas in November 1994. EPA
TUT 
OO6 
2150
/-v 
Ms. Caroline Kwan
April 17, 1995
Page 4
apparently agreed with the importance of this information. The EPA comment letter on the draft
Rl and Development and Screening of Remedial Alternatives report states that the Gore-Sorber
data should be incorporated into the Rl (Rl General Comment #8 and Specific Comment on Rl
Page 1-13) and the FS (General Comment #2 on the Development and Screening of Remedial
Alternatives report). Given the above, the omission of the Gore-Sorber data from the FS is
disturbing.
The FS proposes soil remediation at the Curriculum Center based on soil gas data (FS, Section
4.1.1.3). The Gore-Sorber data not only are not used in a similar manner but are essentially
omitted from the report. This treatment of the two different data sets is inconsistent.
• 
The FS does not consider information in the Western Auto Tank Closure Report that
demonstrates the source of the contamination behind Western Auto.
The FS fails to recognize some of the main findings of ENSR's sampling and analysis conducted
during the fall of 1993 and reported in the Tank Closure Report of June 1994, corroborated by
the Gore-Sorber soil gas survey of July, 1994.
These findings are presented in detail in the Tank Closure Report (Sections 2.5.1 and 2.5.2
y\ 
[pages 2-15 and 2-16]; sections 3.4 and 3.5 [pages 3-13 through 3-28]; section 4.0 [pages 4-1
and 4-2]; and Appendix G and are summarized below.
although the oil in the gravel layer and the oil in the waste oil tank are similar in
that they can both be characterized as a heavy range lubricating oil (e.g., a motor
oil), the evidence indicates that the oil in the gravel layer did not originate from the
tank,
there is no transport mechanism for moving the oil from the tank upwards into the
gravel layer that would not have left a smear zone from the tank to the gravel
layer -such a smear zone was looked for and not observed,
PAH ratios of oil from the tank and from the pipe leading from the bin at Western
Auto to the tank were compared to the same PAH ratios in oily soil samples
collected from the gravel layer and they were found to be markedly different,
PAH ratios of the oil and the bin pipe at Western Auto are also markedly different
than the PAH ratios of the samples in the tank grave (T2-1S, T2-2S, T2-3S, T2-4S),
indicating that the oil in the tank grave did not come from the tank but from the
same source as the oil in the gravel and the 4-inch PVC pipe,
the 4-inch PVC pipe that was found during the excavation was found to contain
a heavy weight lubricating oil and found to have numerous leaky joints in its
-— 
traverse behind the Western Auto store,
oil was observed to flow out of the 4-inch PVC pipe when the pipe was broken,
TUT 
006 
2151
EIKR
Ms. Caroline Kwan
April 17, 1995
Page 5
the 4-inch PVC pipe was traced and found to extend at least 100' (the length of
the tracing device) north upgradient under the Four Winds Plaza building,
the Gore-Sorber survey results indicate a hot spot of petroleum hydrocarbon
contamination upgradient under the Four Winds Plaza building.
The conclusion that can be reached from this information is that the petroleum hydrocarbon
contamination found in the gravel layer, soils, and tank graves behind the Western Auto store
did not originate in the former USTs.
• 
We agree with the FS observation that at Western Auto a gravel layer with perched
water overlies a clay layer.
We agree with the sentences on FS page 2-5, end of the first paragraph "...Saturated alluvial
deposits most commonly occur where bedrock surfaces were incised by former streams and
later filled with sediments or fill. An exception to this condition was observed at Western Auto,
where alluvial deposits were only saturated in a perched zone within a gravel layer overlying a
clay layer". Hydraulic conductivity measurements made of the clay layer (Closure Report, June
1994) indicate an extremely low value of 5 x 10~
9 cm/sec, at a depth of 3 feet below the ground.
This value represents a condition which is impermeable to water flow. For example, a value of
10"6 cm/sec is considered an acceptable specification value for a sheet pile wall made of steel.
This type of wall is considered for use at sites where surrounding of an area is required, and
complete isolation of inside water versus outside water is required (contaminated versus non-
contaminated zones). Both Western Auto tanks were installed directly in this clay without any
stone backfill. Also, clay was used to backfill over the tanks, thus providing a near complete seal
of the tank from the surrounding environment. Attached is a figure showing a cross-sectional
view of the waste oil UST and the surrounding clay layer.
• 
Western Auto should not be considered a PRP for the groundwater contamination.
From review of the FS it is unclear whether Western Auto has been specified as a responsible
party for the groundwater remediation. There is considerable data in the Rl, as well as statements
by several of the involved parties (including Geraghty and Miller) that indicate that Western Auto
has had no adverse impact on groundwater quality. If the FS has used this information to
conclude that Western Auto is not responsible for groundwater contamination, we agree with that
conclusion. The most conclusive information demonstrating that Western Auto did not contribute
to groundwater contamination is the analytical data from monitoring well MW-24 (Rl, Figure 5-17),
adjacent to the former waste oil UST location, which showed non-detects for all parameters
except for insignificantly low and estimated (1J) levels of chlorinated VOCs. These levels are
consistent with and attributable to the edge of the overall chlorinated VOC plume in the Tutu
aquifer. In addition, the available geologic and hydrogeologic information indicate that the
subsurface material in the vicinity of the former USTs is composed of low permeability clays
underlain by relatively unfractured bedrock. Because of this, there is a very poor hydrogeologic
connection between the former UST location and the primary area of the Tutu aquifer, which is
located to the east of the Four Winds building.
TUT 
O06 
21.52
t •
Ms. Caroline Kwan
April 17, 1995
Page 6
The Rl in Sections 5.2.1 and 5.2.6.1 identified the Curriculum Center, O'Henry Dry Cleaners,
Texaco Tutu Service Station, Esso Tutu Service Station and Kentucky Fried Chicken as source
areas for contamination of the Tutu aquifer. Based on the existing data, Western Auto was not
included on this list. Since EPA did not comment on this list in their comment letter on the draft
Rl (Peterson to McBurney and Seibel, 12/28/94), it may be concluded that EPA agrees with this
assessment. In addition, it is apparent that Western Auto is not considered to be a contributor
to groundwater contamination in the FS since the capture zone for the proposed recovery wells
(FS, Figure 4-7) does not extend over to the former UST location. On the other hand, the FS
does identify the Curriculum Center and O'Henry Dry Cleaners as potential groundwater
contamination source areas (FS, Section 4.2), and Texaco Tutu and Esso Tutu Service Stations
as locations with groundwater contamination requiring groundwater IRMs.
If the FS is interpreted to state that Western Auto will be held partially responsible for the
groundwater remediation, we strongly object to that interpretation. There are several PRPs (listed
above and in the Rl) that apparently have contributed to the groundwater contamination of the
Tutu aquifer, however Western Auto is not one of these PRPs.
• 
The preferred remedial alternatives presented in the FS require Western Auto to
excavate and dispose of soil which has already been removed from the site.
The FS requires the removal of 133 cubic yards of soil from behind Western Auto. This
requirement is based entirely on samples collected from soil which no longer exists at the site.
The only samples cited in the FS as the basis for the soil removal requirement were collected
after removal of the USTs, but prior to the final excavation of the tank graves. Before the tank
graves were refilled with clean soil, all visibly contaminated soil was removed and additional soil
samples were collected from the walls and floor of each grave (WOT-1, WOT-2, WOT-3, WOT-4,
DT-1, DT-2, DT-3, DT-4). The analytical results of these samples show significantly lower levels
of petroleum hydrocarbons than the earlier samples, yet were not included in the FS. In
addition, the FS does not recognize that any soil has been removed from the site.
These data were first presented to EPA and Qeraghty & Miller in June 1994 via the Closure
Report for Underground Storage Tanks (ENSR.1994). They were omitted from the first draft of
the Rl and, despite the omission having been pointed out in our comments, were again omitted
in the second draft. Our comments on the second draft also addressed this omission.
Other data, collected after the tank closure, also exist which indicate that the soil behind Western
Auto is substantially cleaner than stated in either the Rl or the FS. Analysis of soil samples
collected during the installation of MW-24 showed that BNAs were below detection limits (except
for some phthalates which were likely due to lab contamination) and that BTEX values were very
low. The results of the Gore-Sorber studies also indicate that post-excavation hydrocarbon
values in the vicinity are low, generally in the vicinity of background levels.
006 
2153
Ms. Caroline Kwan
April 17, 1995
Page 7
• 
The extent of the proposed excavation area and volume for Western Auto in Table
4-1 is not explained or Justified.
Table 4-1 of the FS indicates that two separate volumes of soil, totaling 133 cubic yards, must
be removed from behind Western Auto. There is no explanation of how the depth and areal
extent of the excavations were determined. Presumably they are tied to the results of soil
sampling, however, this does not appear to be the case. The area around and including MW-24
has been targeted for excavation in the FS despite soil sampling results which show compounds
at many times lower than the NY TAGM levels. Large areas to the west of each tank grave,
where no soil samples were taken, are also targeted for excavation.
Further incongruities exist between Table 4-1 and Figure 4-1. The areal extent of soil given in
Table 4-1 and the areal extent of proposed soil remediation in Figure 4-1 do not agree. Table
4-1 indicates that one area of 375 ft.2 and another of 300 ft.2 should be excavated, while Figure
4-1 depicts one area of 600 ft.2 (for the waste oil tank) and another area of 150 ft.2 (for the diesel
tank).
Finally, Section 4.1.1.1 states that excavation at Western Auto should include the PVC pipe and
the oily gravel layer. These sources have no connection to Western Auto and should not be
included in the Western Auto section. The discovery of these sources was fully documented in
the Tank Closure Report, as was the fact that contaminants emanating from these sources could
not have originated from Western Auto. This point was reiterated in ENSR's comments on the
Rl (Galya and Bierschenkto Kwan, 11/14/94).
• 
The FS employs an inconsistent use of the IRM concept.
The presence of BNA contaminants in the former UST location is used to include Western Auto
in site-wide remediation scenarios. Even if there is some minor level of contamination in the
former UST location, it is localized and should not be viewed in the context of a site-wide
remediation effort. Section 4.1.2 of the FS discusses localized groundwater interim remedial
measures (IRMs) to be implemented in the vicinity of the Esso and Texaco service stations and
that these IRMs will be implemented separately from the FS. If such a localized IRM that is
separate from the FS is appropriate for groundwater remediation, it is even more appropriate for
soils. It should be noted that, as stated above, a soil removal IRM for Western Auto has already
been completed. The FS should acknowledge and discuss this.
• 
The discussion on contamination at individual PRP sites is inconsistent.
FS Section 2.2.1.2 is intended to discuss soil contamination at each of the individual PRP
properties. For most of the PRPs, such as Texaco and Esso this is the case and only soil data
is presented in the section. However, the paragraph in this section on Western Auto also
includes analytical results for product samples. If product sample results are to be included for
Western Auto, similar data should be presented for all the PRPs.
TUT 
006 2154
ENCR
Ms. Caroline Kwan
April 17, 1995
Pages
• 
We agree with the conclusion that Western Auto does not have soils impacted with
chlorinated VOCs that require remediation.
Section 2.2.1.1 of the FS identifies PRP properties with soils containing chlorinated VOCs above
NYS TAQM values as locations that might require soil remediation. Western Auto is not identified
as one of these PRP properties. This is appropriate because Western Auto does not have any
samples above the NYS TAGM values for chlorinated VOCs, and because:
(1) The only sample with any detectable concentration of a chlorinated solvent was
sample 
SS-1,
(2) Sample SS-1 was essentially a product sample that leaked out of the waste oil vent
pipe 
after the pipe was inadvertently ruptured during the tank removal (as discussed above)
(3) ENSR's split (sample T2-AS) of sample SS-1 did not indicate the presence of any
chlorinated VOCs,
(4) There were significant laboratory control concerns with the analysis of sample SS-1,
and
(5) No chlorinated VOCs were detected in more than 10 additional samples taken in the
vicinity of sample SS-1.
Additional details on items 3,4, and 5 can be found in ENSR's response document on the Rl
(Qalya and Bierschenk to Kwan, 11/14/94).
• 
No documentation is provided for the elimination of soil remediation alternative SRA
2.
It is stated (FS, page 3-8) that soil remediation alternative SRA 2 (institutional controls/capping)
was eliminated during screening. The reason for this elimination is not but should be given.
In conclusion, we would like to re-emphasize several important points with regard to the FS. First
we agree with the conclusion that remediation of soil in the vicinity of the former USTs is not
required for chlorinated VOCs. Second, we disagree that remediation of soils in the Western Auto
area is necessary for petroleum related compounds. This conclusion is based cleanup criteria
that are generic and not risk-based, and samples from gravel and soils that were removed from
the site during the UST removal and closure. If a re-analysis was performed using site-specific
risk-based cleanup criteria and only soils remaining at the site were considered, we believe that
TUT 
O06 2155
Ms. Caroline Kwan
April 17, 1995
Page 9
it would be concluded that no further remediation of soils in the former LIST vicinity would be
required. Finally, though we have no specific objection to any of the groundwater remediation
alternatives evaluated in the FS, we would like to re-emphasize that, in our opinion, Western Auto
did not contribute to the contamination of the Tutu aquifer and should not be responsible for
groundwater remediation.
Thank you for this opportunity to provide EPA with comments on the FS. Please feel free to
contact us with any questions that you may have on our comments.
Sincerely,
Donald P. Galya, P.E.
Program Director
John Bierschenk
Senior Remediation Specialist
cc: 
E. Butler
G. DelVecchio
TUT 
OO6 2156