USVI Public Records

Document text

Post-tt" brand tax transmittai memo 7671 j • ot page* •> #~ Dept. DPNR COMMENTS ON THB DRAFT REMEDIAL INVESTIGATION FOR THB TUTU WELLFIELD SUPERPUND SITE IN ST. THOMAS, USVI General CommentB Data resentation The discussions of work performed by "other consultants" throughout the document is misleading because Geraghty & Miller Inc. has chosen to selectively include and exclude data from various sampling events from the report. As a result, the entire representation of the data and the assumptions and conclusions drawn from the data is biased and potentially incorrect . The RI report must include all available data collected at the site. The following represents instances where data has been excluded to the detriment of the report: Overall site: The exclusion of the majority of CDM (1989) generated soil data in the discussion and on the figures for each property because the locations of the sampling could not be pinpointed is unacceptable. The exclusion of this and other soil data severely hampers both the credibility and the accuracy of this report. The sampling locations must be obtained and the data must be added to the report . Esso Service Station: A November 1987 mult i- site soil vapor investigation conducted by Geosciencefi Consultants, Ltd. for Texaco Caribbean, Inc. h soil sampling and underground storage tank removal conducted by Soil Tech and reported in January 1990. Discussion of the January 1992 soil gas survey conducted by Target Environmental Services, Inc. at Four Winds Shopping center, especially in the northern portion of the Tutu Esso station. kn April 1994 ground-water sampling event conducted by Blasland Bouck and Lee, Inc. Four winds Plaza: water sampling of the 4 inch pvc pipe located under the western side of Four winds in November 1993 by Caribbean Hydrotech. TUT OO6 O533 *64619* 64619 DENR COMMENTS ON THB DRAFT REMEDIAL INVESTIGATION FOR THB TDTU WELLPIBLD SUPSRFDND SITS IN 6T. THOMAS, USVI General comments: Da^a presentation ajid analysis; The discussions of work performed by "other consultants" throughout the document is misleading because Geraghty & Miller inc. has chosen to selectively include and exclude data from various sampling events from the report. As a result, the entire representation of the data and the assumptions and conclusions drawn from the data is biased and potentially incorrect. The Rl report must include all available data collected at the site. The following represents instances where data has been excluded to the detriment of the report: Overall site: The exclusion of the majority of COM (1989) generated soil data in the discussion and on the figures for each property because the locations of the sampling could not be pinpointed is unacceptable. The exclusion of this and other soil data severely hampers both the credibility and the accuracy of this report. The sampling locations must be obtained and the data must be added to the report. ESSO Service Station: A November 199? multi-site soil vapor investigation conducted by Geoeciences Consultants, Ltd. for Texaco Caribbean, Inc. A soil sampling and underground storage tank removal conducted by Soil Tech and reported in January 1990. Discussion of the January 1992 soil gas survey conducted by Target Environmental Services, Inc. at Pour Winds Shopping Center, especially in the northern portion of the Tutu Esso station. An April 1994 ground-water sampling event conducted by Blasland BoucJc and Lee, Inc. Four winds Plaza: water sampling of the 4 inch pvc pipe located under the western side of Four winds in November 1993 by Caribbean Hydrotech. TUT 006 0534 November 30, 1994 viDPNR's Tutu RI Comments Soil sampling of an underground vault within Cost-U-Less in December 1993 by Caribbean Hydrotech. An expanded Gore-Serber study performed in July 1994 by ENSR under the Four winds building. 0'Henry Dry Cleaners: Slug tests performed on O'Henry wells. ft_ngftd.for a supplemental hYdroereoloaie investigation: Currently, the existing hydrogeologic database used for the understanding of the hydrodynamics of the Tutu valley aquifer system is lacking in certain areas. As a result the RI does not adequately address the hydrogeology of the entire site and provides a stunted understanding of the hydrodynamics of the contaminant plume. Ground water elevation data for both the shallow and deep aquifer zones is severely lacking in the southern part of the site. Ground water elevation data for the KFC-l, KFC-2, Karthman- wilfred well, Harthman-Bakery, CHT-5, EA Corp. #1, EA Corp. #2, EA Corp. #3, EA Corp. #4, Rodriguez, New Gassett, REMW-2, REMW-3, REMW-4, LaPlace, MW-21D, MW-22D, Mathias, Delegarde and Smith supply wells and the Fort Mylner and Versailles wells should be collected, in addition, monitoring wells should be installed in the lots south-west of O1Henry Cleaners and in the pasture southeast of the Fort Mylner Shopping Center in the vicinity of "Fort Mylner." This information is critical for understanding the nature and direction of ground-water flow to the south. The extent of the plume to the south, southeast and southwest as well as north has not been determined adequately. Alternate contaminant sources could be located north of the Curriculum center and VIHA and in the area of Kentucky Fried Chicken (KFC) to the south. The installation of additional monitoring wells or the institution of a temporary well point program is necessary at the site prior to the completion of the Remedial Design to define these areas. At a minimum, the investigation of the Tutu aquifer needs to be extended r.o include the known and suspected gasoline contamination emanating from the vicinity of KFC, formally Home Petroleum, and from Rodriguez Esso. Existing ground water data show MTBE as well as chlorinated hydrocarbon contamination reaching at least as far southeasterly as the Delegarde well. It TUT OO6 O535 November 30, 1994 s Tutu Rl Comments is not clear, given the lack of information between the Delegarde well and the Fort Mylner mall, whether this contamination is coming solely from the sites to the north (ESSO, O'Henry, etc.) or is also coming from Fort Mvlner and Rodriguez or the contamination is combined from both sites. The ground water component associated with Turpentine Run, i.e. that portion of ground wacer flow that is coincident with surface topography and the stream bed of Turpentine Run, has not been investigated to any extent. Ground water levels taken during 1990 investigation of the former Home Petroleum station (present KFC location) indicate a shallow ground water gradient in the direction of Turpentine Run. It is very likely that there is a co-mingling of contamination plumes in this vicinity as it is probable that ground water flow ia structurally as well as topographically controlled. This possibility must be considered in the design phase of the remedial strategy. Ground- water contamination should also be investigated further along the eastern portion of the plume to define its easternmost extent, preferably in the area of Tillet Gardens. Pum A larger scale pump test should be performed which monitors a wider array of wells in differing vertical and horizontal locations. This pump test will provide greater coverage, more extensive site wide aquifer characteristics. This information is needed especially west and east of Route 3d in the center of the site as it is now defined. The information can be used for ground-water flow modeling that must be completed during the Remedial Design. o - contanation The methodology used to discuss source soil contamination and determine the origin of ground-water contamination is unclear and implemented inconsistently. It is DPNR's position that if the contaminant was historically used, stored or disposed of at a facility then that facility is a potential source property. This is regardless of the presence or absence of a contaminated soil sample on the property. DPKR feels, in the case of many properties, that insufficient soil samples have been collected and analyzed to adequately determine if that property is unquestionably not a source. TUT 006 Ob-56 M O V - -5 O - '^ "* W E T . 1 e. : 3 -5- E H V I . R R O T I O M November 30, 1994 viDPNR's Tutu RI Comments DPNR is particularly in disagreement with the assumptions that Geraghty & Miller used in the soils discussion. These assumptions have severely biased the interpretation of the soil data presented in the RI report. The following assumptions listed in Section 5.1 are disputed: Values of vocs in soil samples that are less than 10 ppb do not indicate a release or source of vocs. Instead these values are attributable to ground-water contamination: This assumption may be valid in some instances but almost surely not in all instances. It applies ail arbitrary concentration instead of a risk based level to discuss or dismiss an area's potential for remediation. A more valid concentration should be generated through ground-water fate and transport modeling for each contaminant to determine which areas should be recognized as sources requiring potential remediation. If a soil sample displays a low level concentration of VOCs then it may indicate a nearby source and additional investigation is necessary to confirm the absence of nearby contamination. If no soil sampling was completed at a property but soil gas sampling was negative then the property has not been imp "ted by contamination: No property that is considered a potenti . source can be declared free of impact unless soil sampling is used in combination with soil gas. A soil gas survey alone is not sufficient to rule out a property as a source. The determination of SNA impact on soils is not based on concentration alone and must be based on other factors: The description of these factors, how they weighted and how the determination of impact is made must be clarified. if contaminants are found only in an excavated soil pile then the property is not considered impacted by that contaminant: If an excavated soil pile is contaminated then it has likely originated from the property it presently is located on. This indicates that the property has contained contaminated in-situ soils at one time. In the least, extensive sampling should take place to determ-ne if the in-situ soils in the area of the waste pile have riot been impacted. Isolated samples that are contaminated but surrounded by samples that are non-detect then soil on the property has not been impacted: This assumption is too general to be implemented across the entire site. The spatial distribution of the surrounding samples and the concentration of the contaminated sample must determine whether additional sampling is needed, the TUT OOfe O537 November 3D, 1994 VIDPNR'S Tutu RI Comments property is a source area or the soil has not been impacted. For metals and other inorganics, impact to soil for a particular compound was considered if the analytical result exceeded the background value by at least one order of magnitude: This assumption places an arbitrary distinction on what is considered an inorganic impact to soil. A risk generated concentration or a statistically modelled (based on background samples) cleanup goal should be used for each inorganic instead of this meaningless designation. Adequacy of soils data: Baaed on the presentation in the draft RI report the available soils data is sporadic and was collected in a haphazard uncoordinated manner. Currently, there may not be adequate soils data for the entire site as a whole to prepare a Record of Decision (ROD). DPNR recommends the implementation of a supplemental Remedial Investigation for soils. This investigation would include expanded sampling at the Curriculum Center, VIHA, Tillet Gardens, Antilles Auto, Esso Service station, western Auto, Rodriguez Service Station and Kentucky Fried Chicken, sampling should include the necessary parameters to address the contaminants of concern and to undertake a fate and transport modeling effort. This data would be used in concert with previously collected data to determine cleanup goals for soils based on future impact to ground water through fate and transport modelling. This modelling may have to be supplemented with site-cpecific soil leaching studies (column studies) to determine realistic partition coefficients for contaminants in this type of aquifer. This supplemental field and modelling work could be completed as a focused RI under a separate operable unit (Operable unit II with Operable Unit I being ground-water remediation) or during the Remedial Design of a ROD with a generalized remedy. Specific Comments: Page 1-9, Paragraph 1: site inspections must occur at all facilities that may represent a potential source of contamination to soil and ground water at the Tutu site. It is not acceptable ;o not conduct site inspections on key properties because "difficulties were encountered obtaining access." Furthermore, it is unclear upon what basis Geraghty & Miller Inc. determined that enough information was already available for each property to preclude a site inspection. Sites that were not inspected <->O6 OS38 November 30, 1994 VIDPNR's Tutu Rl Comments still represent a potential source of contamination. page 2-11, Paragraph 3: Data collected during the supply well sampling should be presented in an appendix to the RI report. DPNR recommends that annual supply well sampling be continued through the Remedial Design phase of the project to monitor ground-water quality and plume dynamics. DPNR is not satisfied that the ground water in the supply wells has been adequately characterized. The recent discovery of contamination in the Delegarde well is an example of the lack of adequate plume characterization. Page 3-14, Paragraph 3: All materials to be disposed of at the Bovoni Landfill in St. Thomas must be sampled for all RCRA characteristics, TCL. organics, TAL metals and TPH and must be approved for disposal by DPNR prior to shipment off-site. Page 3-15, Paragraph 4: All the results from the various soil gas studies should be summarized and presented in the RI report. Page 3-20, Paragraph 2: EPA Standard Operating Procedures (SOPs) for well development require stability of pH, temperature, conductivity and turbidity within 10% following each of at least 3-5 purged well volumes prior to completion of development, it is unclear what is meant by "well development was continued until the discharged water was relatively sediment free." Apparently, wells may not have been appropriately developed prior to sampling. 2age 3-22, Paragraph 2: well sampling data from the first phase of the Rl should be summarized in this RI report and related to the Phase II sampling round. i?age 4-2, Paragraph 2: The RI report fails to present a complete inventory of potential contaminant sources in the Tutu valley. Fcr example, KFC {formerly Home Petroleum) is not adequately characterized or discussed. The report also fails to thoroughly identify previous land use and current industrial use in the valley. Page 4-13, Section 4.5.3 and Page 4-20, Section 4.6.2: The orientation of turpentine run basin and its relationship to the hydrodynamics of the site should be expanded and related to the fracture analysis. Additional discussion is needed to clarify the relationship between ground-water flow patterns and the orientation of the major fracture lineaments shown in Figure 4-8. DPNR considers the deep and shallow aquifer zones as part of an interconnected heterogeneous water table aquifer, in some areas TUT OOfo O539 November 30, 1994 viDPNR's Tutu Ri Comments this connection is not in a highly transmissive zone. The RI report is unclear regarding the relationship between the bedrock and alluvial water bearing zones. Page 4-22, Paragraph 2: Shallow ground-water flow appears to shift to the southwest in the southern portion of the site. However, a dearth of ground-water elevation data in the area does not allow for the accurate prediction of ground-water flow patterns. Page 4-23, paragraph 2; Water level data from existing wells is absent in the southern portion of the site area along and to the east and west of route 38 which is needed to describe the ground- water flow in this area. Page 4-35, Section 4.6.3.: This section is in need of substantial expansion to include more detail regarding the hydrodynamics of the site. Figures should created and referenced to illustrate the site hydrogeologic model. Specific information such as flow rates and flow direction should be discussed. This model should provide the framework for understanding contaminant distribution at the site and provide a basis for remedy analysis. Section 5.1.1 and Figure S-l (V.I. Housing Authority): The only soil samples shown in the RI are from drilling activities of MW- 13 and MW-13D. Soil sampling needs to be done in the vicinity of soil gas survey sites SGS-10 and SGS-13 to confirm/check elevated 3TEX readings reported during that survey. Additionally, chlorinated compounds were found in waste oil samples OIL 10, OIL 02 and OIL 891718. Soils on site should be analyzed for those chlorinated compounds (1,1,1 TCA, PCE, 1,1 DCE). Soil gas sampling was conducted in the area of the former UST on the property. Because the soil gas reveal no contamination, the area is considered not impacted by VOCs based on the criteria for soil impact. As stated earlier this criteria is grossly flawed and unacceptable for final determinations such as this. Soil gas must be followed up by extensive confirmatory sampling in the 5.rea of the former UST. Furthermore, based on Figure'5-1, only two soil gas samples were collected within forty feet of either drum storage area. This coverage is inadequate considering that chlorinated and nonchlorinated VOCs were detected in the drums stored at the northwest corner of the building. Section 5.12, Figure 2 (LAQA facility/DOE curriculum Center): Soil sampling needs to be done to confirm/check the soil gas survey that was used at the site for a screening tool. Existing soils data for the LAGA site do not show extremely elevated TUT 006 0540 H s^pli^ 10Cw for^tais or ^ were not *» detecte „ „ ^fjfi o. «• 5; - i ' . . . UT OO6 O541 ' W4 5 RI Comments * ssrto-^ay-ru. mat be dec ^s^«.^rL ^ana «• *« ssa j t^fi-S"a?J«SS4aj, 7116 statement «„ assunPtion is criteria d Pa9e 5 is S^-siSSrS i'S^X -sr? - - ndicated <in r*-«« TUT OO6 O54: etfort •MV a Boreenxna e11 as MW-1° O-2 . HW-lOD 12-14' MW-18 °'2 ,*.d tnrougftouc to* elated r-.^-es.S-S.'SSs."*"—" -8^-«/r cnr-^l———' . _ for Four *- • . 5.8 and l"y.^ ** a re»«*-- nder 0 i - ° . = store and*e ..wtj,. •£ SS"toil S--* °" ^ "on o* K. »» g& Sof»" rne tW- •!?!£*. P«« »*e the t°ut,, discussed- 10 TUT OO6 O543 November 30, 1994 VIDPNR's Tutu Rl Comments Section 5.1.9, Figure S-10 (ESSO): 21 COM samples are excluded from the discussion and the figure because locations were unknown to Geraghty & Miller. This is unacceptable an severely skews the interpretation of the soils from this property. The following detailed information from the CUM study have been omitted from this section: Measurements of 58,000 ppb of 1,1,1 TCA and 110,000 ppb of PCE in an oil/water separator in August 1987 Measurements of 100,000 ppb of 1,1,1 TCA and 65,000 ppb of PCE in an oil/water separator in September 1988 Measurements of 42,000 ppb of 1,1,1 TCA and 91,000 ppb of PCE in an oil/water separator in July 1989 Sampling at ss-8 location should continue below the 7.0 foot level to determine vertical extent of chlorinated contamination. No soil values were reported for CHT-7D and HW-9S. Soil data should be available for these wells during their installation. section 5.1.12.2.: Ramsay Motors, and Texaco should be included as properties that have been impacted by chlorinated VOCs. ?age 5-34, paragraph 4: Elevated BTEX in the vicinity of KFC indicates that the BTEX plume has not been defined in this area. Because of the lack of information regarding the ground-water flow in this area it is imperative that this portion of the plume be investigated further. page 5-35, Paragraph 2: MTBE has been found at substantial distances from the suspected source areas (e.g. the Delegarde well). No other source has been shown to exist in the south eastern portion of the site where detections have occurred. The nature and extent of this MTBE plume must be further investigated and discussed because it provides important information on ground-water flow and transport. MTBE should not be ignored just because there is no Federal drinking water MCL. Section 5.2.1.3.1.: The northern extent of the plume has not yet been adequately delineated. No wells north of VIHA i have been sampled and VIHA i contained chlorinated solvents. Sources areas in the northern part of the Tutu valley should be investigated. Section 5.2.2. : All sites should be evaluated for the presence of DNAPL, not just 0'Henry and the Curriculum Center. For example, the Esso Service Station has had detections of PCE 11 TUT OO6 O544 November .JO, 1994 viDPNR's Tutu Ri Comments in excess of 1,000 ppb in the soil and, as a result, should be discussed here. Section 5.2.3.: The report should indicate that DNAPLs may be present at. the site, but the probability can only ba assessed through a more detailed analysis that includes a study of soil characteristics (TOC, grain size, etc.), ground-water chemistry (pH, temperature, etc.) and fate and transport modeling. Such a study should occur during the Remedial Design phase of the project. No attempt has been made to establish the oxidation- reduction conditions in the aquifer. This information is necessary to determine the appropriate breakdown pathway of PCS for fate and transport modeling. Section: 5.2.6.: Page 5-47, Paragraph 2: Chlorinated VOC contamination was high in the soil gas survey conducted at the Curriculum Center. Chlorinated VOC contamination was not conclusive as a significant source from the soil sampling data, which predates the soil gas survey. Detailed soil sampling to confirm the soil gas survey results needs to be done at the site. The statement that "the ESSO Tutu Service Station does not appear to have contributed chlorinated VOCs to the groundwater" is not convincing since the highest values of chlorinated solvents was found in soils at the ESSO station - with sampling to at least 7 feet below surface. Section 5.2.6.1.2.: This section omits any responsibility of ESSO as a source for vocs. The contouring is too coarse to indicate some important trends. Overall, there is a marked decrease in chlorinated contamination values downgradient from the Curriculum center, in the vicinity of Tutu Texaco, then the chlorinated values increase in the Tillett area to as high as 585 ppb in the Tillett weal. The values decrease slightly downgradient from Tillet to the 130 to 200 ppb range in the vicinity of ESSO. These values remain relatively consistent for about 300 feet south of the station before falling off below the 100 ppb threshold. This constancy of values of chlorinated VOCs under and in the vicinity of ESSO does not preclude the soils under ESSO from being a source. The soils could well be a lower-level source to the ground water, keeping the chlorinated values constant, rather than allowing the concentrations to decrease steadily from the source. The statement made throughout the document that there are two distinct plumes of chlorinated VOCs in ground water is misleading. The RI has not adequately determined all the potential chlorinated VOC sources where distinct plumes may be 12 TUT OO6 O545 November 30, 1994 VIDPNR's Tutu RI Comments emanating from and has not adequately demonstrated a distinction between the areas that are mapped. Esso and Texaco are two potential source areas that have not been adequately addressed in this section. Page 5-46, Paragraphs i & 2: Although two "separate and distinct BTEX plumes" are indicated on the map, and the distribution of BTBX downgradient from ESSO seems limited, the MTBE values effectively smear the two plumes together. Further, and more important, MTBE values extend downgradient from the vicinity of the ESSO station all the way to the Delegarde well. MTBE values are present in MW-19, DW-2, Eglin-I, Eglin-lII, Steel, Laplace, Smith, Delegarde and KFC-1. Limiting discussion to the BTEX plumes only does not adequately describe the extent of gasoline contamination in this aquifer. Page 5-46, Paragraphs 3 & 4: BTEX contamination was reported during the 1990 excavation of USTs at the former Home Petroleum (RFC) site. This extent of this contamination needs to be evaluated further, but there is no clear evidence to show that the Home Petroleum contamination is the source for the MTBB in the vicinity of the Delegarde, Smith and LaPlace wells. Section £; A quantitative fate and transport mode11ing.study is necessary to determine risk to ground-water based remediation goals for soil. 13 TUT OO6 0546