d1_2016_Watermark-Draft_Final_2016_AMR.pdf

Draft Final

2016 Annual Groundwater Monitoring Report

Naval Support Activity, Mechanicsburg Mechanicsburg, Pennsylvania

October 2016

Prepared for:

Department of the Navy Naval Facilities Engineering Command 9324 Virginia Avenue, Building Z-144

Norfolk, Virginia 23511-3095

Prepared by:

Watermark Environmental, Inc.

175 Cabot Street

Lowell, Massachusetts 01854

Watermark

10407-45 Draft Final 2016 Annual Groundwater Monitoring Report i October 2016 Naval Support Activity, Mechanicsburg, PA WLD2041

TABLE OF CONTENTS

SECTION PAGE

1.0 INTRODUCTION .................................................................................................................. 1-1

2.0 BACKGROUND .................................................................................................................... 2-1

3.0 Project Activities .................................................................................................................... 3-1

3.1 Field Work Performed ................................................................................................................... 3-1

3.2 Laboratory Analysis ...................................................................................................................... 3-1

3.3 Data Validation .............................................................................................................................. 3-1

4.0 GROUNDWATER FLOW EVALUATION........................................................................ 4-1

5.0 MAY 2016 ANALYTICAL RESULTS ................................................................................ 5-1

6.0 DISCUSSION OF ANALYTICAL RESULTS AT SITE 3 ................................................ 6-1

6.1 Extent of Contamination .......................................................................................................... 6-1

6.2 Contaminant Trend Evaluation ................................................................................................ 6-2

6.3 Statistical Evaluation of Contaminant Trend Data ................................................................... 6-5

7.0 CONCLUSIONS AND RECOMMENDATIONS ............................................................... 7-1

7.1 Conclusions .............................................................................................................................. 7-1

7.2 Recommendations .................................................................................................................... 7-1

8.0 REFERENCES ....................................................................................................................... 8-1

TABLES

Table 2-1 Contaminants of Concern Table 3-1 Well Construction and Water Level Data, Site 3, Site 8, and Site 9 – 2016 Table 3-2 Sampling Location Summary – May 2016 Table 4-1 Vertical Gradient Analysis – May 2016 Table 5-1 May 2016 Groundwater Sampling Event Frequency of Detection and Statistical Summary Table 5-2 Summary of Groundwater Sample Analytical Results – 2016 Table 5-3 Comparison of 2011, 2012, 2013, 2014, 2015, and 2016 COC Exceedances Table 5-4 Water Quality and Natural Attenuation Parameters – May 2016 Table 6-1 Total VOC Concentrations – Pre- and Post-Injection Table 6-2 Trichloroethene Concentrations – Pre- and Post-Injection Table 6-3 1,2-Dichloroethene (Total) Concentrations – Pre- and Post-Injection Table 6-4 Vinyl Chloride Concentrations – Pre- and Post-Injection Table 6-5 Chlorobenzene Concentrations – Pre- and Post-Injection Table 6-6 Arsenic Concentrations – Pre- and Post-Injection Table 6-7 Manganese Concentrations – Pre- and Post-Injection Table 6-8 Summary of Mann-Kendall Trend Analysis – Total VOCs Table 6-9 Summary of Mann-Kendall Trend Analysis – Trichloroethene Table 6-10 Summary of Mann-Kendall Trend Analysis – 1,2-Dichloroethene (cis- or Total) Table 6-11 Summary of Mann-Kendall Trend Analysis – Vinyl Chloride Table 6-12 Summary of Mann-Kendall Trend Analysis – Chlorobenzene Table 6-13 Summary of Mann-Kendall Tend Analysis – Arsenic Table 6-14 Summary of Mann-Kendall Trend Analysis – Manganese Table 7-1 2016 Groundwater Monitoring Program

10407-45 Draft Final 2016 Annual Groundwater Monitoring Report ii October 2016 Naval Support Activity, Mechanicsburg, PA WLD2041

GRAPHS

Graph 6-1 Contaminant Trends – Source Area Well S03M50 Graph 6-2 Contaminant Trends – Source Area Well S03M54 Graph 6-3 Contaminant Trends – Source Area Well S03M17 Graph 6-4 Contaminant Trends – Source Area Well S03M18 Graph 6-5 Contaminant Trends – Downgradient Well S03M44 Graph 6-6 Contaminant Trends – Downgradient Well S03M46 Graph 6-7 Contaminant Trends – Arsenic by Well Graph 6-8 Site 3 Groundwater Trends – Trichloroethene Graph 6-9 Site 3 Groundwater Trends – 1,2-Dichloroethene (cis- or Total) Graph 6-10 Site 3 Groundwater Trends – Vinyl Chloride Graph 6-11 Site 3 Groundwater Trends – Chlorobenzene Graph 6-12 Site 3 Groundwater Trends – Arsenic Graph 6-13 Site 3 Groundwater Trends – Manganese

FIGURES

Figure 2-1 Site Location Map and Well Location Plan Figure 4-1 Potentiometric Surface Map – Shallow Wells, May 9, 2016 Figure 4-2 Potentiometric Surface Map – Deep Wells, May 9, 2016 Figure 6-1a Trichloroethene Isoconcentrations – Shallow Wells, May 2016 Figure 6-1b Trichloroethene Isoconcentrations – Intermediate Wells, May 2016 Figure 6-1c Trichloroethene Isoconcentrations – Deep Wells, May 2016 Figure 6-2a cis-1,2-Dichloroethene Isoconcentrations – Shallow Wells, May 2016 Figure 6-2b cis-1,2-Dichloroethene Isoconcentrations – Intermediate Wells, May 2016 Figure 6-2c cis-1,2-Dichloroethene Isoconcentrations – Deep Wells, May 2016 Figure 6-3a Vinyl Chloride Isoconcentrations – Shallow Wells, May 2016 Figure 6-3b Vinyl Chloride Isoconcentrations - Intermediate Wells, May 2016 Figure 6-3c Vinyl Chloride Isoconcentrations – Deep Wells, May 2016 Figure 6-4 Chlorobenzene Isoconcentrations – May 2016 Figure 6-5 Manganese Isoconcentrations – May 2016 Figure 6-6 Arsenic Isoconcentrations – May 2016 Figure 6-7 Trichloroethene Isoconcentrations – May 2016 Figure 6-8 Trichloroethene Isoconcentrations Over Time (µg/L) Figure 6-9 cis-1,2-Dichloroethene Isoconcentrations Over Time (µg/L) Figure 6-10 Vinyl Chloride Isoconcentrations Over Time (µg/L) Figure 6-11 Chlorobenzene Isoconcentrations Over Time (µg/L) Figure 6-12 Manganese Isoconcentrations Over Time (µg/L) Figure 7-1 2016 Site 3 Groundwater Monitoring Program

APPENDICES

Appendix A – Field Forms

Appendix A-1 Water Level Sheets Appendix A-2 Field Log Sheets Appendix A-3 Chains of Custody

Appendix B – Historic Data

Appendix B-1 Well Construction Data and Historic Water Level Surveys Appendix B-2 Historic Groundwater Analytical Data

Appendix C – Data Validation Reports

10407-45 Draft Final 2016 Annual Groundwater Monitoring Report iii October 2016 Naval Support Activity, Mechanicsburg, PA WLD2041

LIST OF ABBREVIATIONS AND ACRONYMS

bgs below ground surface cis-1,2-DCE cis-1,2-dichloroethene COCs contaminants of concern EPA United States Environmental Protection Agency ft feet J estimated value MCLs Maximum Contaminant Levels NAPLs non-aqueous phase liquids Navy Department of the Navy NSA Naval Support Activity PCBs polychlorinated biphenyls PRGs Preliminary Remedial Goals ROD Record of Decision Site 3 Ball Road Landfill and Burn Pits TCE trichloroethene TOC top of casing trans-1,2-DCE trans-1,2-dichloroethene TtNUS Tetra Tech NUS, Inc.

μg/L micrograms per liter VC vinyl chloride VOC volatile organic compound Watermark Watermark Environmental, Inc.

10407-45 Draft Final 2016 Annual Groundwater Monitoring Report 1-1 October 2016 Naval Support Activity, Mechanicsburg, Pennsylvania WLD2041

1.0 INTRODUCTION

This report provides a summary of the results of the 2016 annual groundwater sampling at Naval Support Activity (NSA) Mechanicsburg, Pennsylvania. The 2016 sampling event was completed at Sites 3 and 9 due to site-specific groundwater contamination that requires monitoring and evaluation. The remaining environmental sites at NSA Mechanicsburg are not monitored for groundwater contamination. A brief overview of the site, a summary of the work performed, and a presentation and discussion of the water level data and analytical data for the sampling event are included in this report.

Groundwater flow patterns have been evaluated and current analytical results were compared to prior analytical data to identify changes in groundwater flow patterns and groundwater quality trends since the pilot-scale in-situ chemical oxidation work performed in 2004. Tabular summaries and graphical analytical trend analyses are provided for selected parameters and wells across the site. Finally, conclusions and recommendations are provided. Field forms are provided in Appendix A, historical data tables are provided in Appendix B, and data validation reports are provided in Appendix C.

The sampling for 2016 and this report were completed by Watermark Environmental, Inc. (Watermark) under contract to the Naval Facilities Engineering Command.

10407-45 Draft Final 2016 Annual Groundwater Monitoring Report 2-1 October 2016 Naval Support Activity, Mechanicsburg, Pennsylvania WLD2041

2.0 BACKGROUND

The 2016 sampling event was completed at Sites 3 and 9 as these sites have groundwater contamination that requires monitoring and evaluation. The remaining environmental sites at NSA Mechanicsburg are not monitored for groundwater contamination. In the near future (Summer 2016), an in-situ treatment remediation technique will be performed at Ball Road Landfill and Burn Pits (Site 3). In preparation for this event, the 2016 sampling analyte list included several chemical and geochemical parameters needed to evaluate baseline concentrations and aquifer geochemistry.

Site 3 covers approximately 7.5 acres across the southwestern portion of NSA Mechanicsburg as depicted on Figure 2-1. Originally, the site served as a quarry for borrow materials that were used in on-site construction.

Two large borrow pits were developed then later used as disposal areas (burn pits) for liquid wastes (solvents, lubricants, paints, varnishes, and gasoline) from the mid-1940s to 1977. A total of 47,000 tons of soil were removed from the two burn pits in 1999 as part of a soil removal action (EA Engineering, 2000). A post-removal action Record of Decision (ROD) for Site 3 soils was signed in 2000 [Department of the Navy (Navy), 2000]. The ROD specified the implementation of institutional controls (land use restrictions) to prohibit residential use of the property in the future. Currently, most of the site is covered by asphalt that was installed in 1994 as part of an interim remedial action for soils and re-installed in 1999 over the burn pit area after their excavation.

A series of investigations performed by the Navy at Site 3 from 1990 through 2003 indicated that groundwater was impacted by releases of contaminants associated with past activities at the burn pits, evidenced by a variety of detected organic and inorganic contaminants, most notably chlorinated volatiles and petroleum hydrocarbons. The contaminants have been found as both dissolved-phase constituents in groundwater and as components in non-aqueous phase liquids (NAPLs). The NAPLs (as floating product) have been sporadically found in fractures within the karst limestone at various depths. NAPLs can be a long-term residual source of groundwater contamination because of the slow dissolution rate of contaminants in this phase.

A 2004 ROD for Site 3 groundwater specified a remedy including in-situ chemical oxidation, land use controls, groundwater monitoring, and Comprehensive Environmental Response, Compensation, and Liability Act Five Year Reviews. The Navy implemented two phases of chemical oxidant (hydrogen peroxide/chelated iron catalyst) injection activities at Site 3 in four injection events. The two rounds of Phase I injections (March to May 2004) were performed at 25 injection points, while the two rounds of Phase II injections (August to October 2004) were performed at 17 new locations plus eight injection locations used during the Phase I in-situ chemical oxidation activities. Groundwater sampling from on-site wells, off-site wells, and downgradient wells was performed during each phase of injection. The results of the Phase I and Phase II sampling indicated that the chemical oxidation process had reduced volatile organic compound (VOC) concentrations, although cleanup levels for identified contaminants of concern (COCs) had not been achieved for many parameters. Table 2-1 presents the COCs and associated cleanup levels identified in the

ROD.

Three additional rounds of groundwater sampling and analysis were performed in 2005 and 2006 to further evaluate long-term changes in groundwater quality following the 2004 chemical injections. The results were compared to the ROD requirements [Tetra Tech NUS, Inc. (TtNUS), 2007]. From 2008 to 2016, annual groundwater sampling has been performed. This report documents the work performed and presents the sampling results for the May 2016 monitoring event. COCs from this round of groundwater sampling are compared to the previous rounds of monitoring in this report to identify concentration trends over time.

10407-45 Draft Final 2016 Annual Groundwater Monitoring Report 3-1 October 2016 Naval Support Activity, Mechanicsburg, Pennsylvania WLD2041

3.0 PROJECT ACTIVITIES

3.1 Field Work Performed

A comprehensive round of synoptic water level measurements was performed on May 9, 2016 prior to beginning well purging/sampling activities. There was no precipitation on May 9th, but there was some precipitation on the days preceding the measurement of water levels. There was a small amount of precipitation on the May 8th (0.01 inches) and May 7th (0.05 inches), and more rain on May 6th (0.57 inches) (Weather Underground, 2016). Monitoring well S03M51 was covered with a vehicle on May 9th; therefore, the water level for this well was measured on May 19th. Water levels were measured at 78 Site 3 monitoring wells and at two Site 9 monitoring locations to allow interpretation of groundwater flow patterns within and exterior to Sites 3 and 9. During the visit to each well, depth to water and total well depth below top of casing (TOC) were measured and recorded. In addition, an observation of the well condition was recorded in the field logbook. Table 3-1 is a listing of wells in which water level was measured with both water depth [feet (ft) below TOC] and water level elevation (feet above National Geodetic Vertical Datum).

From May 10 through May 19, 2016, 48 monitoring wells were sampled at Site 3 with two monitoring wells sampled at nearby Site 9 in accordance with the Sampling and Analysis Plan (Field Sampling Plan and Quality Assurance Project Plan) for Annual Groundwater Monitoring at Site 3-Ball Road Landfill and Burn Pits, Naval Support Activity Mechanicsburg, PA (TtNUS, 2008) and three corresponding amendment letters dated May 25, 2012 (Watermark, 2012), April 30, 2015 (Watermark, 2015), and March 28, 2016 (Watermark, 2016). Sampled wells are indicated on Figure 2-1 and listed in Table 3-2. Stainless steel submersible bladder pumps (with polyethylene bladders) utilizing nitrogen air-lift were used to collect groundwater samples. Dedicated polyethylene tubing was used for well purging and sampling. The pump intake depth for all monitoring wells was set at a depth consistent with previous sampling rounds to correlate with water producing zones as indicated on Table 3-1 (monitored interval).

As discussed in the 2015 Annual Report, monitoring well S03M69D2 is obstructed preventing placement of a bladder pump at the mid-screen depth of 220 ft. Monitoring well S03M69D2 was sampled using a bladder pump placed at approximately 170 ft below ground surface (bgs).

On May 18, 2016, an unsuccessful attempt was made to sample S03M64D1 using a QED® bladder pump.

The pump could not be placed at the mid-screen depth of 175 ft due to an obstruction at approximately 40 ft bgs. On May 19, 2016, S03M64D1 was sampled using a ¾ inch diameter MicroPurge® bladder pump, which was lowered past the obstruction and placed at the mid-screen depth of 175 ft. Groundwater quality parameters stabilized and were recorded along with the details of the issues encountered at S03M64D1 on the field log sheets include in Appendix A-2.

3.2 Laboratory Analysis

All groundwater samples were analyzed for VOCs, Aroclor-1260, manganese, and arsenic at an offsite analytical laboratory. To prepare for the planned Site 3 in-situ treatment remediation event, groundwater samples from nine monitoring wells (S03M15, -18, -41, -44, -48, -49, -50, -70, and -73D1) were also analyzed for alkalinity, ammonia, chloride, dissolved iron, total iron, nitrate, total organic carbon, sulfate, sulfide, dissolved gases (ethane, ethane, and methane), and dechlorinating bacteria. In addition, three additional wells were inadvertently sampled for total iron.

3.3 Data Validation

The laboratory analytical results were validated consistent with EPA Region III Data Validation Guidelines, the National Functional Guidelines and in accordance with the project SAP/QAPP (Watermark, 2016).

Analytical data was evaluated to determine if the data met data quality objectives (DQO); completeness and to document data quality and usability. Particular emphasis was placed on holding time compliance, equipment calibration, spike recoveries and blank results although all required elements of the validation process was considered. During the validation process, qualifiers were assigned to select results as needed and are reflected in the electronic data deliverables and analytical data tables. Data validation reports are included in Appendix C.

10407-45 Draft Final 2016 Annual Groundwater Monitoring Report 4-1 October 2016 Naval Support Activity, Mechanicsburg, Pennsylvania WLD2041

4.0 GROUNDWATER FLOW EVALUATION

Groundwater elevations were calculated from depth to water measurements (Table 3-1) and used to generate potentiometric surface maps centered on Sites 3 and 9. This section presents an interpretation of the contoured maps for the shallow and deep bedrock potentiometric surfaces with a discussion of vertical gradients. Appendix B presents historical groundwater elevation data and well construction information.

Figure 4-1 is a shallow bedrock potentiometric surface map indicating localized Site 3 potentiometric surface highs are present in the area of Burn Pit #1 (S03M47 through S03M51, S03M53, and S03M54) and near Burn Pit #2 (S03M58 and S03M57). Elevated groundwater levels are also observed in the wells near the storm water management pond and into the Defense Reutilization Marketing Offices area to the east (S03M66D1 and S03M46). Locally, groundwater flows radially from these potentiometric surface highs before flowing to the north, south, east, and west. Comparison of Figure 4-1 to groundwater flow maps in historic monitoring reports show that the localized mounding at Site 3 is consistently observed. The cause of the localized mounding is unknown, but may be related to past activities at the burn pits (i.e., injection of the chemical oxidant may have caused iron precipitation as a “halo” effect, decreasing the formation permeability and causing localized mounding).

Figure 4-2 depicts the deep bedrock potentiometric surface for Site 3. Deep bedrock groundwater at Site 3 is mounded at well S03M66D2 and groundwater flow to the northeast, northwest, and southwest. A steeper slope is present towards the west, in the direction of wells S03M73D1 and S03M73D2, than in other directions.

The measurement of all well water levels in a short period of time allows evaluation of vertical gradients.

This analysis was performed for 13 well clusters using the 2016 measurements. The following well clusters were evaluated:

• eastern perimeter of Site 8 (S03M01/S03M05);

• northern perimeter of Site 3 (S03M02/S03M06);

• northern and eastern perimeters of Site 3 (S03M03/S03M07);

• southern perimeter of Site 8 (S03M04/S03M08);

• eastern side of Site 3 (S03M61S/S03M61D and S03M62S/S03M62D);

• east of Site 3 (S03M66D1/S03M66D2 and S03M67D1/S03M67D2);

• west of the Site 3 (S03M69D1/S03M69D2);

• south of the Site 3 (S03M68D1/S03M68D2);

• between the two burn pits (S03M73D1/S03M73D2); and

• within the former burn pits (S03M63D1/S03M63D2/S03M63D3 and S03M64D1/S03M64D2/ S03M64D3).

Well locations are indicated on Figures 4-1 and 4-2 with the results of the vertical gradient evaluation presented in Table 4-1.

An upward gradient was observed at three well clusters while downward gradients were observed for eight well clusters (Table 4-1). The vertical gradients for well clusters S03M02/S03M06, S03M61S/S03M61D, S03M63D1/S03M63D2, S03M64D1/S03M64D2, and S03M73D1/S03M73D2 seem to be significantly downward, although magnitude gradient may be related to dynamic conditions due to variable or slow recharge and pressure build-up within the wells. Vertical gradients less than 0.1 feet were determined too small to accurately determine a gradient direction. Overall, the vertical gradient data suggests that large-scale groundwater movement is primarily downward, and that vertical movement is mainly attributed to well clusters S03M02/S03M06, S03M61S/S03M61D, S03M63D1/S03M63D2, S03M64D1/S03M64D2, and S03M73D1/S03M73D2.

Following previous recommendations, all of the Site 9 monitoring wells were abandoned with the exception of DD-1 and DD-7D. As a result, there is insufficient data to effectively determine the groundwater flow direction in this area.

10407-45 Draft Final 2016 Annual Groundwater Monitoring Report 5-1 October 2016 Naval Support Activity, Mechanicsburg, Pennsylvania WLD2041

5.0 MAY 2016 ANALYTICAL RESULTS

The 2016 sampling results indicate the site COCs (Table 2-1) were detected in multiple samples at concentrations above and below their respective Preliminary Remedial Goals (PRGs) or Maximum Contaminant Levels (MCLs) concentration. The site PRGs are defined in the ROD for Site 3 Groundwater (Navy, 2004). The detections include:

• All seven VOCs in one or more samples;

• Eight detections of Aroclor-1260; and

• Multiple detections of arsenic and manganese.

Table 5-1 provides a summary level analysis for the detected COCs, geochemical parameters, gases, and dechlorinating bacteria. The table includes frequency of detection, minimum and maximum concentrations (location of sample containing maximum are provided), averages of positive detections, and averages of all results. Table 5-2 provides detailed data results for all analytes and shading to indicate if a concentration for a COC exceeds the target concentration.

A comparison of results for the COCs is best evaluated per analyte group.

VOCs

• 1,4-Dichlorobenzene was detected in 27 percent (13 of 48) of the locations sampled with a range from 0.35 estimated value (J) micrograms per liter (µg/L) to 168 µg/L. Two locations had results that exceeded the MCL (75 µg/L).

• Benzene was detected in 40 percent (19 of 48) of the locations sampled with a range from 0.52 J µg/L to 515 µg/L. Nine locations had results that exceeded the MCL (5 µg/L).

• Carbon tetrachloride was detected in one sample at a concentration of 2.0 µg/L.

• Chlorobenzene was detected in 38 percent (18 of 48) of the locations sampled with a range from 0.35 J µg/L to 2,400 µg/L. Nine locations had results that exceeded the MCL (100 µg/L).

• Cis-1,2-dichloroethene (cis-1,2-DCE) was detected in 75 percent (36 of 48) of the locations sampled with a range from 0.43 J µg/L to 63,600 µg/L. Eight locations had results that exceeded the MCL (70 µg/L).

• Trichloroethene (TCE) was detected in 69 percent (33 of 48) of the locations sampled with a range from 0.63 J µg/L to 237 µg/L. Fifteen locations had results that exceeded the MCL (5 µg/L).

• Vinyl chloride (VC) was detected in 52 percent (25 of 48) of the locations sampled with a range from 0.44 J µg/L to 2,390 µg/L. Fourteen locations had results that exceeded the MCL (2 µg/L).

Polychlorinated Biphenyls (PCBs)

• Aroclor-1260 is the brand name for a Monsanto Chemical Company product that is classified as a PCB. This group of chemicals has a MCL of 0.5 µg/L [United States Environmental Protection Agency (EPA), 2009]. Aroclor-1260 was detected in 17 percent (8 of 48) of the locations sampled.

Three locations had results that exceeded the MCL (0.5 µg/L).

Metals

• Arsenic was detected in 65 percent (31 of 48) of the locations sampled with a range from 1.0 J µg/L to 58 µg/L. Sixteen locations had results that exceeded the MCL (10 µg/L).

• Manganese was detected in 92 percent (44 of 48) of the locations sampled. Three locations had results that exceeded the PRG (314 µg/L).

Table 5-3 presents a listing of the wells where PRGs/MCLs were exceeded in the May 2011, June 2012, May 2013, May 2014, May 2015, and May 2016 sampling events. The shaded cells in Table 5-3indicate a concentration increase between the previous and current year values.

10407-45 Draft Final 2016 Annual Groundwater Monitoring Report 5-2 October 2016 Naval Support Activity, Mechanicsburg, Pennsylvania WLD2041

The parameters that exceeded their respective PRGs/MCLs in May 2016 were relatively consistent with parameters that have exceeded their respective PRGs/MCLs in past sampling rounds. All analytical data were validated according to EPA Region III validation procedures.

Table 5-4 presents the results of groundwater quality parameter measurements taken during sampling (collected using the flow-through cell and turbidity meter). Appendix B-2 presents historic groundwater analytical results. The parameters that were detected in May 2016 were generally consistent with parameters that have been detected in past sampling rounds.

10407-45 Draft Final 2016 Annual Groundwater Monitoring Report 6-1 October 2016 Naval Support Activity, Mechanicsburg, Pennsylvania WLD2041

6.0 DISCUSSION OF ANALYTICAL RESULTS AT SITE 3

The presence of TCE and three reductive dechlorination by-products [cis-1,2-DCE, trans-1,2-dichloroethene (trans-1,2-DCE), and VC] indicates that degradation is on-going at the site. In some cases, concentrations of TCE were lower than that of VC and cis-1,2-DCE, most notably in samples from well S03M50. However, in the deep groundwater at well cluster S03M63, the TCE concentration was greater than VC. Chlorobenzene was predominantly detected around former Burn Pit #2, with only minimal detections in the center of former Burn Pit #1. Historically, chloroform has also been found in a number of wells at Site 3, although in 2016 results, only four Site 3 wells had low-level concentrations of chloroform. In general, exceedances of groundwater quality criteria for chlorinated VOCs at Site 3 extend approximately 310 ft east of Burn Pit #2.

The TCE, cis-1,2-DCE, and VC isoconcentration contours are shown in three separate hydro-stratigraphic maps to allow the depiction of VOC concentration changes with depth. The hydro-stratigraphic intervals shown are the shallow interval (represented by monitoring wells screened from 7 to 110 ft bgs), intermediate interval (represented by monitoring wells screened from 145 to 270 ft bgs), and deep interval (represented by monitoring wells screened from 300 to 358 ft bgs). The TCE isocontours are shown on Figures 6-1a through 6-1c, the cis-1,2-DCE isocontours are shown on Figure 6-2a through 6-2c, and the VC isocontours are shown on Figure 6-3a through 6-3c. Isoconcentration maps for three remaining COCs (chlorobenzene, manganese, and arsenic) are presented on Figures 6-4 through 6-6.

The two Site 9 wells, to the north of Site 3, have only low concentrations of VOCs. TCE in DD-1 and DD-7D was the only VOC with a concentration exceeding the MCL of 5 µg/L (Figure 6-7). The VOC concentrations detected in wells DD-1 and DD-7D are consistent with historical data (Appendix B-2).

6.1 Extent of Contamination

The isoconcentration contours depicted in Figure 6-1a through Figure 6-4 indicate the Former Burn Pits as the two primary areas of elevated VOC concentration. As with the mapping of any well concentrations and interpretation of plume extent, bias is introduced by the number and location of wells available for each interval. Observations related to TCE, cis-1,2-DCE, and VC distribution are provided in the following bullets.

• TCE distribution in the shallow interval wells of Former Burn Pit #1 (Figure 6-1a) are highest in well S03M49 (60.7 µg/L) with the concentrations in other wells less than the MCL. In Former Burn Pit #2, a plume of TCE is depicted from the highest concentration in well S03M41 (65.4 µg/L) to the east towards S03M44 and S03M46. It is difficult to constrain TCE concentrations to a specific interval within these wells as most have long open hole intervals ranging from approximately 15 to 100 ft bgs.

• Figure 6-1b depicts TCE concentrations as measured in the wells screened from 145 to 270 ft bgs (intermediate interval) with the maximum concentration shifted to the north in Former Burn Pit #1 to well S03M63D2. Of the three wells sampled in this area, well S03M63D2 has the deepest screen interval (250 to 270 ft bgs) and highest TCE concentration indicating TCE concentrations increasing with depth. In Former Burn Pit #2, the plume retains the same general shape of the shallow interval with an orientation to the east based on the location of wells S03M66D1 and D2.

• TCE concentrations for samples collected from the deepest screen interval wells are depicted in Figure 6-1c and indicate plume limits essentially centered over each of the Former Burn Pits. Five wells sampled in or near the footprint of Former Burn Pit #1 indicate the highest concentrations directly beneath the pit footprint (S03M63D3) with a low concentration extension to the north to well S03M71. Plume extent in the deep interval cannot be interpreted with only a single well sampled in the Former Burn Pit #2 area however it should be noted the TCE concentration in well S03M64D3 (160 µg/L with screen interval of 310 to 330 ft bgs) is similar to S03M63D3.

• Figures 6-2a to c and 6-3a to c depict the plume configurations of cis-1,2-DCE and VC, respectively.

As degradation byproducts, plume shape of these two compounds is typically correlative to the plume of their parent compound (i.e., higher concentrations of cis-1,2-DCE with higher TCE

10407-45 Draft Final 2016 Annual Groundwater Monitoring Report 6-2 October 2016 Naval Support Activity, Mechanicsburg, Pennsylvania WLD2041 concentration). An exception to this observation occurs in the depiction of cis-1,2-DCE and VC in well S03M15 (Figures 6-2b and 6-3b) as this well was non-detect for TCE yet cis-1,2-DCE and VC are elevated. As shown in Table 5-2, DHC, BVC, and VCR counts are highest in this well indicating local conditions are favorable for the sustainment of VOC-degrading bacteria. High concentration of chloride and methane in this well are also indicative of the ongoing degradation.

In general, the VOC contaminant plumes shown in Figure 6-1a through Figure 6-4 are similar in size and shape as those observed in previous sampling events performed since 2011, that is centered on the former Burn Pits with an eastward component at former Burn Pit #2.

A comparison of total VOC and individual VOC groundwater concentrations prior to and following the 2004 chemical treatment are provided in Tables 6-1 to 6-5 with similar tables for arsenic and manganese in Tables 6-6 and 6-7. An average value for the pre-injection time period (May 2000 to 2004) and post injection period (October 2004 to May 2016) is provided for each Burn Pit area well set and non-source area wells.

The average value is not provided if the well has only a single value, such as S03M50 for the pre-injection period. In addition, an overall average concentration was calculated for each time period using all sample values for each set of wells (i.e., Former Burn Pit Area #1, Former Burn Pit Area #2, and Other Wells).

Using Table 6-2 for discussion, the following observations related to TCE concentrations can be made:

Former Burn Pit Area #1

• Only three average pre-injection values can be calculated for wells S03M22, S03M54, and S03M55 due to limited samples. Relative to post-injection averages, minimal change in TCE concentration is indicated for these three wells.

• Well S03M50 has a single pre-injection concentration of 46,000 µg/L with a post-injection average concentration of 415 µg/L indicating a significant and sustained reduction of TCE in the capture area of this well.

• The overall TCE concentration for Former Burn Pit Area #1 in the pre-injection period is 7,688 µg/L with a post-injection average concentration of 82.6 µg/L. If the S03M50 value of 46,000 µg/L is removed from the calculation, the Burn Pit #1 overall pre-injection period average reduces to 25.9 µg/L, indicating no significant change in overall TCE concentration except at well S03M50.

Former Burn Pit Area #2

• Four of the ten wells in this area show a significant to slight decline in average TCE concentration from pre-injection to post-injection (S03M03, S03M17, S03M18, and S03M36).

• The overall average TCE concentration for Former Burn Pit Area #2 in the pre-injection period is 1,064 µg/L with a post-injection average concentration of 223 µg/L, indicating an approximate 89 percent concentration reduction.

Other Wells

• A significant to slight decline in average TCE concentration (pre-injection to post-injection) occurred in wells S03M21, S03M45, and S03M46.

• The overall TCE concentration for all wells showed a reduction from an average pre-injection concentration of 72.9 µg/L to a post-injection average concentration of 6.66 µg/L.

6.2 Contaminant Trend Evaluation

An evaluation of chemical trends over time was performed to assess changes in COC concentrations and evaluate plume stability. Concentration versus time plots (Graphs 6-1 through 6-6) were prepared for the six prevalent contaminants [TCE, cis-1,2-DCE (or total 1,2-DCE if cis-1,2-DCE is not provided in historical results), VC, chlorobenzene, manganese, and arsenic].

The wells were selected to represent source and downgradient locations and include the following monitoring wells/well clusters:

10407-45 Draft Final 2016 Annual Groundwater Monitoring Report 6-3 October 2016 Naval Support Activity, Mechanicsburg, Pennsylvania WLD2041

• S03M50 and S03M54 (former Burn Pit #1 source area);

• S03M17 and S03M18 (former Burn Pit #2 source area);

• S03M44 (lesser area of elevated VOCs downgradient of former Burn Pit #2); and

• S03M46 (southeastern edge of the Site 3 plume).

Note that the concentration scales for VOCs are on the left side of the graphs and on the right side for manganese and arsenic. Arsenic concentrations versus time were also plotted for the wells on one graph (Graph 6-7) because trends were difficult to discern in the concentration versus time plots for individual wells.

The following is a summary of the trend analyses from the concentration versus time graphs.

• S03M50 (Graph 6-1): Following the Phase II chemical oxidant injection in October 2004, VOC concentrations rapidly decreased as indicated by the results from the August 2005 sampling event.

TCE concentrations continued to steadily decrease since 2005. Cis-1,2-DCE, and to a lesser degree VC, have increased from 2009 to 2014, and have been fairly stable since from 2014 to 2016. The increase in cis-1,2-DCE and VC indicate that reductive dechlorination processes are active.

Manganese concentrations spiked following the Phase II injection activities and slowly decreased through 2006 before rebounding in 2008; however, from 2010 through 2016, manganese concentrations have decreased. Chlorobenzene, cis-1,2-DCE, and VC concentrations were all above their respective MCLs.

• S03M54 (Graph 6-2): Overall VOC concentrations increased after the chemical oxidation injection activities were performed, then decreased since 2008. Manganese concentrations spiked following the Phase II injection activities but have slowly decreased. Manganese and VC concentrations exceeded their respective PRG and MCL values in 2016. The presence of TCE breakdown products cis- and trans-1,2-DCE and VC, along with the decreasing concentration of TCE, indicates that reductive dechlorination is occurring.

• S03M17 (Graph 6-3): VOC concentrations in this well fluctuated between 2001 and 2006 before stabilizing near their current levels. Chlorobenzene is the primary contaminant found in this well, with lesser concentrations of other VOCs. The chlorobenzene concentration in 2016 is relatively consistent with the 2015 concentration. The TCE, cis-1,2-DCE, VC, manganese, and arsenic concentrations are relatively consistent with previous recent results. Concentrations of benzene, chlorobenzene, cis-1,2-DCE, VC, and Aroclor-1260 were greater than their respective MCLs in 2016. Overall VOC concentrations in this source area well are either stable or decreasing slightly.

• S03M18 (Graph 6-4): VOC concentrations decreased significantly during the Phase I and II chemical oxidant injections then rebounded somewhat before leveling off by 2009. Chlorobenzene, the primary contaminant found in this well, presents an overall decreasing but unstable trend.

Manganese concentrations spiked following the Phase I and II injection activities and have decreased since. Concentrations of TCE, VC, and arsenic have remained relatively constant from 2009 to 2016. In the 2016 sample results, concentrations of benzene, chlorobenzene, TCE, VC, and arsenic were all greater than their respective MCLs; however, the concentration ratio of TCE to cis-1,2-DCE and VC has decreased, indicating an increase in anaerobic biodegradation activity since the completion of chemical oxidant injection activities.

• S03M44 (Graph 6-5): VOC concentrations increased after the chemical oxidation injection activities were performed, but have decreased since 2005. Manganese concentrations spiked nearly one year following the Phase I and II injection activities, decreased through 2011, rebounded in 2012, and continued to increase through 2015 but have decreased in the past year. The delay in seeing the initial manganese concentration spike is likely due to the well’s location more than 200 ft east of the injection area. Cis-1,2-DCE concentrations increased much more sharply than TCE concentrations following the injection. This change, along with an initial slight increase in VC concentrations, suggests an increase in anaerobic biodegradation activity following completion of chemical oxidant

10407-45 Draft Final 2016 Annual Groundwater Monitoring Report 6-4 October 2016 Naval Support Activity, Mechanicsburg, Pennsylvania WLD2041 injection activities. Overall VOC concentrations in this downgradient well have been variable with a decrease in 2016. TCE and VC were both greater than the respective MCLs in 2016.

• S03M46 (Graph 6-6): VOC concentrations in this downgradient well dropped significantly after the second round of chemical oxidant injection and have been generally stable since 2006 and decreasing since 2012. Chlorobenzene and VC, if present, have concentrations below the method detection limit (i.e., non-detect) since at least 2009. In 2016 there are no parameters present at a concentration exceeding its MCL.

• Arsenic data (Graph 6-7): Arsenic levels in wells S03M17, S03M18, S03M44, S03M46, S03M50, and S03M54 remained low through 2016, with little change from pre- to post-injection sampling.

The data suggest minimal impacts on arsenic levels by the chemical oxidant injection activities.

Spikes in arsenic concentrations in monitoring wells S03M17 and S03M18 occurred following the Phase I injections. Arsenic concentrations in wells S03M17 and S03M18 have decreased and/or stabilized since 2006.

Graphs 6-8 through 6-13 present the same six contaminants that were presented in the line graphs, grouped by analyte. The concentration data are presented on a logarithmic scale to facilitate the presentation of a wide concentration range. Zero (non-detect) values are not plotted. Presentation of data in this format allows for visual assessment of long-term trends (or lack thereof) on an analyte-specific basis. Analysis of arsenic and manganese was not performed in 2013. The 2016 observations are summarized in the following bullets.

• TCE (Graph 6-8): TCE concentrations have generally decreased or remained the same among the wells graphed. Concentrations have decreased significantly overall in wells S03M17, S03M18, S03M41, S03M50, and S03M54.

• cis-1,2-DCE (Graph 6-9): Overall, cis-1,2-DCE concentrations remained steady between 2001 and 2016, possibly due to reductive dechlorination of TCE and subsequent generation of cis-1,2-DCE.

Concentrations in six of the wells decreased slightly in 2016. A fairly significant increase continues to be observed in well S03M50. Although the concentration is greater than the MCL (70 µg/L), this increase, in conjunction with the decrease of TCE concentration, could be an indication of the progression of reductive dechlorination.

• VC (Graph 6-10): Similar to cis-1,2-DCE, VC concentrations are overall stable with slight decreases in seven of the wells in 2016. VC was not detected in the 2016 sample from wells S03M46 and S03M49.

• Chlorobenzene (Graph 6-11): No overall trends in chlorobenzene concentrations are evident from the graph. The concentration in three wells increased slightly in 2016 while concentrations in two wells decreased. Chlorobenzene was not detected in wells S03M44, S03M46, and S03M49 in 2016.

• Arsenic (Graph 6-12): Arsenic concentrations have remained stable or decreased slightly over time in the wells that historically had higher levels (S03M17 and S03M18, both within former Burn Pit #2). Arsenic concentrations increased in two wells and decreased in two wells in 2016 compared to 2015. Arsenic was not detected in samples from wells S03M44, S03M46, and S03M49.

• Manganese (Graph 6-13): Manganese concentrations increased during the injection activities and generally decreased or remained stable in most wells through 2016.

Figures 6-8 through 6-12 present TCE, cis-1,2-DCE, VC, chlorobenzene, and manganese isoconcentrations, respectively, in Site 3 groundwater during the period from 2006 to 2016. Presenting the isoconcentration contours of six different sampling events on the same figures allows for a visual assessment of the change in shape and size of the contaminant plume. In general, the shapes of the plumes appear consistent and the sizes and concentrations have remained stable or decreased. Only the wells sampled during each event are shown in the appropriate frame of the figure.

The TCE plume proximal to former Burn Pit #1 and #2 appears to have remained the same size over the 2006-2012 timeframe, although the concentrations of TCE in the center of the plumes have decreased by an

10407-45 Draft Final 2016 Annual Groundwater Monitoring Report 6-5 October 2016 Naval Support Activity, Mechanicsburg, Pennsylvania WLD2041 order of magnitude since 2006 (Figure 6-8). As indicated in the panels for 2008 to 2016, additional wells to the east of Burn Pit #2 were sampled, enabling the eastern-most edge of the TCE plume to be delineated.

There has been a slight increase in the northern extent of the TCE plume at former Burn Pit #1 first shown on the 2014 and 2016 maps.

The cis-1,2-DCE (Figure 6-9) plumes in former Burn Pit #1 and #2 have generally remained the same size from 2006-2012. There has been an increase in the northern extent of the plume at former Burn Pit #1 shown on the 2014 and 2016 maps. There also appears to be an area northeast of former Burn Pit #2 at S03M15 with increased concentrations shown on the 2014 and 2016 maps.

The lateral extent and concentrations within the VC plumes (Figure 6-10) associated with former Burn Pits #1 and #2 have increased since 2008. There has been an increase in extent and magnitude northeast of former Burn Pit #2 and in an area north of former Burn Pit #1. Assuming VC generation is a byproduct of biodegradation, it is typical to measure higher VC concentrations proximal to TCE hot spots, such as the Pit areas.

The chlorobenzene plume (Figure 6-11) associated with former Burn Pit #2 also appears stable with minimal variance during the period from 2006 to 2016. Chlorobenzene concentrations at S03M50 in former Burn Pit #1 have increased in 2015 and 2016 to above 100 µg/L.

Manganese concentration (Figure 6-12) in Burn Pits #1 and #2 appear to have remained stable from 2006 to 2016, except for a temporary increase in extent shown in 2014 northwest of former Burn Pit #2 caused by a temporary increase in concentration at S03M65D in 2014 and 2015.

6.3 Statistical Evaluation of Contaminant Trend Data

A quantitative evaluation of concentration trends over time was performed using Mann-Kendall statistic at the 80, 90, and 95 percent confidence levels for all six primary COCs (TCE, cis-1,2-DCE, chlorobenzene, VC, arsenic, and manganese) and total VOCs on the same sampling round data as presented in Tables 6-1 through 6-7. The Mann-Kendall analysis was constrained to data from August 2004 (during the Phase II injections) through May 2016. The results of the statistical trend analyses are presented in Tables 6-8 through 6-14.

A number of monitoring wells were removed from the sampling program between the 2012 and 2016 sampling events; however, 27 wells have sufficient data for statistical trend analysis. The Mann-Kendall trend analysis compares all of the pairs of concentration values, counts the number of pairs where values are increasing, and subtracts the number of pairs that are decreasing. If there are more pairs that are increasing in concentration, then an upward trend is predicted. If there are more pairs that are decreasing in concentration, then a downward trend is predicted. A trend was considered statistically significant if it passed the 80 percent confidence level test and not significant if a trend was suggested by the data but did not pass the 80 percent confidence level test.

For Tables 6-8 through 6-14, the results are separated into the Pit #1 and #2 source areas and their plumes, the down gradient plume fringe, and Site 9 with trend predictions for the 80, 90, and 95 percent confidence levels (Alpha = 0.20, 0.10, and 0.05, respectively). The results for the 80 percent confidence level are discussed below for total VOCs and the COCs.

Table 6-8 (Total VOCs)

- The analysis for the Burn Pit #1 source area indicates two wells with significant downward trends and one well with an increasing trend (S03M50). In the Burn Pit #2 wells, two wells have no discernible trends with three wells showing a decrease in total VOC concentration over time. The results for the Burn Pit #1 wells plume are inconclusive; however, an overall significant downward trend is evident for the Burn Pit #2 plume wells. The majority of the plume fringe wells also indicate a decreasing total VOC concentration trend at the 80, 90, and 95 confidence levels. Of the two Site 9 wells, only DD-7D indicates a discernable trend of decreasing concentrations. The indication of overall decline in total VOC concentration may be a result of the past in-situ treatments but is likely a result of the progression of on-going biodegradation.

10407-45 Draft Final 2016 Annual Groundwater Monitoring Report 6-6 October 2016 Naval Support Activity, Mechanicsburg, Pennsylvania WLD2041

Table 6-9 (TCE)

- The Mann-Kendall trend analysis revealed significant downward trends in the two Pit source areas, plumes, and plume fringe wells. Of the 23 wells in this data set, three well locations had a significant upward trend (S03M07, S03M15, and S03M52), and four well locations had neither statistically significant upward or downward trends. Of the two Site 9 wells, only DD-7D indicates a discernable trend of decreasing concentrations. Overall, the average total post-injection TCE concentration (105 µg/L) is well below the average concentration from the pre-injection/injection time period (2,332 µg/L), as shown on Table 6-2.

Table 6-10 [1,2-DCE (cis- or total)]

- The Mann-Kendall trend analysis revealed significant downward trends at 13 well locations, significant upward trends at two well locations (S03M21 and S03M50), and 10 well locations had neither statistically significant upward or downward trends. Overall, the average total post-injection 1,2-DCE concentration (1,258 µg/L) is slightly below the average concentration from the pre-injection/injection time period (1,448 µg/L), as shown on Table 6-3.

Table 6-11 (VC)

- For this compound, the trend appears to be well-specific and is likely a function of the proximity of the well to an active area of TCE biodegradation. VC concentrations in the Burn Pit #1 source area well are increasing (S03M50) yet both upward and downward in the two source area wells of Burn Pit #2. Similarly, the Pit plume wells show upward and downward trends. For the plume fringe wells, both downward and no trends are indicated. Overall, the average total post-injection VC concentration (120 µg/L) is slightly greater than the average concentration from the pre-injection/injection time period (86 µg/L), as shown on Table 6-4.

Table 6-12 (Chlorobenzene)

- The Mann-Kendall trend analysis revealed significant downward trends at three well locations, significant upward trends at four well locations (S03M51, S03M50, S03M44, and DD-1), and 10 well locations had neither statistically significant upward or downward trends. Overall, the average post-injection chlorobenzene concentration (425 µg/L) is relatively consistent with the average concentration from the pre-injection/injection time period (439 µg/L), as shown on Table 6-5.

Table 6-13 (Arsenic)

- The Mann-Kendall trend analysis revealed significant downward trends at three well locations, significant upward trends at five well locations (S03M54, S03MW52, S03M50, S03M48, and S03M20), and 10 well locations had neither statistically significant upward or downward trends.

Overall, the average post-injection arsenic concentration (12.9 µg/L) is consistent…