Corrective_Action_Plan_Part_B,_AVGAS_Sludge_Burial_Site_A_(DP-05).pdf

CORREcnvE ACTION PLAN

PARTB

AVGAS Sludge Burial Site A (DP..06) Dobbins Air Reserve Base

841ndustrlal Drive Marietta, Cobb County, Georgia

EPD Facility ID No. 9-033018 ._.ease No.1

February 2001

Prepared for:

Georgia UST Management Program 4244 International Parkway

Suite 104 Atlanta, Georgia 30364

Prepared by:

Tetra Tech 101 Marietta Street, Suite 2400

Atlanta, Georgia 30303 Tetra Tech EMI Project No. FGWB 00-4302

Submitted for Tetra Tech By: Approved By:

CONTENT$

Section

2o

3o

INTRODUCTION AND BACKGROUNDooooooooooooooooooooooooooooooooooooooooooooooooooOoooooooooooooooooooooooooooooooooooooooo1

1o1 Site Location and Description 0000000 00000000000 ooooooooooooooooooo 0000000 oooooooooooooooooooooooooooooooooooooooooooooooooo1

1o2 Site History oOOOOOOoOOOOOOoOoooooooooooooooooooooooooooooooooooooooooooOoooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooo2

CORRECTIVE ACTION PLAN PART B CERTIFICATION oooooooooooooooooooooOoooooooooooooooooooooooooooooooooooooooo5

SITE INVESTIGATION REPORTooooooooooooooooooooooooo00000000000000oooooooooooooooo0000000ooooooooooooooooooooooooooooooooooooooo5

3o1 Horizontal and Vertical Extent of Petroleum Hydrocarbon Constituentsooooooooooooooooooooooooooooo5

3o1o1 Delineation of Impacted SollooooooooooooOOOOOoOOOOooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooo5

3o12 Delineation of Groundwater Plume

3o1o3 Delineation of Free Product Plume

3o1.4 Delineation of Surface Water lmpact

3o2 Local and Site Hydrogeology .. o

3o2o1 Documentation of Local Groundwater Conditions

3o2o1o1 Area Geology ............ o ........................ o ...... o

3o2o12 Local Groundwater Use .......................................................................... o10

3o2o1o3 Slug Tests

3o2o2 Stratigraphic Boring Logs

3o2o3 Stratigraphic Cross Sections

3o2o4 Referenced or Documented Calcuiations ............................................................ o12

3o2o5 Direction of Groundwater Flow

32o5o1 Monitoring Well Construction Information

32o5o2 Potentiometric Map

32o5o3 Flow Net

4o REMEDIAL ACTION PLAN ........................................................ , ............................ 0

4o1 Corrective Actions Completed or in Progress

4o1o1 Recovery and Removal of Free Product ............................................................ o13

4o1o2 Remediation and Treatment of Hydrocarbon-impacted Backfill and Native Solls ............ o .......................................................... o

42 Objectives of the Corrective Action

42o1 Remove Free Product that Exceeds One-Eighth inch

4o2o2 Provide Risk-based CorrectlveActlon ................ o ................................................ o14

4o2o2o1 Potential Receptor Survay ............................ o

4o2o2o2 Potential Exposure Pathways .. o .............................................................. o14

·Section

CONTENTS (Continued)

4.2.2.3 Fate and Transport Cslculations ........•...•.••......•..............•...••••..•.•....••....• 15

4.2.2.4 Conclusions and Recommendations ..•.....•••.....••.............•..•..•.....•....••

4.3 Design and Operation of Corrective Action System .............•.....••......•......••..........•.....•

4.4 Implementation .....................................................................•..................••............•

4.5 Public Notice ....•...................................................................••......•..............•..........•.....•....• 20

5. REIMBURSEMENT CLAIM ...............................................................•......•......•............•....•......•

REFERENCES ................................•.......................•.................•...................•......•.......•............•...•

TABLES

Table 1 Soli Analytical Results

Table2 Groundwater Analytical Results

Table3 Groundwater Level Data

Table4 Groundwater Monitoring Well Construction Details

Table5 Groundwater Geochemical Parameters/Biodegradation Indicators

TableS Free Product Recovery Summary

Table7 Surface Water Analytical Results

TableS Aquifer Characteristics

FIGURES

·Figure 1 Site Location Map

Figura 2 Site Layou1 Plan

Figura 3 Monitoring Well Location Map

Figura4 Groundwater Contour Map, October 2000

Figure 5 Soli Quality Map

Figure 6 Benzene Plume Map

Figure 7 Estimated Free Product Plume

FigureS Geologic Cross Section Map

Figura 9 Geologic Cross Section A-A'

Figura 10 - Geologic Cross Section B-B'

Figura 11 - Potentiometric Map, February 199S

Figura 12 - Groundwater Flow Net Map

FINAL

II

• CONTENTS (Continued)

FIGURES

Figure 13 - Proposed DVE Well System Layout

Figure 14 - Proposed ORC injection Point Locations

APPENDICES

A - CAP-B Certification

B - Groundwater Row Calculations

C - Identification of Potential Exposure Pathways

D • BIOSCREEN Calibration Model Run

E - BIOSCREEN Model Run - 2 Years into Future

F - BIOSCREEN Model Run- 30 Years into Future

G - Public Notification Letters

ATTACHMENTS

• A - EMP Report (MontgomeryWatson,1996b)

B - Analytical Date Sheets, Chain-of-Custody Forms, Boring Logs and Well Completion

Forms

C - Documentetlon of Potential Recaptors

D - Slug Test Date

E - Field PDotTest (Montgomery Watson, 1999)

F - Oxygen Releasing Compound Manufacturer information

• FINAL

ill

CORRECTIVE ACTION PLAN PART B

AVGAS SLUDGE BURIAL SITE A (DP-05)

DOBBINS AIR RESERVE BASE

841NDUSTRIAL DRIVE

MARIETTA, COBB COUNTY, GEORGIA

EPD FACILITY NO. 9-033018 RELEASE #1

TETRA TECH PROJECT NO. FGWB 00-4302

Tetra Tech, on behalf of the Alr Force Reserve Command (AFRC), has prepared this Corrective Action

Plan (CAP) Part B for the Aviation Gasoline (AVGAS) Sludge Burial Site A (Site DP-05), located at the

Dobbins Alr Reserve Base (ARB), 84 Industrial Drive, Marlette, Cobb County, Georgia. This report Is based prlmanly on a supplemental site Investigation (SSI) conducted by Montgomery Watson from

August 1997 to February 1998 (Montgomery Watson 1998). Additional information was also obtained from a comprehensive analytical report on the environmental monitoring program prepared by

Montgomery Watson in September 1996 (Montgomery Watson 1996b). Soli-vapor extraction (SVE) and dual-phase vacuum extraction (DVE) pOot tests ware performed between January 11 and 15, 1999 by

Montgomery Watson. The results of the test are summarized In the Final Pilot Test Flndlrltgs Letter Report from Montgomery Watson, dated May 18, 1999. Information from these documents has been compOed and formatted In accordance with the Georgia Underground storage Tank (GUST) Management regulations and the Georgia Environmental Protactlon DMslon's (EPD) CAP Part B guidance document entitled, Underground Storage Tank Release: Corrective Action Plan (CAP) Part B (EPD 1995~ The purpose of the SSI was to attempt to delineate the free product plume and soli and groundwater contamination resulting from the release of fuels from the former underground storage tank (usn system at the site. The Montgomery Watson documents referenced are Included as Attachments to this plan.

1. INTRODUCTION AND BACKGROUND

1.1 Site Location and Description

Dobbins ARB Is located at 84 Industrial Drive, Marlette, Cobb County, Georgia, and consists of about 1,666 acres of land under the command of AFRC. The base Is adjacent to the cities of

Smyrna and Marlette, about 5 mDes from the city limits of Atlanta. The AVGAS Sludge Burial Site

A (Site DP-05) Is a former storage and handling facility for aviation fuels, located northwest of the

AFRC petroleum, oil, and lubricant (POL) tank farm area. The coordinates of this site are latitude

33° 55' 20" north and longitude 84° 31' 39" wesl A site location map Is provided in Figure 1.

The adjoining 709 acres house Alr Force Plant 6 (AFP 6), operated by Lockheed-Martin

Corporation for Aeronautical Systems Center, which borders the site to the northwasl The

• 1.2

Georgia Army Reserve National Guard (AVGAS Burial Site B [Site DP-06D is located southeast of the POL fuel area, across Atlantic Avenue. A small, wooded area borders the POL area to the north and to the easl The ground slopes gently downhill from the POL fuel area to a base drainage-way, which borders the site to the north and east and flows from northwest to southeast

Water In this drainage-way Is fed from dewatering operations discharges from AFP 6. The drainage-way discharges on base Into a manmade retention area, referred to as Big Lake, located about 1,200 feet southeast of DP-05.

The POL area has been used since the 1940s for storage and handling of aviation fuels.

Disposed waste consisted of sludge residues collected from cleaning fuel tanks. Sludge was buried In shallow pits adjacent to the tanks through 1973, when this practice ceased. A site plan

Is provided In Figure 2. The site consists of two JP-8 aboveground storage tanks (AST), one

84,000iJallon and one 210,000-gallon, mounted on a monolithic, concrete slab. Immediately to the east of the JP-8 storage tanks are two 10,000ilallon and two 5,000iJallon diesel ASTs. A gravel surface surrounds the secondary containments for the JP-8 ASTs; however, the perimeter of the POL fuel area Is paved to provide Dobbins ARB fuel trucks access to the ASTs. Access to the POL fuel area is provided by gates to the southeast and southwest of the fenced-in POL fuel area .

Site History

The following data was compiled from Montgomery Watson and previous contractors on the

Dobbins ARB site. Site DP-05 has been used since the 1940s for storage and handling of fuels, including AVGAS, gasoline, JP-4, and JP-8. Wastes disposed of at the site Included sludge residues collected from cleaning of fuel tanks. Prior to 1973, some of the sludge (consisting of water, rust, and sediment) was removed from the fuel tanks and buried In shallow pits located northwest of the ASTs. Subsequent investigations have not confirmed the location of the burial pits (see Figure 2). In August 1991, six 25,000-gallon USTs were removed from the area just west of the ASTs and about 100 feet southwest of the presumed sludge burial pil Figure 2 presents the location of the former USTs at the site. The tanks had been anchored to a monolithic, concrete slab that Is still In place. The top of the slab Is about 14 feet below ground surface (bgs).

In April1982, DP-05 was identified as an Installation Restoration Program (IRP) site during the lnltlaiiRP records search conducted by CH2M Hill, Inc. (CH2M Hill) (CH2M Hlll1982). The purpose of the records search was to identify the potential for environmental contamination from past waste management practices and to assess the probability for contaminant migration. The records search Identified the potential for contaminants at the s.ite, resulting In the Inclusion of

DP-05in the IRP (Parsons Engineering Science (Parsons]1995a).

In 1986, Environmental Science and Engineering (ESE) performed a Phase II Confirmation and

Quantification study to Investigate hazardous material contamination and migration at the site

(ESE 1986). DP-05 was found to have significant potential for contaminant migration and was recommended for further study. Subsequently, ESE conducted a Site Characterization Study

(SCS) In 1991. The SCS Report recommended that a Remedial investigation (RI) be conducted

(Parsons 1995a). In August 1991, contaminated soli was Identified during removal of five of the siX 25,000-gallon USTs and associated piping at the site. Twenty-two soli samples were collected In accordance with Georgia's UST guidelines and analyzed for total petroleum hydrocarbons (TPH) and benzene, toluene, ethylbenzene, and xylenes (BTEX). Ten of the samples exceeded Georgia's TPH action level of 500 mDIIgrams par kDogram (mglkg), and siX of the samples exceeded the total BTEX action level of 1 00 mglkg (Parsons 1995a). No free petroleum product was observed beneath the tanks or the piping during the UST removal.

Samples of the stockpiled soil were tested using the Toxicity Characteristic Leaching Procedure and found to be nonhazardous under Resource Conservation and Reoovery Act definitions. The stockpiled soil was returned to the tank excavation and covered with clean backfill (Parsons

1995a). The clean backfill was mounded and compacted to promote surface water runoff and limit water Infiltration.

In 1992, Parsons conducted a Remedial investigation (RJ) at DP-05. The Rl Included a human health risk assessment (HHRA) of currant and hypothetical future exposure scenarios at the site.

The risk ~;~ssessment concluded that no current risk existed at the site, because complete exposure pathways were not present. However, hypothetical future risks, If excavation and construction activities were to occur, exceeded generally accepted U.S. Environmental Protection

Agency (EPA) risk levels for the typical worker and construction worker scenarios. Georgia

Department of Natural Resources (DNR) reviewed the Draft Rl Report prepared by Parsons for the site In 1995. The comments were Incorporated Into the Anal Rl Report (Parsons 1995a).

Dobbins ARB received correspondence In 1996 from Georgia DNR Indicating that the IRP activities at the site presented In the Final Rl Report had been accepted by Georgia DNR as an alternative to the preparation of a CAP Part A. The Georgia DNR Underground Storage Tank

Management Program (USTMP) typically requires a CAP Part A.

In October 1995, Parsons prepared a Feasibility Study (FS) for the site. Four groundwster remedial alternatives and siX soli remedial alternatives Oncludlng no action alternatives) were evaluated for the site (Parsons, 1995b).

In 1995 and 1996, Montgomery Watson conducted an Environmental Monitoring Program (EMP) at DP-05. The purpose of the EMP was to (1) obtain current analytical data, (2) revise the HHRA for the site, and (3) provide recommendations based on the analytical results (Montgomery

Watson 1996a and 1996b). The section of the EMP report describing the Investigations and results at DP-05 are presented In Attachment A.

Two rounds of soU, groundwatar, and surfaca water samples ware coOected and analyzed by

Montgomery Watson In 1995 and 1996. An additional monitoring well was Installed Immedlataly northeast and downgradlent of the former UST excavation. Subsequently, this new well (D5-

W47, sea Figure 3) was found to contain approximately 0.03 ft of frea product. As a result.

Montgomery Watson recommended additional activities for this site (Montgomery Watson 1996b).

SVE and DVE pilot tests were performed between January 11 and 15, 1999 by Montgomery

Watson. The results of the tast are summarized In the Final Pilot Test Findings Letter Report from

Montgomery Watson, dated May 18, 1999, a copy of which Is Included In Attachment E. The data generated from the pilot tests lndicatad that the DVE system provides a relatively large vacuum lnfluenca that is vertically uniform, but horizontally elongated because of the strike of the schist foliation in the saprolite layer abova the bedrock. The SVE tests were conducted over an 8-hour period. The effective radius of lnfluenca (ROI) of the vacuum (> 8-inches of watar) was about 30 to 40 feet. However, there were Indications of potential short-circuiting of air from the surfaca.

The DVE tests were conducted ovar a 64-hour period. The effective ROI for groundwater drawdown (> 8-inches of watar) was also about 30 feet. During the DVE test. the petroleum vapor recovary rata avaraged 24 pounds per day. A total of 3,314 gallons of groundwatar were exlr!\cted yielding an average production rate of 0.8 gallons per minute.

In October 2000, Tetra Tech verified groundwatar elevations at 14 monitoring walls. Tetra Tech also probed the wells for the potential presenca of free product. Groundwater monitoring wells

Investlgatad are presented In Figure 3. Water and free product levels were measured to the nearest O.o1-foot using an oll-watar Interfaca probe. Photolonizatlon detector (PID) readings ware also taken at the wellhead as an additional Indicator for potential groundwater contamination. The results of the field Investigation, conducted on October 12, 2000, are presentad In Table 3. Based on lnterfaca probe measurements, Monitoring Well D5-W9 had a measured, free product thickness of 0.02 foot, and wells D5-W26 and D5-W34 each had a measured, free product thickness of O.o1 foot. However, confirmation of the presenca of free product with balled samples was Inconclusive sinca no oil sheen was observad. Monitoring Wells

D5-W36 and D5-W51 had wellhead PID readings of 1,700 and 1,250 parts per million by volume

(ppmv), respectively; however, no product was detected with the Interface probe at these wells .

Monitoring Well D5-W47 had a PID reading of 1,030 ppmv, about 1 foot of measured free product, and no water. Well D5-W47 was purged with a bailer until It was dry. The well was resampled after about 1hour. The balled sample revealed the recovery of an additional small volume of free product and water Into the wall. All recovered free product and associated groundwater were placed In a drum on site and labeled as "petroleum contact water".

2. CORRECTIVE ACTION PLAN PART B CERTIFICATION

The CAP - Part B certification has been signed and Is Included In Appendix A of this report.

3. SITE INVESTIGATION REPORT

3.1 Horizontal and Vertical Extent of Petroleum Hydrocarbon Constituents

3.1.1 Delineation of Impacted Soli

Regulatory corrective action limits for site soils are specified In GUST Rule 391-3-15.09(3), Table B (Georgia DNR 1996). The site is located in a lower pollution susceptibility area, as defined by Hydrologic Atlas 20, with the contaminated soil in the former tank pit located over 500 feet from a surface water body (Big Lake). No private water supplies have been Identified within

0.25 mile, and no public water supplies are within 1 mile of the site. Comparing this Information to GUST Rule 391-3-15.09(3), Table B, the appropriate soU corrective action limits for the site are

11.30 mg/kg benzene, 500 mg/kg toluene, 140 mg/kg ethylbenzene, and 700 mg/kg xylenes. No corrective action limits have been established for polynuclear aromatic hydrocarbons et this site, because the health-based threshold levels for these compounds exceed the expected soli concentrations under free product conditions.

Based on previous reports describing removal of the USTs, contaminated soli was returned to the subsurface and capped with clean soli. Additional historic data indicates thet contaminated soli Is largely limited to the regions within or adjacent to the location of the former USTs. Historical soli data through 1996 Is presented In the EMP report In Attachment A (Montgomery Watson 1996b ).

None of the soli samples analyzed exceeded GUST soli threshold levels (STL) presented above.

Howaver, analytical results for three locations revealed concentrations thet exceeded the Georgia

Hazardous Site Response Act (HSRA) Type 3 Risk Reduction Standards (RRSs) (Georgia DNR, 1999) for either benzene or total xylenes. In 1990, benzene was detected at a concentration of

36 micrograms per kilogram (!'Q/kg) near well D5-W36 at a depth of 10 to 12 feet bgs (see Figure

3 for groundwater monitoring wall locations). Two sampling locations, one just south of the suspected sludge burial location taken In 1992 and one at Well D5-W47 taken In 1995, revealed total xylene concentrations of26 mglkg at 7 to 8 feat bgs and 41 mglkg at 9 to 11 feat bgs, respectively.

During Montgomery Watson's 1997 SSI, 19 soD samples from borings ware collected and analyzed to delineate the horizontal and vertical extent of soil contamination at the site. Borings were drilled using a direct-push unit, and 4-ft cores ware collected continuously. Samples for laboratory analysis of volatile organic compounds (VOC) were selected based on field screening, using visual inspection for staining and odor detection, and elevated PID readings. A total of seven samples were salected for laboratory analysis, five of which were collected from boring locations where no evidence of contamination was observed. Analytical results are summarized in Table 1. Laboratory data sheets and associated chain-of-custody (C-O-C) forms are Included In

Attachment B. Figure 5, taken from the SSI (Montgomery Watson, 1998, Figure 5-1), shows the locations of the samples sent for analysis and Indicates those samples with detectable contaminant concentrations above Georgia HSRA Type 3 RRSs. None of the soD samples analyzed revealed contaminants in excess of GUST STLs. Results Indicate that contamination from low concentrations of VOCs and semlvoiatlle organic compounds Is limited to the Immediate vicinity of the former USTs.

3.1.2 Delineation of Groundwater Plume

A total of 13 monitoring walls have bean Installed during various site Investigation activities In order to determine the horizontal extent of the groundwater plume. Additional wells used In evaluating recovery operations at AFP 6 are located Immediately adjacent to the western and northern boundaries of the site (wells labeled "ITS"). One AFP 6 recovery well, RW101, Is located near DP..Q5 (sea Figure 2). The 14 monitoring walls used In thesa Investigations are depleted In Figure 3. Historical groundwater data through 1996 Is presented In the EMP report In

Attachment A (Montgomery Watson, 1996b).

Water levels from all site wells were measured on October 12, 2000. These water levels and the historical water level data for the site wells are shown In Table 3. Well construction details ere summarized in Table 4.

Montgomery Watson sampled 10 groundwater monitoring points during November 1997, using a direct-push method to Install %-Inch temporary wells. During December of 1997, groundwater samples ware collected from eight monitoring walls on site (D5-W34, -W35, -W37, D5-W48 through -W51, and ITS-7). Savannah Laboratories and Environmental Services, Inc., analyzed the samples for petroleum hydrocarbon constituents. Samples were analyzed for VOCs, Including BTEX, using EPA Method 82609, for SVOCs (base neutrals) using EPA Method 8270C, and for diesel-range organics using modified EPA Method 8100. Additionally, samples from the monitoring wells were collected for analysis of the following geochemical and biodegradation parameters: dissolved oxygen, oxidation reduction potential, and conductivity by field analysis;

dissolved Iron (Method 6010): total organic carbon (Method 415.1): nitrate (Method 353.3): sulfate

(Method 375.4): and alkalinity (Method 310.1) by laboratory analysis. Analytical results for organics In groundwater samples from monitoring wells are summarized In Table 2.

Concentrations of benzene In groundwater at the time of the 1997 sampling events, Including direct-push points, are shown In Figure 6 (Montgomery Watson 1998, Figure 5-5). Analytical results of geochemical and biodegradation parameters are shown In Table 5. Laboratory data sheets and associated C-0-C forms are Included In Attachment B.

The EPD In-Straarn Water Quality Standards (ISWQS) applicable for the groundwater beneath this site are 71.28 f1911. benzene, 200,000 flg/L toluene, and 28,718 fl9ll. ethylbenzene.

Analytical results from the 1997 sampling events Indicate that the In-Stream Standard for benzene was not exceeded In any push-point samples. The ISWQS for benzene was exceeded

In monitoring wells D5-W34 (480 flg/L) and D5-W51 (3000 flg/L). Free product was discovered In

Well D5-W47; therefore, the well was not sampled. Groundwater samples collected from downgradlent wells and push-points located near the dralnageway to the north and east contained detectable BTEX constituents at concentrations well below ISWQS requirements.

Based on the analyticel results from this sampling event, the horizontal extent of the petroleum hydrocarbon plume In groundwater has bean delineated.

3.1.3 Delineation of Free Product Plume

Free product was originally detected at Well D5-W47 upon initial sampling, following Installation In

November 1995 (Montgomery Watson 1996a). The well Is located adjacent to and downgradlent from the UST removal area (see Figure 3). Based on field measurements with an Interface probe, the frea product layer had an apparent thickness of 0.05 foot. Field sampling logs Indicate that an approximate thickness of 0.03 ft was observed when the well was sampled with a bailer.

The recovery method and volume recovered were not given In the report or field sampling logs.

Montgomery Watson began a weekly removal of free product from site monitoring wells In August

1997 (Montgomery Watson, 1998). Activities Included gauging to the nearest 0.01 ft., all monitoring wells where free product was detected. Free product removal activities were conducted If free product was Identified In a monitoring well at a thickness exceeding O.o1 ft. A hydrophobic cloth (oil-absorbent pad) was usad to recover free product. Results from monitoring well gauging and recovery of free product are presented in Tables 3 and 6, respectively. In addition, Montgomery Watson conducted an Investigation In November of 1997 to horizontally delineate frea product at the site. The Investigation Included use of temporary direct-push monitoring points. Dedicated, disposable bailers were used to retrieve groundwater samples.

The presence or absence of free product was established by vtsual observation of a sheen, film, or petroleum product layer on groundwater. Upon completion of the free product delineation program, each monitoring point location was survayed by a State of Gaorgia-iicensed surveyor.

Figure 7, taken from the SSI (Montgomery Watson 1998, Figure 5-2), shows the free-product monitoring points and monitoring wells where free-product gauging and recovery wes conducted.

Table 3 provides results from each of the free product gauging events conducted by Montgomery

Watson and Tetra Tech. Free-product thickness for monitoring wells given In the figure ere the maxbnum value detected from gauging between August 1997 and January 1998. The figure also presents an Interpreted extent of the free product plume based on data taken during the SSI.

In general, free product at the site was a sheen of limited area In the vicinity of the former USTs.

The free product thickness ranged from 0.01 to 0.02 ft. Sheens were detected In many of the direct-push monitoring points at locations for which no hydrocarbon constituents had hlstoricelly been detected In groundwater. Thesa locations were primarily upgradlent, toward AFP 6. The only location at which free product was consistently delected was within the Immediate vicinity of the former USTs (Wells D5-W47 and -W51 ). Estimates of the product recovered from Well D5-

W47 are presented In Table 6. N. other locations, free product was detected Infrequently and rarely reappeared In wells following recovary. Thesa characteristics suggest that free product at the site primarily consisted of resldual-phasa hydrocarbons trapped In pore spaces above the water table. Volumes of free product collected were not provided In the SSI. However, volumes can be estimated from the product thickness and well diameter.

Montgomery Watson collected a free product sample from Monitoring Well D5-W471n 1997. The sample was "fingerprinted" by the laboratory to Identify the specific fuel(s) presant (modified

Method 8015). The sample was quantitated as gasoline, exhibiting large peaks for benzene and toluene. Additionally, the fractions Indicated the presance of commingled hydrocarbons such es

JP-4 or diesel .

3.1.4 Delineation of Surface Water Impact

Surface water drainage flows overland across the site to the north and east Into a base dralnageway system. The dralnageway is an unnamed tributary of Rottenwood Creek. The UST removal area at DP-05 Is located about 300 feet south and 600 feet west of the drainage-way.

Water In the drainage-way largely originates from outfalls for dewatering operations at AFP 6.

The drainage-way flows from northwest to southeast and discharges on base Into Big Lake, located about 1,200 feet east-southeast of the POL area. The lake is also part of the environmental monitoring program at Dobbins ARB and Is currently closed to base personnel for fishing, swimming, and wading. The outfall from the lake discharges Into a tributary of

Rottenwood Creek, a primary tributary of the Chattahoochee River. Limited data Is available to determine whether surface waters on base are gaining from, or losing to, the groundwater table.

Shallow and deep well clusters at AFP 6 Indicate a slight upward gradient from the bedrock toward the shallow aquifer {Parsons 1995a). However, current drought conditions In the area and start-up of the AFP 6 groundwater recovery system In 1999 are likely to have changed local groundwater gradients.

Surface water samples were collected and analyzed from two locations by Montgomery and

Watson In 1995 and 1996 {Montgomery Watson 1996b). Analytical results are summarized In

Table 7. Complete analytical results from the sampling are available In a copy of the DP-05 section from the EMP report {Montgomery Watson 1996b) provided In Attachment A. Sample

D5-SW03 wes collected upgrsdlent of the POL, northwest of Well D5-W34, just downstream of the AFP 6 outfall. Sample D5-SW05 was collected downgradlent from the POL, east-northeast of

Well D5-W37. Analysis of the samples revealed the presence of chlorinated VOCs (CVOC) In upstream and downstream locations. No source of CVOC exists at DP-05. A documented source for the chlorinated VOCs has been confirmed to ba the B-1 0 Aerallon Basin Site located at the adjacent AFP 6 facility northwest of DP-05 {Montgomery Watson 1998). Benzene was not detected In any upstream sample or In the 1995 downstream sample. However, benzene wes detacted at 13 J.lg/L In the downstream sample (D5-SW05) collected In 1998. The 1996 benzene concentration wes below EPD ISWQS of 71.28 J.lg/L.

3.2 Local and Site Hydrogeology

3.2.1 Documentation of Local Groundwater Conditions

3.2.1.1 Area Geology

FINAL

Dobbins ARB Is located on a gently rolling plateau In the Central Uplands district of the

Piedmont physiographic province. The district Is characterized by a series of low, linear ridges separated by broad, open valleys. Several small stream channels, Including

Rottenwood and Poorhouse Creaks, dissect the plateau. Elevations range from about

1,100 faet above mean sea level (ft msl) at the northwestern oomer of the base to approximately 950ft msl at the southwestern oomer. Figure 1 presents a USGS topographic map of Dobbins ARB with Site DP-05 delineated. Topographic lines at the site are also depleted In the site layout presented In Figure 2.

Bedrock strata In the area are termed the Sandy Springs Group, consisting of the Powers

Ferry Formation (Interbedded gneisses, schists, and amphibolltes ), the Chattahoochee

Palisades Quartzite (quartzite, locally feldspathlc or mlcecaous), and the Factory Shoals

Formation (Interbedded matagraywache and schists). Naturally occurring soils at

Dobbins ARB are saprolltlc, consisting primarily of mlcecaous, clayey slits and mlcecaous, sandy slits that retain the structural features of the parent bedrock. The weathered erosional surface of the rock Is Irregular, so depth to competent bedrock Is variable across the base.

3.2.1.2 Local Groundwater Usa

FINAL

As discussed In the synopsis provided by Montgomery Watson In the SSI (Montgomery

Watson 1998), a local water resources survey was conducted between November 26, and December 2, 1997. Survey activities Identified that no public and nonpubllc drinking water resources existed within 1 and 0.25 mile, respectively, from the site (sea Figure 1 }.

Montgomery Watson contacted the United States Geologic Survey (USGS) Georgia district office, which located three walls. These three wells were confirmed to be unused by USGS. A copy of the USGS water survey report Is presented in Attachment C.

Montgomery Watson reviewed the Cobb County Health Department database that held no records of "intent to drill a well" or of a request for a sample to be collected from any private well within 1 mile of the site. In addition, the Health Department had no record of a surface water intake within 1 mile of the site. In May 1995, Parsons conducted a drinking water supply survey and Identified no record of water supply within 3 miles of the

AFP 6. boundary.

Sixlean water supply wells are located on the Dobbins military reservation; however, the wells have bean inactive since the early 1950s. Fifteen of these previous well locations were filled in and covered during the early 1980s and are no longer accassible (Parsons

9958). The remaining Inactive well was converted to a groundwater monitoring watl for restoration activities at AFP 6 .

3.2.1.3 Slug Tests

On December 1997, slug tests were performed by Montgomery Watson to characterize the aquifer horizontal hydraulic conductlvlty under the site. Slug tests were performed on

Monitoring Wells D5-W50 and D5-W51, using a solid slug to displace the water column and a pressure transducer with an internal data logger to record the change In the water level. The recovery data wes then analyzed following the Bouwer and Rice (1976) method for unconfined aquifers, using the AQTESOLVlM computer program. The program was used to calculate the hydraulic conductlvlty (K) of the shallow aquifer. The computer printouts are Included In Attachment D.

Slug test results show K ranges between 8.99 x 1 0 ... centimeters per second (em/sec) at

D5-W50 and 1.27 x 1 o"" em/sec at D5-W51. These values forK should be considered to be approximations of the hydraulic conductivity, because the tests reflect localized conditions only. The geometric mean for Kat the site Is 1.07 x 10"" em/sec or 3.03 ft per day. SoU geotechnical testing and slug tests Indicated slmller values for hydraulic conductivities within the level of accuracy of these methods. Therefore, hydraulic conductivity of the water-bearing zone Is assumed to be within the range of 1 o"" to 1 0 ...

em/sec.

3.2.2 Stratigraphic Boring Logs

Stratigraphic boring logs and associated well completion forms for all site monitoring wells are provided In Attachment B.

3.2.3 Stratigraphic Cross Sectlo ns

Two stratigraphic cross sections of the site were prepared by Montgomery Watson using boring logs from existing and newly Installed groundwater wells and from soli borings. A map depleting the locations of the cross sections and the cross sections are provided In Figures 8 through 10

(Montgomery Watson 1998, Figures 4-3, 4-4, and 4-5). Cross Section A-A' extends from the southwest, at Soli Boring D5-SB06, to the northeast, through the sludge burial site up to Soil

Boring D5-SB11 and Is oriented roughly in the direction of the hydraulic gradient, as measured In

1998. Cross Section B-B' extends from northwest, at Soli Boring D5-SB1 0 bordering AFP 6, through the former UST location to Soil Boring D5-SB117,located In the southeastern comer of the POL area. Again, the site Is underlain by unconsolidated silty sands, with trace clay, and by weathered saprolite, which extends to the bedrock layer. Bedrock baneath the site consists of weathered gneiss and schist The unconsolidated deposits form a single, phreatic (water table), heterogeneous aquifer. The water table generally follows the surface elevation, sloping gently to the north and east toward the drainageway.

3.2.4 Referenced or Documented calculations

Based on water level data obtained in February 1998, the average hydraulic gradient is approximately 0.028 feet per feet, as reflected in the groundwater contour map (potentiometric map) from the SSi presented in Figure 11 (Montgomery Watson 1998, Figure 4-6). An updated groundwater contour map based on the recent water levels is presented in Figure 4. The two primary differences betwaen the potentiometric maps reflect a lowering in the water table caused by recent drought and an apparent Induced gradient in the western and northwestern region of

DP-05 from groundwater recovery operations at AFP 6. Because the former UST aree is located in the western portion of DP-05, groundwater flow in that area could be influenced by recovery operations at AFP 6. Calculations for determining the hydraulic gradient are shown in

Appendix B. Table 8 shows the pertinent aquifer-characteristic calculations for the shallow aquifer.

3.2.5 Direction of Groundwater Flow

3.2.5.1 Monitoring Well Construction Information

Groundwater monitoring wall construction details are provided in Table 4, and boring logs and wall construction diagrams for several of the walls installed during previous field investigations are included in Attachment B. Ali wells are completed with cement grout surface seals, stick-ups of from 2 to 2.5 ft above ground surface, and steel protective covers. The wells are surveyed relative to mean sea level. The elevations of the top of the well casings range from 1,043.83 to 1,074.93 feet, relative to the assigned benchmark. Montgomery Watson measured water levels in all 14 walls on February 9, 1998. The depths to groundwater ranged from 5.05 (D5-W49) to 28.57 (D5-W9) feet below the top of the wall casings. Tatra Tech measured water levels in all14 wells on

October 12, 2000. The depths to groundwater ranged from 6.8 (D5-W49) to 35.53 (D5-

W9) feet below the top of the well casings.

3.2.5.2 Potentiometric Map

FINAL

Figure 11 presents a potentiometric surface map for the site using groundwater elevations measured on February 9, 1998. Relative groundwater elevations ranged from

1,055.17 feet above meen sea level (msi) (D5-W48) to 1,038.78 ft msi (D5-W49). The direction of groundwater flow beneath the site in the area of Interest is primarily to the north and east. Figure 4 presents a potentiometric surface map for the site using the groundwater elevations measured on October 12, 2000. The relative groundwater elevations ranged from 1,047.86 feet msl (D5-W48) to 1,035.85 ft msl (D5-W37). The direction of groundwater flow beneath the site In the area of Interest Is still primarily to the north and east; however, an addltlonallnfluenca Is apparent from groundwater recovery wells at AFP 6 (sea Recovery Well RW1 01 In Figure 4 ), creating a small gradient toward the northwest at the western region of DP-05.

3.2.6.3 Row Net

A flow net superimposed on the site plan map Is provided In Figure 12. The map Includes selected flow lines to show the dlrectlon(s) of groundwater flow based on stetlc water level measurements made by Tetra Tech on Oclober 12, 2000.

4. REMEDIAL ACTION PLAN

4.1 Corrective Actions Completed or In Progress

4.1.1 Recovery and Removal of Fraa Product

Montgomery Watson Initially detected frea product at the site In 1995 In well D5-W47 at a thickness of 0.05 foot. Product recovery appears to have bean performed by hand balling from

August 1997 until February 1998, when free product was no longer detected In the weD. Tetra

Tech Inspected 14 of the site monitoring wells on Octot)er 12,2000. About 1 foot offrea product and no water was detected In Well D5-W47. The product was Immediately belled out until the well was dry. After 1 hour, the well was Inspected again, and a small amount of free product was removed. As a result of recant drought conditions In the eraa, It appears that the free product may be present because of the very low water table at the site, which Is about 5 to 7 feet lower than the elevations measured at the site In January 1998. The raappaaranca of free product Is not associated with a second release at the site, because all six of the USTs were removed from the site in August 1991.

4.1.2 Remediation and Treatment of Hydrocarbon-Impacted Backfill and Native Solis

No removal or treatment of contaminated soli has bean performed at the site .

• 4.2

Objectives of the Corrective Action

4.2.1 Remove Free Product that Exceeds One-Eighth Inch

About 1 foot of free product was encountered In Well D5-W471n October 2000. The normal response would be free product recovery at the site unW no measurable free product remains for about 12 consecutive months. However, as a result of telephone conversations between AFRC and Georgia USTMP personnel, a deviation from normal practices Is being proposed. The proposed corrective action requires that product removal be delayed until the planned remediation system Is In place.

4.2.2 Provlda Rlsk·based Corrective ActlonO

4.2.2.1 Potential Receptor Survey

Tetra Tach conducted a survey of potential receptors In the vicinity of the AVGAS Sludge

Burial Site on June 16 and October 31, 2000, to Identify potable water supplies, adjacent property owners, site workers, subsurface utilities, and any other pathway receptors that could be impacted by or could Influence constituent migration. All of the area

Immediately surrounding the site Is within the boundaries of Dobbins AFB. No private wells are In use within 0.25 mile of the site, and no public wells exist within 1 mile of the site (see Figure 1}. Municipal water Is supplied to the area. Subsurface water lines run

In a north-south direction, about 15 feat wast of the pad, for the 84,000-gallon AST.

Subsurface, Industrial wastewater lines are located In the area and used to collect storm water from secondary containments for ASTs and pads for fuel Issuing and recaMng areas. Both the domestic water lines and Industrial waste lines are located lass than 6 feet below land surface, and the shallowest depth to groundwater recorded at the site In the source area Is about 27 feat. No other burled utilities were Identified In the site area.

All of the supporting documentation for the potential recaptor survey Is provided In

Atlachment C.

4.2.2.2 Potential Exposure Pathways

FINAL

A Site Conceptual Exposure Model for all potential exposure pathways Is Illustrated In flow cherts In APPendix C. One complete exposure pathway has been identified at this site at the present time (see Aow Chart 1 In Appendix C). The site Is currently used only occasionally for vehicle fueling, so no site employees work at this location on a dally basis. Site buildings do not contain basements, and groundwater Is greeter than 27 feet bgs over most of the site area, so It Is unlikely that volatiles will migrate Into any enclosed spaces from soil or groundwater. None of the surrounding buildings have a basement or water well, and municipal water Is supplied to the base by Cobb County.

A construction worker was considered as a potential future receptor at the site (sea

Appendix C, Aow Chart 2). A hypothetical construction worker could be Involved In

Intrusive activities In soD around utilities or structures. Exposure pathways for this receptor are dermal contact with soil and groundwater, Inhalation of volatiles from soil and groundwater In a construction trench, and Incidental ingestion of soil. The depth to groundwater, over 27 feat, Indicates that it Is unlikely that construction activities would result In dermal contact with groundwater. Also, based on the depth to groundwater, It dces not appear that utilities In the area are Impacted.

No private drinking water wells are located within 0.25 mile, and no public water supply wells are located within 1.0 mile of the site. All of the propartles In this vicinity receive drinking water from the county. Domestic, or parsonal use wells, are typically drilled Into deeper strata. Therefore, potential impact to a drinking water well was considered to be unlikely and was not quantified. The nearest downgredlent surface water body Is located

640 feat east of the former tank pit source area. The potential Impact to the creek was evaluated as a worst-case recaptor.

4.2.2.3 Fate and Transport Calculations

FINAL

The nearest downgredlent surface water body was evaluated as a worst-case potential recaptor. The nearest surface water body Is the Intermittent stream, located 640 feat from the source area. No other potentially complete groundwater exposure pathways exist, so the pathway was evaluated as a worst-case exposure pathway.

The model selected to evaluate this pathway was BIOSCREEN, version 1.4.

BIOSCREEN Is a screening model that was wrltlen by C. J. Newell, R. K. Mcleod, and

J. R. Gonzales for the Air Force Center for Environmental Excellence. It Is based on the

Domenico analytical solute transport model (1987). The model simulates remediation through natural attenuation of dissolved petroleum hydrocarbon constituents at petroleum release sites. BIOSCREEN has the ability to simulate advactlon, dlsparslon, adsorption, and aerobic decay, which have been shown to be the dominant biodegradation processes at many petroleum hydrocarbon release sltas.

Benzene was chosen es the constituent to be modeled, because It Is the most mobile of the BTEX compounds and It was tha only constituent detected In groundwater at

• FINAL

concentrations greater than ISWQS requirements. The source concentration was input to be 28,000 fLg/L, which Is estimated as the maximum benzene concentration In groundwater at equilibrium with gasoline. The source duration was assumed to be infinite. A degradation half-life of 2 years was used, as required by GUST guidance. The one-tenth rule was used to estimate a longitudinal dlspersMty of 20.1 feat. The organic cerbon-water partitioning coefficient for benzene Is 81 milliliters par gram (mUg). Site specific date were unavailable for dry bulk density; therefore, the default value of 1.6 kg/L was used. The value for organic carbon fraction, f., was based on a default value of

0.0003, because no organic carbon was detecled In soli samples above the detection limit of 30 mg/kg. The simulation time was set at 6 years, assuming a release date of

1991 and comparison to benzene Ievals measured In 1997. The remaining Inputs

(hydraulic conductivity, hydraulic gradient, and effective porosity) were site-specific values which are presented In Table 8. The calibration model run and all Input parameters are shown In Appendix D. The model could not be calibrated with respecl to observed benzene levels In Monitoring Wells D5-W49 and direct-push sampling locations

D5-GW04 and D5-GW11, primarily because of the lack of adequate date on the exact configuration of the plume (for example, there was no current benzene date for all site walls, and no date for walls W36, W11, W9, andW26) and the true extentoffrae product

(this report relied solely on the previous extent of frae product, as determined by

Montgomery Watson In August 1997 and January 1998). The model Indicates an Impact to the downgradient creek at a concentration of 6,628 fLg/L. Surface water date collecled in 1996 Indicated that benzene was prasant at 13 fLg/L. No additional surface water samples have bean collecled In order to determine what the current benzene level is In the creek. The 2 and 30-year future model runs are presented In Appendices E and F, respactlvely.

The different model runs prasant concentrations with respect to time and the concentration profile with respecl to distance to the creak. The model predicts a maximum concentration of 6,800 fLg/L of benzene at the creek after an elapsed time of 36 years after the Initial release. This concentration Is considerably higher than the In

Straarn Standard ·for benzene of 71.28 fLg/L. This estimate Is extremely conservative, since the predicted values In the calibration run are greater than observed values, and actual future benzene concentrations are expecled to be much Iowar than predicted:

However, without additional benzene concentration data, the model cannot be raflned further at this time. Also, a proposed Alternate Concentration Limit (ACL) for benzene cannot be calculated at this time, because the available Information suggests that site benzene levels must be reduced to below 200 !191L before the stream will not be

Impacted at the site.

4.2.2.4 Conclusions and Recommendations

Current levels of hydrocarbons at this site pose no Immediate risk to human health, but the levels do pose a risk to the downgradlent stream receptor. Additionally, frae product has baen discovered at the site, and frae product recovery efforts are required. No sufficient analytical data exist at this time for comparison to determine If the groundwater plume has stabilized. A conservative fate and transport model indicates a current Impact to the cr~k at levels above the In-stream Standards, but no current surface water analytical data Is available to verify the Impact to the craek. As a result, Tetra Tech proposes to Install an aggressive DVE remediation system that Is complemented with an accelerated natural attenuation approach, as described In further detail In Sactlon 4.3.

Tetra Tech recommends that quarterly groundwater monitoring be performed at the site during operation of the remediation system in order to demonstrate the effectiveness of the system and to refine the fate and transport model. An additional monitoring well should be Installed near the stream receptor to sarve as a downgradlent compliance point.

At this time, Tetra Tech does not propose any additional corrective action for site soils. All

BTEX constituents were detected In soli at levels below appllceble sns and therefore do not require remediation.

4.3 Design and Operation of Corrective Action System

The recommended remediation design for corrective action at the DP-5 site will Include DVE, based on the results of the pilot test. The DVE design Is similar to the pilot test system. During the pilot test the DVE system consisted of a 20-horsepower, 230-volt, alternating current, oil cooled, liquid-ring vacuum pump, powered by a diesel generator and a 60-gallon air/water separator.

Liquid effluent from the air/water separator was discharged to a tanker truck for disposal, and the air effluent was discharged to the atmosphere. The system operated at a vacuum of 20-lnches of mercury ("Hg}, and the ROI for soli vapor was about 30 to 40 faet, and the ROI for groundwater was about 30 feet. In addition, It Is indicated from previous studies that the advancing edge of the dissolved petroleum plume Is threatening the water quality In the nearby downgradlent stream.

Consequently, the remediation design recommended by Tetra Tech Includes accelarated natural attenuation via Oxygen Releasing Compound"' (ORC) placed In the subsurface near the advancing edge of the dissolved plume.

Figure 13 Illustrates the conceptual placement of DVE wells that will result In…