USVI Public Records

Document text

SDMS Document w U I n Federal Programs Corjioracion 115565 A Subsldl/>iy Of CL.-.yp Dresser A McKee Inc. 125 Maiden LanG. 5lh Floor New York. New York 10038 Tel: 212 785-9123 Fax: 212 785-6114 October 8, 1996 Ms. Alison Devine U.S. Environmental Protection Agency 290 Broadway New York, New York 10007-1866 ^ ^ ^ PROJECT: DOCUMENT NO. SUBJECT: ARCS 11 Contract No. 68-W9-0024 Work Assignment 078-2PN7 7720-078-EP-CLWT Technical Review of Draft Supplemental Data Summary Report Virgin Island Chemical Site St. Croix, United States Virgin Islands Document No.: 7720-078-LR-CLWV Dear Ms. Devine: CDM FEDERAL PROGRAMS CORPORATION (CDM Federal) is pleased to submit this letter report entitled, "Technical Review ofthe Supplemental Data Summary Report for the Virgin Island Chemical Site" as partial fulfillment ofthe reporting requirements for this work assignment. If you have any questions regarding this submittal, please do not hesitate to call Pamela Philip at (212) 785-9123. Sincerely. CDM FEDERAL PROGRAMS CORPORATION Robert D. Goltz , P.E. ARCS II Program Manager cc: C. Kwan, EPA S. Henry, EPA J. Litwin, CDM Federal P. Phihp, CDM Federal B. Mende, CDM Federal Project File Document Control 303350 0T/S0-d 020T926609T6 01 d3/AN ayaa 0d3 sn UOdJ 80:60 966X-2e-lD0 9962 ^.£9 Z\c • ^ ^ D I l A Federal Programs Corporacioii A Subsidiary of Camp Dreaaer i McK«? Inc. m 125 Maiaen Lane. 5 n Floor New York. Nev/ York 10038 Tcl: 212 785-9123 Fax; 212 785-6114 Octobers, 1996 Ms. Caroline Kwan Remedial Project Manager U.S. Environmental Protection Agency 290 Broadway New York, NY 10007-1866 Ms. Sherrel Henry Remedial Project Manager U.S. Environmental Protection Agency 290 Broadway New York, NY 10007-1866 PROJECT: DOCUMENT NO. SUBJECT: ARCS II EPA Contract No.: 68 -W9-0024 Work Assignment No.: 078-2PN7 Virgin Island Chemical Site St. Croix, United States Virgin Islands 7720-078-LR-CLWV Technical Review of Draft Supplemental Data Summary Report Dear Ms. Kwan and Ms. Henry: CDM FEDERAL PROGRAMS CORPORATION (CDM Federal) has reviewed the Draft Supplemental Data Summary Report (DSDSR) for the Virgin Island Chemical site on St. Croix in the United States Virgin Islands. Tlie report was prepared by Harding Lawson Associates (HLA) on behalf of the potentially responsible party (PRP). The DSDSR is a supplement to the initial Data Summary Report (DSR) dated August, 1995. CDM Federal has the following general and specific comments regarding the DSDSR. General Comments 1. By design a data summary report is intended to summarize and evaluate all data collected during a site investigation. The analytical data should be combined in a comprehensive manner and include historical data and environmental and physical setting information so a conceptual site model can be formed. Once this model is developed, areas of potential concern can be 0T/£0"d 020T9£:6609I6 01 dO/AN aaa3 ud3 sn 996C i C 9 ^TP WOdd 303351 60:60 966T-2S-ijO CDM Fcdcr;il Programs Cnqxiraiioo Ms. Kwan Ms. Henry Octobers, 1996 identified and it can be detemiined, if necessary, what further actions or solutions are required. The DSDSR provides only the results from the Phase II sampling with little, if any, interpretation ofthe data. No recommendations for further investigation actions are offered. The DSDSR does not synthesize together or reference the results of previous site investigations and, therefore, presents an incomplete analysis ofthe identified concems. A final data summary report should be compiled that synthesizes and evaluates all site data as well as recommends additional actions to be conducted at the site. HLA states that the extent of soil contamination has been defined in all directions from the .AST Area except to the west. However, HLA must explain ftirther how the vertical and lateral extent of soil contamination to the south ofthe AST Area has been defined. Samples collected from the southernmost boring, SBB2, contained elevated levels of several volatile organic compounds. The extent of contamination beyond this point has not been defined. Further delineation ofthe lateral and vertical extent of soil contamination in the areas both south and west ofthe AST area is recommended. A minimum of four borings in each quadrant should be installed to delineate the extent of soil contamination. As stated above, the report does not compile or interpret past sampling results. Chloroform historically has been identified as a site contaminant of concem but its recent detection in the shallow groundwater in MW-2 (Process Pit) at a concentration above the Federal MCL, 470 ppb, is not subject to any interpretation by HLA. It should be noted that lower concentrations ofthe compound were previously detected in the two deep process wells P-1 (> 90 feet) and P-2 (> 90 feet). P-1 and P-2 are upgradient and laterally gradient of the former lab pit area, respectively, and upgradient (P-1) and downgradient (P-2) ofthe process pit area. Chloroform was detected in P-1 at a concentration of 31 ppb and in P-2 at a concentration of 130 ppb. The chloroform value detected in MW-2 is approximately 15 times higher than the P-1 concentration and approximately four times the P-2 concentration. The detection of chloroform in both the shallow and deep aquifer zone with a higher concentration present in the shallow groundwater zone strongly suggests the site is the source ofthe chloroform. It should also be noted that no potential upgradient sources have been identified in the area ofthe topographic high located to the south ofthe property. OJ o Ui tn to 0X/t70"d 020T9£:6609T6 01 d3/AN aaad dd3 sn 9962 A£9 Z\Z WOdd 60:60 9661-22-130 ^ # l J l w l Fedcnil Pro^rr;ii-ns Corporanoti Ms. Kwan Ms. Henry Octobers, 1996 The extent of chloroform in both the shallow and deep aquifer units needs to be further delineated. As per the EPA-approved HLA work plan, the production wells should be resampled to establish the present quality ofthe groundwater. The effects of pumping and sampling ofthe deeper production wells on adjacent shallow wells should be also monitored. It is also recommended that additional shallow and deep monitoring wells be installed and sampled to frilly delineate the extent ofthe VOC groundwater contamination. Based on a review of historical data, CDM Federal believes that the most likely origin of the chloroform was a direct injection ofthe compoimd rather then a surface spillage of contaminants that verticaUy migrated to the water table. Support for this contention is based on the physical characteristics of chloroform, regional climate conditions, and the regional geologic and hydrogeologic properties at the site. Chloroform is a very soluble and volatile compound (Vapor Pressure: 160 mm @20 C and 245 mm @ 30 C). Given the climate, a surface spill of chloroform in a either a dilute or concentrated fomi would tend to evaporate or vaporize in the shallow soil zone, minimizing the potential impact to the groundwater quality. Conversely, a subsurface discharge or injection ofthe compound would minimize vaporization and increase the potential impact to groundwater quality. During previous meetings concerning the site, several below grade structures (i.e., catch basins and recharge pits) that historically discharged untreated effluent directly into the soils were identified. These structures included tlie centrally located Process Pit and interconnected catch basins and the Laboratory Pit. Operations in the lab pit area resulted in a direct discharge of untreated liquid wastes to an unlined pit while process pit operations discharged contaminated effluent into a drain system whose design specifications are only partially known. Soil samples collected during HLA's onsite investigations did not indicate chloroform in the soils. However, a chloroform concentration of 200 parts per billion was detected in a composite soil sample previously collected in the laboratory pit area. The DSDSR neither examines these data together nor does it speculate or provide an analysis of this discrepancy. Ui o OJ OJ O l OJ 0T/S0'd 020192:660916 01 dD/AN aaa3 bdd sn WOdd 60:60 9661-22-130 QQi:;C > CQ PTT" W D I w l Fcdcml Programs Corporanon Ms. Kwan Ms. Henr>' Octobers, 1996 Due to the presence of chloroform in the groundwater, additional soil sampling in this area appears to be warranted. The absence of chloroform in HLA soil sample results could be the direct result ofthe sampling procedures utilized combined with the high volatility ofthe compound. Consideration should therefore be given to revise fiiture soil sampling protocols to include the methanol extraction procedure which has been shown to be effective in reducing volatilization during the sample collection process. 4. Volatile organic compounds were detected in groundwater samples collected from MW-1 and MW-6, wells located along the site's boundary', at concentrations above Federal MCLs. The groundwater flow direction presented in the DSDSR indicates that a strong potential for the offsite migration of contaminants exists. It is recommended that offsite wells be installed so the offsite migration of site contamination can be delineated. 5. The groimdwater elevation data presented suggest the monitoring well network may be insufficient to definitively depict groundwater flow directions. Mapping of groundwater elevations indicate inconsistencies in flow directions with the more recent contour maps showing a "groundwater divide". Insufficient data currently exist to support HLA's assertion that a groundwater divide is present onsite. Evaluations of previous groundwater measurements and local lithology are required to support this assertion. Ln addition, postulafion conceming the cistern does not appear to be supported by the data provided. If the cistern were the source of water recharging the shallow water table it would be anticipated that a mounding effect would be visible around the cistern rather than upgradient of it. 6. A comparison ofthe Tubers-generated groundwater elevation data and May 1995 groundwater sampling water level measurements indicates that the groundwater samples were collected following a major storm event. The large influx of fresh water from the storm event resulted in a significant rise of the local water table over levels measured the preceding day. Tliis influx of fresh water possibly, and most probably, diluted the contaminants present m the groundwater. Therefore, the May 1995 sampling results should not be considered representative ofthe actual groundwater quality'. Groundwater elevations measured at the time ofthe most recent groundwater OJ sampling event (June 18, 1996) show that the water table had equilibrated, g OJ Ul 0T/90'd 020T9e6609T6 01 dD/AN Qdda bdd SO WOdd 01:60 9661-22-130 i qq^r x'r'=i PT? W C i P f l Feder-al Programs Corpor-.^r ion Ms. Kwan Ms. Henry Octobers, 1996 and minimal, if any, effects from storm related water influx were noted. Therefore, the June 1996 analytical results appear to be representative of actual groundwater quality conditions. The DSDSR should be revised so all previous elevation data is presented and the effects of storm events on analytical data thoroughly evaluated. Since the validity ofthe May 1995 data is questionable, it is strongly recommended that a complete round of groundwater samples be collected at the time of equilibrium conditions. Data collected will be used to determine if the current shallow monitoring well network is adequate for groundwater characterization. Tliere appears to be some confusion regarding HLA's data validation procedures and data reporting. Methylene chloride values in several soil samples are flagged with a "B" to indicate blank contamination. If the values in the samples are less than 10 times the associated blank concentrations, then the detected values should be qualified as nondetected and flagged with an "U". However, there are several high concentrations of methylene chloride that are flagged with a "B" but do not appear to be attributable to blank contamination. According to the USEPA CLP National Functional Guidelines for Organic Data Review, EPA 540/R-93/012, February 1994; Section V, Blanks, E (page 21), a "B" qualifier remaining on validated data indicates that the analyte was detected in the associated method blank but can not be attributed to the method blank contamination. These values were most, likely not disqualified during validation because the sample results were greater than five or ten times the associated blank value. HLA should revise the data summary so the actual detections of methylene chloride are reported and evaluated. Due to the high volatilit}' ofthe groundwater contaminants that have been detected in the site monitoring wells, it is recommended that any future groundwater sampling be conducted in accordance with the EPA Region 2 Low Stress Purging and Sampling Guideline. This procedure is recommended since it will reduce sample aeration and will result in the collection of samples that are more representative of actual aquifer conditions. OJ o OJ OJ U l U l 0 T / i 0 ' d 0201926609T6 01 d3/AN Qdda bd3 SO WOdd 01:60 9661-22-130 qqi^c -ICQ p x ~ w D I V I Federal Programs Coq.Xirauon Ms. Kwan Ms. Henry Octobers, 1996 9. A proper evaluation ofthe hydrographs cannot be perfomied without area precipitation records. The records should be submitted for review and comparison. Specific Comments 1. Section 3.2.1: HLA states that organic vapors were not detected in any surface soil sample, including those samples in which volatile organic compounds were detected in soil samples collected at depth. This statement supports CDM Federal's previous contention that PID surface soil screening or surface soil collection and analysis is inadequate to delineate the extent of soil contamination due to the volatilization of compounds in shallow soils. 2. Section 3.3.3: The volatile organic groundwater sample results (volatile fraction) have high detection limits indicating that these samples were diluted. These high detection limits render the results unusable for comparison against regulatory standards. The accepted practice in EPA Region II is to report the undiluted sample results and replace any results exceeding the calibration range (E qualified data) with the diluted result for that particular analyte. In addition, the principal results ofthe validation of only the organic data are completely discussed in this section. The validation ofthe inorganic data should be more thoroughly discussed. 3. Section 3.2.2: Based on CDM Federal's review of the volatile organic compound (VOC) soil sampling results, several VOCs were reported by the laboratory, but not identified in the text. These compounds were: 1,1 Dichloroethene; Benzene; and Trichloroethene. The text needs to be expanded to incorporate these compounds. 4. Section 3.4,1: Fifth Paragraph: The term "relatively dr>'period" requires definition. Precipitation records should be provided. 5 Section 4.1; It should be noted in the conclusion that methylene chloride and acetone were also detected in soil samples at concentrations above the risk based criteria. In addition, the text should be edited to reflect that the lateral and vertical extent of soil conteimination has iiot been defmed to the area south ofthe AST area. See General Comment #6. ^ o OJ OJ I O l a\ 0T/80'd 020T926609T6 01 d3/AN addd bd3 SO WOdd 01:60 966T-22-130 996£ L£3 212 w W l V I Federal Progiams Qirp(,)ranon Ms. Kwan Ms. Henr\' Octobers, 1996 6- 7. Section 4.2: During the June 18, 1996, only the newly installed monitoring wells (MW-2 and MW-6) were sampled. Since the results of the May 1995 sampling results should be invalidated (See General Comment U 4), a complete round of groundwater sampling is required to evaluate the quality of the shallow and deep aquifer systems at the site. As per the EPA approved HLA work plan, both deep (P-1 and P-2) production wells should also be sampled. Sampling should be conducted only when monitoring ofthe groundwater elevations reveals equilibration ofthe water table- Figure 3-1 and Figure 3-2: The May 1995 groundwater sampling data can not be presented as being representative of groundwater quality due to the time of sampling. See General Comment # 4. 8. Figure 3-3: No explanation or interpretation ofthe trough observed over the period covering the end of April beginning of May 1996 is provided. The hydrograph for well MW-3 (and to a lesser extent MW-1) is provided. The hydrograph shows a significant, almost instantaneous, water loss followed two weeks later by a storm event. Immediately following the water loss, the slope or trend ofthe stabilization ofthe water levels remains constant. Measured precipitation levels should be incorporated (graphed) onto this figure and presented in tabular form. 9. Figure 3-4: This figure is confusing. It is presented in black and white with identical line weights for each ofthe wells graphed. The graphed lines overlap and crossover making it difficult to interpret the data. It is extremely difficult to discern the data provided for wells MW-1, MW-4 and MW-5 for the two weeks leading up to and including the storm event. It also appears that the designation for MW-3 is incorrect. 10. Figure 3-3 and Figure 3-4: The MW-1 Tubers data plot shows a regular pattern of peaks and valleys. This significant flow pattern is not seen for any ofthe other monitored wells. Over the 8-month monitoring period, the depth ofthe valleys and the rate of recovery are fairly consi.stent. These peaks and valleys as could be interpreted as Drayton and recovery from a pumping well located to the west ofthe site. As groundwater contamination, above the Federal MCLs, has been reported in the AST area, the pumping well should be identified and its impacts on site contamination should be evaluated during additional groundwater delineation studies. OJ o OJ OJ U l 0T/60"d 020T9e6609T6 01 d3/AN aad3 bd3 sn woad TT:60 966T-22-130 • I ' d "IblGl ^ * U i V I Ffcxleral Programs Corpor.-ition Ms. Kwan Ms. Henry Octobers, 1996 Pages 11. 12. Figure 3-7 and Figure 3-8; Further examination and evaluation ofthe "groundwater divide" is required. Table 3-8: Nickel, chromium and cobah result.s in samples MW-2 and MW- 2(Dup) are qualified with both "B" and "U". This is not possible. The result is either non-detect and is qualified "U" or the concentration is between the instrument and the contract detection limit and is qualified '"B", Aluminum in samples MW-2 and MW-2(Dup) are reported on the table with "UJ" qualifiers, however, as the report notes (section 3.3.2.4) these results are above CRDL. No "U" qualifiers should be used with these resuhs. Table 3-9: Samples MW-2 and MW-2(Dup) Aluminum, chromium, cobalt, iron, nickel and vanadium on MW-2 are qualified with both "B" and "U". As explained above, the result is either non-detect and is qualified "U" or between the instrument and the contract detection limit and is qualified "B". In addition, the report notes (section 3.3.2.4) that dissolved copper was not detected, however, the table reports copper at 4.2B ug/I. The text should be clarified. Please contact me at (212) 785-9123 if you have questions regarding this submittal. Sincerely, CDM FEDERAL PROGRAMS CORPORATION 13. <r-L; •"K.' Z'c. .- ZUi Z' Pamela J. Philip Work Assignment Manager CO o OJ 00 cn 00 0 T / 0 f d 020T926609T6 01 d3/AN QddS bd3 sn 9962 A£9 2T2 WOdd TT:60 966T-22-130