Sovereign_2015_LTMMP_Update_-_Final_-_Complete.pdf

Prepared for:

US Army Corp of Engineers

New England District Concord, Massachusetts

Prepared by:

Sovereign Consulting Inc.

Contract No.: W912WJ-10-D-0003 Delivery Order: 0013

LONG-TERM MONITORING AND

MAINTENANCE PLAN UPDATE

SHEPLEY’S HILL LANDFILL

FORMER FORT DEVENS ARMY INSTALLATION, DEVENS, MA

FINAL

REVISED

SEPTEMBER 2015

SHL Long-Term Monitoring and Maintenance Plan Update Sovereign Consulting Inc.

Final Version

NOTICE

The United States Department of Defense, Department of Army, funded wholly or in part the preparation of this document and work described herein under Contract No. W912WJ-10-D- 0003 and Delivery Order 0013. Mention of trade names or commercial products does not constitute endorsement or recommendation for use.

Long-Term Monitoring and Maintenance Plan Update

FINAL

Devens, Massachusetts

Revised

September 2015

CERTIFICATION:

I hereby certify that the enclosed Report, shown and marked in this submittal, is that proposed to be incorporated with Contract Number W912WJ-10-D-0003 DO#0013. This Document has been prepared in accordance with USACE Scope of Work and is hereby submitted for Government Approval.

Reviewed By:

11/2/15

Sovereign Project Manager Date

11/2/15

Sovereign Quality Control Manager Date

Received By:

USACE Project Manager Date

REDDY.PENELOPE.W.15021050

Digitally signed by REDDY.PENELOPE.W.1502105066 DN: c=US, o=U.S. Government, ou=DoD, ou=PKI, ou=USA, cn=REDDY.PENELOPE.W.1502105066 Date: 2015.11.06 11:50:02 -05'00' i

TABLE OF CONTENTS

1.0 INTRODUCTION

1.1 OBJECTIVES AND REPORT ORGANIZATION

1.2 BACKGROUND AND SITE DESCRIPTION

1.3 REMEDIAL ACTION OBJECTIVES

1.4 SUMMARY OF REMEDY COMPONENTS TO ADDRESS RAOS

1.5 BACKGROUND OF EXISTING LTMMP

2.0 EXISTING LTMMP PROGRAM AND CONCEPTUAL SITE MODEL

2.1 SUMMARY OF THE CURRENT LTMMP

2.2 CONCEPTUAL SITE MODEL

2.2.1 Background / Summary

2.2.2 Contaminant Fate and Transport

2.2.2.1 North Impact Area

2.2.2.2 Red Cove

2.3 GROUNDWATER MODEL UPDATE

2.4 EVALUATION OF NEW DATA TO EXISTING LTMMP

3.0 UPDATED LTMMP PROGRAM

3.1 DATA QUALITY OBJECTIVES FOR THE UPDATED LTMMP PROGRAM

3.1.1 DQOs for the Landfill Cap/Containment Remedy Monitoring, Maintenance and Performance Evaluation

3.1.2 DQOs for the Groundwater Remedy

3.1.3 DQOs for Barrier Wall Monitoring

3.2 SAMPLING DESIGN

3.2.1 Limits of Decision Errors

3.2.2 Data Acquisition

3.3 LANDFILL MONITORING AND MAINTENANCE

3.3.1 Annual Inspections

3.3.1.1 Landfill Inspection Checklist

3.3.1.2 Corrective Action

3.3.2 Vegetative Maintenance

3.3.3 Settlement Monitoring

3.3.4 Landfill Gas Monitoring

3.3.4.1 Frequency and Parameters

ii

3.3.4.2 Monitoring Equipment and Sample Analysis

3.4 ATP OPERATION AND MONITORING

3.4.1 System Description, Operations, and Maintenance

3.4.2 Influent/Effluent Monitoring

3.5 BARRIER WALL MONITORING

3.5.1 Hydraulic Head Monitoring

3.5.2 Arsenic Flux to Red Cove

4.0 SAMPLING PROCEDURES

4.1 ENVIRONMENTAL MEDIA MONITORING

4.2 PRE-SAMPLING ACTIVITIES

4.2.1 Equipment and Supplies

4.2.2 Equipment Calibration

4.2.3 Site Location, Security and Access

4.2.4 Initial Well Opening and Inspection

4.2.5 Water Level Measurements

4.3 SAMPLING ACTIVITIES

4.3.1 Well Purging

4.3.2 Sample Containers and Preservatives

4.4 SAMPLE COLLECTION

4.4.1 Sample Identification

4.4.2 Quality Assurance/Quality Control Samples

4.4.2.1 Field Duplicate Sample

4.4.2.2 Rinsate Blank

4.4.2.3 Matrix Spike/Matrix Spike Duplicates

4.5 POST-SAMPLING ACTIVITIES

4.5.1 Chain-of-custody

4.5.2 Sample Delivery/Shipment to Laboratory

4.5.3 Equipment Decontamination

4.5.4 Investigation-Derived Waste

4.5.5 Data Validation

4.6 FIELD DOCUMENTATION

4.6.1 Field Log Books

4.6.2 Field Sample Collection Sheets

iii

4.6.3 Photographic Documentation

4.6.4 Project File

5.0 INSTITUTIONAL CONTROL MONITORING PLAN

5.1 LAND USE CONTROL OBJECTIVES

5.2 INSTITUTIONAL CONTROLS

5.3 LAND USE CONTROL MAINTENANCE AND INSPECTION

6.0 REPORTING REQUIREMENTS

7.0 REFERENCES

List of Figures

Figure 1 Site Location Map Figure 2 Shepley’s Hill Landfill Overall Site Plan Figure 3 Downgradient Arsenic Limit of Impact Figure 4 Area of Land Use Controls Figure 5 Long-Term Monitoring Well Network

List of Tables

Table 1 LTMMP Sampling and Hydraulic Monitoring Program Table 2 LTMMP Hydraulic Monitoring Only Table 3 Groundwater Sampling Analytical Program Table 4 Well & Piezometer Abandonment List Table 5 Historic Barrier Wall Hydraulic Monitoring Data iv

Appendices

Appendix A NIA Supplemental Investigation Data Appendix B Baseline Data for Existing LTMMP Monitoring Wells Appendix C Site Specific Quality Assurance Project Plan (CD) Appendix D Landfill Inspection Checklist Appendix E Landfill Gas Monitoring Form Appendix F Mass Development Landfill Discharge Permit No. 020 Appendix G EPA Region I Low-Stress (Low Flow) SOP (EQASOP-GW 001) (CD) Appendix H Monitoring Well Sampling Log Form Appendix I Response to Regulator Comments v

ABBREVIATIONS, ACRONYMS, AND SYMBOLS

ADR Automated Data Review AOC Area of Contamination As Arsenic ATP Arsenic Treatment Plant BOH Board of Health BCT Base Closure Team bgs Below Ground Surface BRAC Base Realignment and Closure CERCLA Comprehensive Environmental Response, Compensation, and Liability Act CH4 Methane CIP Clean in Place ClO2 Chlorine Dioxide CO Carbon Monoxide CO2 Carbon Dioxide COC Contaminant of Concern CQC Contractor Quality Control CSM Conceptual Site Model cy Cubic Yard Devens Former Fort Devens DOD Department of Defense DPT Direct Push Technology DO Dissolved Oxygen DQO(s) Data Quality Objectives DTW Depth to Water ELAP Environmental Laboratory Accreditation Program EW Extraction Well FBRO Filter Bed Roll-off FM Flux Maintenance ESD Explanation of Significant Differences FS Feasibility Study gpm Gallon per Minute H2S Hydrogen Sulfide IDW Investigation-Derived Wastes LEL Lower Explosive Limit LTM Long Term Monitoring LTMMP Long Term Monitoring and Maintenance Plan LUC Land Use Control LUCIP Land Use Control Implementation Plan MA Massachusetts MassDEP Massachusetts Department of Environmental Protection MCL Maximum Contaminant Level MCP Massachusetts Contingency Plan MF Microfiltration MS/MSD Matrix Spike/Matrix Spike Duplicate NAE New England District vi

NAVD North American Vertical Datum NELAC National Environmental Lab Accreditation Conference NIA North Impact Area NTCRA Non-Time Critical Removal Action O2 Oxygen ORP Oxidation-Reduction Potential PID Photoionization Detector POTW Publically Owned Treatment Works PVC Polyvinyl Chloride QA/QC Quality Assurance/Quality Control QAPP Quality Assurance Project Plan QSM Quality Systems Manual RAO Removal Action Objective RI Remedial Investigation ROD Record of Decision RQS Rock Quality Designation SAR Supplemental Assessment Report SDG Sample Delivery Group SHL Shepley’s Hill Landfill SOP Standard Operating Procedure Sovereign Sovereign Consulting Inc.

PPE Personal Protective Equipment USACE United States Army Corp of Engineers USEPA United States Environmental Protection Agency VOC Volatile Organic Compound

1.0 INTRODUCTION

Pursuant to the Contract Modification for #W912WJ-10-D-0003 Task Order 0013, Sovereign Consulting Inc. (Sovereign), on behalf of the US Army Corps of Engineers New England District (USACE-NAE) and the Army Base Realignment and Closure (BRAC) Environmental Office at Devens, Massachusetts has updated the 2007 Shepley’s Hill Landfill (SHL) Revised Long Term Monitoring and Maintenance Plan (LTMMP) (CH2M Hill, 2007) and the 2009 SHL Revised LTMMP Addendum (ECC, 2009). The updated LTMMP includes revisions, as deemed appropriate, to the groundwater monitoring program, treatment plant monitoring, landfill gas monitoring, and landfill cap inspection/maintenance. Further, this update documents long-term monitoring associated with the installation of a hydraulic barrier wall on the eastern side of SHL which is designed to restrict arsenic flux from SHL towards Plow Shop Pond and the implementation of land use controls (LUCs) in the north impact area (NIA) north of SHL.

Lastly, the updated LTMMP provides information such that the long-term effectiveness of the cap and Contingency Remedy may be evaluated per the remedial action objectives of the 1995 Record of Decision (ROD), the 2005 and 2013 Explanations of Significant Differences (ESDs).

1.1 Objectives and Report Organization

The objectives of this updated LTMMP are as follows:

Summarize the site description and historical background;

Summarize the current Removal Action Objectives (RAOs) and the remedy components applied to address these RAOs;

Summarize the Conceptual Site Model (CSM);

Define and evaluate the existing LTMMP program by assessing the fate and transport of

Contaminants of Concern (COCs), the CSM, and the groundwater model;

Specify all Data Quality Objectives (DQOs) to be utilized in remedy performance assessments within the established groundwater decision framework;

Incorporate the barrier wall remedy for the SHL and the Long Term Monitoring (LTM) for the NIA; and Incorporate necessary monitoring for the LUCs in the NIA.

Section 2.0 of this report summarizes the existing LTMMP technical approach as it relates to the CSM, remedy performance objectives, and new data collection. Section 3.0 of this report presents the DQOs for the revised monitoring and maintenance for the landfill, barrier wall, arsenic treatment plant, and other monitoring locations. Section 4.0 summarizes all the updated LTMMP monitoring procedures, analyses, frequencies, and quality assurance/quality control and data validation. Section 5.0 summarizes the Institutional Control Monitoring Plan.

Finally, Section 6.0 outlines all necessary reports to be completed following each monitoring event within the LTMMP.

1.2 Background and Site Description

Devens, Massachusetts (MA) is located approximately 35 miles northwest of the city of Boston, within the towns of Ayer, Shirley (Middlesex County), Harvard and Lancaster (Worcester

County). The former Fort Devens was established in 1917 for military training and logistical support during World War I. Fort Devens became a permanent base in 1931, and continued service until its Base Realignment and Closure Committee closure in 1996. Figure 1 depicts the area and topography of the former base and surrounding area.

SHL encompasses approximately 84 acres in the northeast corner of the main post of the former Fort Devens (Figure 2). The landfill is bordered to the northeast by Plow Shop Pond, to the west by Shepley’s Hill, to the south by recent commercial development, and to the east by land formerly containing a railroad roundhouse. Nonacoicus Brook, which drains Plow Shop Pond, is located north of the landfill.

SHL was reportedly operating by the early 1940s, and evidence from test pits within the landfill suggests earlier usage, possibly as early as the mid-nineteenth century. The landfill contains a variety of waste materials, including incinerator ash, demolition debris, asbestos, sanitary wastes, glass, and other wastes. The maximum depth of the refuse occurs in the central portion of the landfill and is estimated to be about 40 feet below ground surface (bgs). The volume of waste in the landfill has been estimated at over 1.3 million cubic yards (cy), of which approximately 160,000 cy (11%) is below the water table. The saturated wastes appear to be emplaced in a wetland; at least two areas previously mapped as wetlands were filled (Harding ESE, 2002) and have been found to be underlain by peat deposits (Sovereign, 2011).

The landfill was closed in five phases between 1987 and 1992-93 in accordance with Massachusetts Regulations at 310 CMR 19.000. The Massachusetts Department of Environmental Protection (MassDEP) approved the closure plan in 1985. Closure consisted of installing a 30 to 40-mil polyvinyl chloride (PVC) membrane cap, covered with soil and vegetation and incorporating gas vents. Closure also included installation of wells to monitor groundwater quality around the landfill, and construction of drainage swales to control surface water runoff. MassDEP issued a Landfill Capping Compliance Letter approving the closure in February 1996.

Subsequent to closure of the landfill, remedial investigations (RIs) completed under the Comprehensive Environmental Response, Compensation, and Liability Act (CERCLA) evaluated soil, sediment, surface water, and groundwater conditions at and in the immediate vicinity of the landfill. The results confirmed the presence of various contaminants, particularly certain inorganics including arsenic and volatile organic compounds (VOCs), in groundwater, sediment, and surface water at or adjacent to SHL. A Feasibility Study (FS) and ROD resulted in a remedy that required long term monitoring and maintenance of the existing landfill cap and groundwater monitoring.

The ROD (USAEC, 1995) required the Army to perform groundwater monitoring and five-year reviews to evaluate the effectiveness of the selected remedial action, which relied heavily on the previously installed landfill cap to attain groundwater cleanup goals by 2008 and to reduce potential exposure risks. If groundwater contaminant concentrations, primarily arsenic, met risk-based performance standards (cleanup goals) over time, the ROD did not require further action; however, if cleanup goals were not met, the ROD required implementation of a groundwater extraction contingency remedy. Due to continued elevated contaminant concentrations, the Army installed and operated a groundwater extraction and treatment system in March 2006 as a contingency remedy to address groundwater contamination emanating from the northern portion of the landfill (CH2M Hill, 2005a).

In 2011, the AOC 72 RI (AMEC, 2011) concluded that components of the current remedy – landfill capping and groundwater extraction – did not eliminate groundwater flow and arsenic migration from SHL into Red Cove / Plow Shop Pond, identified as Area of Contamination (AOC) 72. The AOC 72 RI results suggested that groundwater discharge contributed arsenic to sediment that could accumulate to levels resulting in conditions that posed unacceptable risks, and therefore a remedy that minimized such arsenic-in-groundwater flux to Red Cove would be most protective. Consequently, a low-permeability groundwater barrier wall was installed between the SHL and AOC 72 as part of a Non-Time Critical Removal Action (NTCRA) from August to September 2012 to mitigate arsenic flux to Red Cove/Plow Shop Pond by groundwater flow from the SHL. Documentation of the barrier wall installation was provided in the Removal Action Completion Report (Sovereign, 2013d).

1.3 Remedial Action Objectives

The following Remedial Action Objectives (RAOs) were stipulated in the ROD (USAEC, 1995):

Protect potential residential receptors from exposure to contaminated groundwater migrating from the landfill having chemicals in excess of maximum contaminant levels (MCLs).

Prevent contaminated groundwater from contributing to the contamination of Plow

Shop Pond sediments in excess of human health and ecological risk-based concentrations.

1.4 Summary of Remedy Components to Address RAOs

The current components of the SHL remedy selected to address the RAOs are as follows:

Landfill Capping Remedy Component: The landfill was closed in five phases between 1987 and 1992-93 in accordance with Massachusetts regulations at 310 CMR 19.000. The MassDEP approved the closure plan in 1985. Closure consisted of installing a 30 to 40-mil PVC membrane cap, covered with soil and vegetation and incorporating gas vents. Closure also included installation of wells to monitor groundwater quality around the landfill, and construction of drainage swales to control surface water runoff. MassDEP issued a Landfill Capping Compliance Letter approving the closure in February 1996. Inspections of the landfill cap are conducted yearly and include vegetative maintenance, landfill gas monitoring, and visual inspections of the capped area. Results and/or corrective actions are detailed in annual reports.

Groundwater Extraction Contingency Remedy Component: In the years following the capping of the landfill, data gathered at SHL indicated that the capping of the landfill was not resulting in a reduction of arsenic concentration in groundwater north of the landfill as originally expected. This triggered the installation of a contingency supplemental remedy for

SHL, a groundwater extraction system/Arsenic Treatment Plant (ATP). The ATP was designed to remove arsenic from extracted groundwater through co-precipitation with iron followed by microfiltration (MF). The extraction system consists of two extraction wells (EWs) located at the northwestern portion of the landfill cap. These extraction wells, EW-1 and EW-4 are capable of achieving the required combined target extraction rate of 50 gallons per minute (gpm) by either operating simultaneously or independently of one another to maximize plant influent flow.

Subsequently, groundwater enters the ATP influent stream, and then is dosed with chlorine dioxide which oxidizes and precipitates the inorganic metals, arsenic, iron, and manganese.

These precipitates are then filtered by a microfiltration system, and the effluent or treated water is discharged to the Devens publically owned treatment works (POTW) collection system.

Every 15 minutes, the MF control unit conducts flux maintenance (FM), which backwashes the filtered precipitates from the membranes. These solids are fed to the inclined plate clarifier (IPC) and allowed to settle out of suspension and form a residual sludge. The backwash effluent supernatant is fed through two bag filters configured in parallel and discharged to the plant effluent sump. The sludge is then pumped out of the IPC, dosed with polymer to increase flocculation, and carried over to the filter bed roll-off (FBRO). The accumulated sludge is removed from the plant at least once a month for disposal. This remedy has been in place since September 2005.

Barrier Wall Remedy Component: Following several years of operation of the ATP and monitoring of the landfill cap, it was determined that neither remedy was preventing the flow of impacted groundwater to the Red Cove area of Plow Shop Pond. To mitigate the arsenic-in-groundwater flux from SHL to Red Cove/Plow Shop Pond and reduce risk to environmental receptors consistent with local conditions in Plow Shop Pond, a low permeable barrier wall was installed along the eastern limit of the landfill and to the west of Red Cove in 2012 as part of a NTCRA. The barrier wall extended from the ground surface, through the landfill cap and a thin mantling of waste, through native sandy glacial deposits and glacial till, and to the bedrock surface. The boundaries and length of the barrier wall were based on the identified areas of impacted sediment in Red Cove, groundwater concentrations along the eastern edge of the SHL, and particle track analysis as predicted by the SHL groundwater model. The barrier wall was designed to intercept and divert groundwater flowing in the overburden soils away from Red Cove. It consists of an 850-foot long minimum barrier that extends through the overburden soils to the top of competent rock, with an effective hydraulic conductivity of 1 x 10-7 cm/sec.

1.5 Background of Existing LTMMP

The ROD and the original LTMMP established incremental reduction of risk rather than incremental reduction in concentration of individual contaminants as a measure of progress toward attainment of cleanup levels to focus on the cleanup of arsenic, which was the primary contributor to risk.

The existing LTMMP provides the basis for monitoring groundwater within and adjacent to the SHL, landfill gas sampling, and landfill inspections that have been conducted since the mid- 1990s and includes monitoring of the arsenic groundwater extraction, treatment, and POTW discharge system. Therefore, as outlined above, the existing LTMMP is germane to only the landfill cap and the ATP remedy components. The LTMMP provides a framework of operation, monitoring, and sampling to meet the objectives of the ROD (USAEC, 1995). During the five-year review in 2007, the revised LTMMP made use of methods utilized historically and optimized the location and frequency of monitoring based upon historical analytical data collected under the LTMMP and the implemented goals of the ATP Contingency Remedy Component. This LTMMP Update is designed to outline a revised monitoring and maintenance plan for all of the planned and implemented remedy components at SHL, inclusive of the landfill cap, the ATP, the barrier wall, and the impacts in the NIA.

2.0 EXISTING LTMMP PROGRAM AND CONCEPTUAL SITE MODEL

The following is a description of the existing LTMMP Program, overall conceptual site model and status of the continued evaluation of selected remedy components in achieving the necessary RAOs at SHL.

2.1 Summary of the Current LTMMP

The current LTMMP program consists of the collection of data to monitor the performance of the landfill cap and the ATP system conducted through the long-term monitoring of groundwater and landfill gas. The objective of the current program was to provide a comprehensive, revised LTMMP, thereby merging previous LTM and remedy performance monitoring activities into a single program. It was/is intended to be a dynamic monitoring program that will be further optimized through the process of annual evaluations of collected data and the issuance of annual reports with recommendations.

The objective and technical approach of the current program consists of a series of quantitative monitoring programs designed to meet the goals of the ROD (USACE, 1995) such as hydraulic monitoring including quarterly and semiannual sampling and gauging events at select wells and treatment system operation and maintenance including monthly and/or quarterly monitoring of system influent and effluent. The Revised Long Term Monitoring and Maintenance Plan for Shepley’s Hill Landfill Devens, Massachusetts, (CH2M Hill, 2007) and Revised Long Term Monitoring and Maintenance Plan Addendum –Shepley’s Hill Landfill and Treatment Plant, Long-Term Monitoring and O&M Services (ECC, 2009) present in great detail the existing LTMMP. The current LTMMP network of wells was selected based on remediation effectiveness from the evaluation of historical data in conjunction with annual landfill cover and treatment plant monitoring.

Annual performance assessments of the current program have been focused on system hydraulics and capture/control of groundwater at the north end of SHL. Consistent with EPA guidance including A Systematic Approach for Evaluation of Capture Zones at Pump and Treat Systems (USEPA, 2008), a multiple lines of evidence approach has been taken with respect to the performance assessment. The individual assessment components, their data requirements, and a brief summary of the results are provided in the various Shepley’s Hill Landfill Annual Reports and within the 2010 Five Year Review Report.

The data quality objectives for the collection of future data outlined in this LTMMP are to gather the data necessary to document and evaluate the performance of all of the remedy components, including the landfill cap/cover, the arsenic treatment plant, the barrier wall and the associated long-term monitoring of environmental media designed to document the performance of the selected remedy. The data are also used to further expand upon the overall CSM as it relates to the long term performance of the remedy. The CSM is detailed in Section 2.2., below.

2.2 Conceptual Site Model

2.2.1 Background / Summary

The CSM for SHL is updated through the collection of new data which include but are not limited to supplemental investigations conducted between 2009 and 2014 as documented in the 2009 Supplemental Groundwater and Landfill Cap Assessment for Long-Term Monitoring and Maintenance Report (AMEC, 2009), the 2011 Shepley’s Hill Landfill Supplemental Groundwater and Landfill Cap Assessment for Long-term Monitoring and Maintenance - Addendum Report (Sovereign, 2011), and the Shepley’s Hill Landfill 2012 and 2013 Annual Reports (Sovereign, 2013c and 2014a) and long term monitoring and operation of the ATP. Potential sources of arsenic in groundwater include bedrock, till, landfill waste, peat, and aquifer sand overlying bedrock and underlying waste or peat. Due to the placement of the cap on the landfill, any potential leachate from the landfill waste is now limited to the ~10% of the waste that is present within the saturated zone. Recent studies (Harding ESE, 2002; Sovereign, 2011) indicate that the predominant source of the dissolved arsenic beneath the landfill is naturally occurring arsenic entrained in iron oxyhydroxides in the aquifer sand that is released into groundwater from the aquifer sands by naturally occurring and landfill-induced reducing conditions caused by carbon degradation and oxygen depletion leading to anaerobic conditions. Evidence for this conclusion is several-fold and includes results of vertical profiling in the landfill that does not exhibit an arsenic vertical concentration profile suggestive of the landfill waste as the primary source of arsenic in the system, results of scanning electron microscopy of aquifer sands detailing the prevalence and volume of arsenic entrained in iron oxyhydroxides, and column studies associated with the flushing of oxygen-depleted water through aquifer sands resulting in the release of dissolved arsenic in the test cells. In addition, concurrent research on arsenic occurrence in groundwater in other parts of the world noted the importance of buried peat layers in mobilizing arsenic through the very same reductive dissolution mechanism (Appelo, 2006).

There is further evidence that indicates the landfill is not the primary source of arsenic and that conditions favoring a natural origin for the elevated arsenic in groundwater are known to be present [e.g., regional occurrence of high arsenic in both bedrock minerals and in overburden iron oxyhydroxide coatings, presence of peat deposits, low oxidation-reduction potential (ORP), etc.] (Gannett Fleming, 2011). If significant amounts of arsenic leached from the landfill waste, then the underlying sands would be enriched with arsenic (Keimowitz et al., 2005). This has not been observed (Sovereign, 2011). The arsenic concentrations in the soil profiles increase with depth to the top of the till and bedrock, and arsenic contents found in the aquifer sands are similar in concentration as those found locally and regionally in the vicinity of Fort Devens (USACE, 2004). Arsenic in all materials (aquifer sand, waste, bedrock, etc.) is mobilized by reducing conditions at the site, and this process will persist for as long as reducing conditions remain.

There are two sources of carbon and reducing conditions at the SHL. Historically, the peat and wetlands underlying the landfill and in the NIA likely provided reducing conditions that mobilized arsenic through reductive dissolution of iron oxyhydroxides on which arsenic was entrained. This would have occurred prior to the development of the landfill extending back over 10,000 years, as this process is well documented in published scientific literature. The landfill was placed on top of the existing wetlands and underlying peat, and degradation of the waste rapidly created additional reducing conditions.

These processes are similar to those noted at the Winthrop Landfill in southern Maine (Keimowitz, 2005). For example, (1) the aquifer at both sites has an arsenic source not derived from landfill waste but from geologically naturally occurring arsenic that was and is mobilized by reducing conditions imposed by landfill waste and by peat deposits and wetlands (bogs); (2) studies at SHL (Harding ESE, 2002; Sovereign, 2011) and Winthrop (MACTEC, 2006) have shown that the source is arsenic entrained in iron oxyhydroxides in the aquifer sand underlying and surrounding the landfill; and (3) evidence indicates the landfills were unlikely significant sources of arsenic. Arsenic leached from the landfill waste would have enriched the sands, especially immediately below the waste, with arsenic (Keimowitz et. al., 2005). This has not occurred at either site, and the arsenic concentrations in the soil profiles at SHL generally increase with depth to the top of the till and bedrock. Further, (4) both landfills are impacted by pre-existing wetlands and underlying peat deposits whereby degradation of the landfill waste created additional reducing conditions that added to the mobilization of the arsenic in the underlying aquifer sands and increased the aerial extent of the reducing conditions beyond the boundaries of the wetlands and peat. The main exception is that peat bogs and wetlands lie adjacent to the Winthrop landfill and not beneath it such as exists at SHL. The ability of peat and reduced wetlands to behave like landfill waste as a source of reducing conditions is well documented (Bozkurt et al., 2001).

Further, the mechanisms responsible for the elevated arsenic at the SHL appear to be the same as those at the Winthrop Landfill where arsenic contamination occurs (Keimowitz et. al., 2005).

The difference between the concentrations of arsenic in groundwater at SHL and at the Winthrop Landfill is attributed to the difference in concentrations of naturally occurring arsenic located in the aquifer material beneath each landfill. The aquifer material at SHL contains an average of 14,000 µg/kg in the upper aquifer sands in contrast to the average of 4,900 µg/kg arsenic reported for aquifer material at Winthrop. More importantly, the bottom 10-20 feet of each boring at the SHL consisted of sand, glacial till or bedrock containing an average 38,000 µg/kg of arsenic. Thus a higher potentially soluble source of arsenic exists at SHL compared to Winthrop, and the SHL inventory of arsenic can be expected to be an order of magnitude greater than that found at the Winthrop Landfill. In addition, it is important to note that the pump and treat system at the Winthrop Landfill was ultimately terminated, as it was determined to be not effective in remediating arsenic to achieve restoration of the aquifer and that land use and institutional controls were sufficient to meet the RAO for receptor protection

(MACTEC, 2006).

Estimates of the time it would take to flush aqueous phase arsenic in the system to background conditions through the landfill and up to Nonacoicus Brook and wetlands approach 300 years under best case conditions (clean, oxic water replaces groundwater with no arsenic remobilization). The residual or background level of arsenic that is achievable by flushing is not known but could approach 1,500 ug/L based on the solubility of expected residual arsenic solid phases (Appelo, 2006; Smedley and Kinniburgh, 2002; USGS, 2011). Thus even with no new additions of arsenic from any non-native source, significant time to achieve local background is required and the new ambient arsenic level will almost certainly be several orders of magnitude above current MCLs for groundwater. Estimates of flushing residual carbon in the landfill footprint to lessen reducing conditions is estimated to be at least two times (2x) the time it takes to flush arsenic from the system or over 500 years (Keimowitz, 2005; Bozkurt et al., 2001).

Restoration of the aquifer to MCLs or even to less than 100 ppb throughout the area downgradient of the landfill in the NIA in a reasonable time frame appears unlikely given (1) the volume of naturally occurring arsenic in aquifer sands beneath SHL and longevity of reducing conditions exacerbated by the presence of the landfill, and (2) the continuing enrichment of the aquifer sands and groundwater with arsenic via the upwelling of arsenic rich groundwater from Shepley’s Hill as documented by recent EPA studies.

The existing data set from groundwater investigations along Nonacoicus Brook does not suggest that arsenic is discharging to the Brook at appreciable concentrations and continues to suggest that an oxygenated zone is present which naturally precipitates arsenic into iron solids near or beneath Nonacoicus Brook as the low-dissolved oxygen groundwater mixes with oxidized water from the north and beneath the Brook (Sovereign, 2014a). Investigations completed in 2013 and 2014 continues to document that arsenic remains at depth, more than 40 feet below the Brook elevation and, taken with the 2010 data collected north of the Brook, indicates that the arsenic concentrations appear to decline rapidly at depth in proximity of Nonacoicus Brook, which appears to represent a groundwater discharge divide.

Historically, elevated arsenic concentrations in groundwater at SHL have impacted Red Cove/Plow Shop Pond which is located down-gradient and in close proximity to the northern portion of the landfill. To mitigate arsenic-in-groundwater flux from SHL to Red Cove/Plow Shop Pond and reduce risk to environmental receptors consistent with local conditions in Plow Shop Pond, a low permeable barrier wall was installed in 2012 along the eastern edge of SHL.

The barrier wall has subsequently intercepted and diverted groundwater flowing in the overburden soils away from Red Cove and toward the northern end of the landfill. With the installation of the barrier wall between the landfill and Red Cove, the arsenic flux to Red Cove is expected to be significantly reduced (Sovereign, 2013d). The effects of the barrier wall are being monitored and the CSM will be updated as necessary to account for the effects of the barrier wall with respect to flow and flux to the east and north.

2.2.2 Contaminant Fate and Transport

As outlined above, arsenic is released into groundwater from the aquifer sands and bedrock by both naturally-occurring and landfill-induced reducing conditions caused by carbon degradation and oxygen depletion that lead to anaerobic conditions. Portions of the landfill overlay pre-existing, buried peat deposits that induced reducing conditions prior to emplacement of the landfill over the buried peat and associated wetlands. Therefore, it can be concluded that the buried peat deposits within the landfill footprint also likely caused arsenic mobilization to the north end of the site toward Nonacoicus Brook as well as east toward Plow Shop Pond prior to the placement of waste.

2.2.2.1 North Impact Area

In order to refine the understanding of the extent of chemically-reducing conditions in the NIA and update the CSM, a supplemental investigation was conducted in the spring of 2013 and the winter of 2014 in the NIA. The scope of this investigation was detailed in the May 2013 Work Plan for Long-Term Monitoring and Maintenance Update (Sovereign, 2013b). As part of this investigation dissolved arsenic concentrations, ORP, dissolved oxygen, and other geochemical parameters were measured at select locations in the NIA. Components of this evaluation included the completion of vertical arsenic profiling, permanent monitoring well installation, and monitoring well sampling and analysis. Data tables summarizing the data collected in the spring 2013 and winter 2014 supplemental investigation are presented in Appendix A, and a full discussion and interpretation of all of the data will be provided in the 2013 and 2014 Annual Reports (Sovereign, 2014a).

Well sampling north of the landfill and in the NIA since 2001 indicates that both in-situ carbon degradation and the presence of the landfill has resulted in reducing conditions in the aquifer.

This has been confirmed by the low dissolved oxygen, elevated dissolved methane concentrations, elevated dissolved carbon, elevated ammonia concentrations, and elevated arsenic and iron concentrations. Thus, both the geochemistry of the landfill has induced reducing conditions and the naturally occurring conditions continue to mobilize arsenic in that area.

Nonacoicus Brook appears to represent a groundwater discharge divide. Recent (2013) sampling continues to document no elevated arsenic in the monitoring wells directly north of the brook. The bedrock delineation and general elevation of the northern-most wells indicates that the bedrock surface is much higher in elevation on the north side of the wetlands and brook than the southern side. Hydraulic data gathered from wells on the north side of the brook suggest a westerly/southwesterly groundwater flow component. This flow of groundwater from the north contains higher concentrations of dissolved oxygen that would create a redox boundary which should precipitate arsenic into iron solids near or beneath Nonacoicus Brook as oxygen-depleted groundwater emanating from the landfill area migrates north and mixes with oxidized water from the north and beneath the Brook.

Recent work (2013) included advancement of vertical profiles and groundwater monitoring wells immediately near the southern edge of Nonacoicus Brook to address concerns that arsenic may discharge to the Brook in localized areas. As presented in the 2013 Shepley’s Hill Landfill Annual Report (Sovereign, 2014a), arsenic-impacted groundwater was encountered at 50 to 60 feet below grade immediately south of the Brook and has not been encountered from 10 to 40 feet below grade at each location based on the results of groundwater profiling activities conducted in 2013 at SHM-13-03 and in 2014 at SHM-13-14S/D and SHM-13-15. Consequently, the existing data set does not suggest that arsenic is discharging to the Brook at appreciable concentrations and continues to suggest that an oxygenated zone is present which naturally precipitates arsenic into iron solids near or beneath Nonacoicus Brook as the low-dissolved oxygen groundwater mixes with oxidized water from the north and beneath the Brook. Work completed in 2013 and 2014 continues to document that arsenic remains at depth, more than 40 feet below the Brook elevation and, taken with the 2010 data collected north of the Brook, indicates that the arsenic concentrations appear to decline rapidly at depth in proximity of the Brook.

Previous modeling work suggested that groundwater flow direction curves westward as groundwater approaches the brook and previous assessments assumed that as groundwater flow curved westward, elevated concentrations of arsenic would be found in a similar pattern.

However, the amalgam of data collected between 2001 and 2014 at this time do not show any elevated arsenic in groundwater in monitoring wells installed in line with the groundwater flow bend to the west at the Brook. Arsenic appears to remain in the aquifer in a relatively narrow band trending north, between profile point SHM-10-21 and SHM-10-25 as shown on Figure 3.

As the existing LTMMP does not include the monitoring of any of the new investigation points in the NIA, Section 3, below, provides updates to the monitoring plan that will provide long term monitoring of locations in the core of the impact area, along the edge of the Brook and in downgradient locations to the west.

2.2.2.2 Red Cove

Elevated arsenic concentrations in groundwater at SHL have subsequently impacted Red Cove/Plow Shop Pond which is located down-gradient and in close proximity to the northern portion of the landfill. Red Cove is a shallow cove with a water depth of less than one meter.

As detailed by AMEC within the 2011 Remedial Investigation Report for AOC 72, arsenic flux to Red Cove was estimated at approximately 14.6 to 20 g/day with the landfill cap in place before the groundwater extraction and ATP were installed (AMEC, 2011).

The Army evaluated whether a significant risk to human health or the environment exists at Plow Shop Pond/Red Cove (AMEC, 2011) and determined the evaluation of a removal action was warranted to reduce current and potential risks to human health and the environment posed by contaminants that originate from SHL. As a result, a low-permeability groundwater barrier wall between the SHL and AOC 72 was determined to be an acceptable SHL remedy component to help mitigate impacts associated with AOC 72. The installation of the barrier wall in 2012 along the eastern edge of SHL, in combination with the landfill cap and ATP remedy components, was intended to meet the RAO objective (i.e., prevent contaminated groundwater from contributing to the contamination of Plow Shop Pond sediments in excess of human health and ecological risk-based concentrations).

Prior to the installation of the barrier wall, supplemental pre-construction data collection activities were performed from 2011 to 2012 in the area of the proposed wall as detailed in the 2012 Removal Action Work Plan for the Shepley’s Hill Barrier Wall (Sovereign, 2012) to refine and update the CSM in the area of the proposed wall and address several of the field data needs identified during the conceptual design of the barrier wall. As part of this investigation, geotechnical composition of the submerged aquifer sands, bedrock depth and competency along the proposed wall, hydraulic conductivity of the shallow bedrock aquifer, and arsenic concentrations in groundwater along the proposed wall were evaluated.

The geotechnical samples collected from the overburden documented a generally homogeneous overburden consisting of loose sand material generally from the surface to bedrock, with locally-absent layers of dense till material immediately above bedrock. No significant geological variation was observed over the length of the wall. The maximum depth to bedrock was 64 feet below grade, and the shallowest depth to rock was 20 feet below grade. Bedrock hydraulic conductivity and transmissivity were determined to be low, and groundwater arsenic concentrations within the bedrock ranged from 71 µg/L in shallower fractures to 3 µg/L in deeper fractures. Conversely, dissolved arsenic profiling in the overburden documented arsenic concentrations ranging from 269 µg/L to 512 µg/L. The difference in concentration from the overburden to the shallow rock fractures suggested that the primary source of the arsenic flux into Red Cove was through the overburden (Sovereign, 2012).

Following the installation of the wall, hydraulic monitoring events were conducted periodically along both the up- and down-gradient sides of the wall to provide hydraulic monitoring data for the barrier wall. Results of the monitoring events demonstrated a positive difference in hydraulic head between the up- and down-gradient monitoring locations along the barrier wall and indicated that the barrier wall was effective in mitigating flow to Red Cove/Plow Shop Pond. With the barrier wall in place, flow patterns in the Red Cove area have changed permanently, with reduced gradient toward the pond east of the wall and greater gradient to the north on the west side of the wall (Sovereign, 2014a). Consequently, with the installation of the barrier wall between the landfill and Red Cove in 2012, the arsenic flux is expected to be significantly reduced (see Section 3.5.2).

The investigation data collected prior to and following the installation of the barrier wall has been utilized in the LTMMP to refine the understanding of the CSM and to evaluate remedy performance. However, the existing LTMMP does not incorporate sufficient monitoring for the long-term evaluation of the effectiveness of the barrier wall in achieving the RAO and, therefore, will be updated to achieve that DQO (Section 3).

2.3 Groundwater Model Update

An update to the Shepley’s Hill Landfill groundwater flow model version SHL104 was completed in 2013(Sovereign, 2013f). The update included a series of significant revisions, as well as a thorough review and modification of various model parameters based upon available data where possible. To address the 2014 BCT comments on the model, ongoing model revisions will be documented in a separate report to be finalized and submitted in 2015.

During the annual reporting process, the model will incorporate LTMMP generated site-wide hydraulic data to continually evaluate model calibration and sensitivity and will be utilized to conduct advective travel time analysis and reverse (backward in time) and forward (forward in time) particle tracking simulations to evaluate remedy performance. The results of the model analysis will be documented in each subsequent Annual Report for SHL.

2.4 Evaluation of New Data to Existing LTMMP

The USEPA installed a series of piezometers in the area of the ATP extraction wells in 2012 to delineate the ATP capture zone and to provide a baseline of data from this area following the construction of the barrier wall. Two piezometers, the first screened across the water table and the second screened within the deep overburden aquifer, were installed at each location (Lockheed Martin, 2012). These wells will be incorporated into the LTMMP monitoring program (see Section 3.1.2), and hydraulic and geochemical data from these piezometers will be used with data collected from other nearfield and downgradient monitoring wells to evaluate remedy performance in the area of the ATP through gradient vector analysis, capture zone width calculation, drawdown assessment, and model simulations.

The results of the data collected to-date define the down-gradient extent of the arsenic impacted groundwater at Nonacoicus Brook in a band measuring approximately 300-350 ft wide bound generally between SHM-10-10 to the west and SHM-10-27 to the east, at a depth of 25 to 50 ft below grade. This observation along with the geochemistry data suggests that either Nonacoicus Brook is protected from arsenic impacts by naturally occurring redox conditions near the Brook and/or the groundwater flow divide north of the Brook and/or the extent of arsenic-impacted groundwater at this area has reached its downgradient extent of migration.

This redox zone appears to be located in the vicinity of SHM-10-10. The boundary appears to consist of three features: (1) a bedrock surface that controls the flow of landfill impacted water to the Brook but also brings groundwater from the north and northeast of the Brook that counters the landfill flow, (2) intrusion of more oxidized groundwater from the north side of the landfill, and (3) mixing of clean water resulting in precipitation of arsenic that does not impact the water quality in the Brook or wetlands.

Further, the data collected during the 2010 through 2014 field investigations are consistent with data collected historically throughout the NIA. This indicates the arsenic plume in the NIA is stable and limited to an area along West Main and Shirley Streets. In addition, data collected from the western area of the NIA does not indicate that the core of the arsenic impacted groundwater extends westward, but rather trends roughly north.

The northern most wells currently monitored as part of the LTMMP are located along Sculley Road. Based on recent data from the NIA, select wells between the current LTM wells and Nonacoicus Brook will be added to the LTM well network to monitor the fate and transport of arsenic as reducing groundwater approaches the Brook and to monitor the overall stability of arsenic concentrations in the core of the impacted area beneath West Main Street.

Furthermore, the existing LTMMP does not incorporate sufficient monitoring for the long-term evaluation of the effectiveness of the barrier wall. Results of the initial monitoring events conducted upon the completion of the barrier wall indicate a positive difference in hydraulic head between the up- and down-gradient monitoring locations along the barrier wall and that the barrier wall is effective in mitigating flow to Red Cove/Plow Shop Pond. Consequently, a long-term monitoring program in the area of the barrier wall will be implemented, and select wells located on both the up- and down-gradient side of the barrier wall will be added to the LTM well network to monitor the long-term effectiveness of the wall.

3.0 UPDATED LTMMP PROGRAM

This LTMMP Update modifies the current monitoring well network at SHL to enhance the ability to evaluate the effectiveness of the individual remedial components underway at SHL that together encompass the remedy. To this end, this LTMMP update addresses five remedial program elements with the overall goal and strategy of providing sufficient data to proceed forward along the groundwater decision framework and monitor the remedy performance for SHL. These five elements include:

1. Continued maintenance of the landfill cap;

2. On-going monitoring and performance evaluation of the ATP remedy;

3. Hydraulic and geochemical performance monitoring of the barrier wall remedy;

4. Performance monitoring of the LUCs for the NIA; and

5. An update to the groundwater monitoring well network at SHL encompassing select monitoring locations installed between 2010 and 2014 and on-going maintenance of institutional controls institutional controls in the NIA.

Concerning the fourth and fifth elements and the completion of additional assessment activities between 2010 and 2014 in the area of impacted groundwater north of Sculley Road and the railroad right of way (referred to as the NIA), it is anticipated that long term monitoring will become a component of the remedy that will address groundwater in the NIA and will be formalized into the ROD through a future ESD. Whereas natural subsurface processes such as dispersion, volatilization, biodegradation, adsorption, and chemical reactions with subsurface materials can reduce COC concentrations, the extent and rate of attenuation depends on a variety of parameters such as COC types and concentration, temperature, moisture, and redox state. For an inorganic COC such as arsenic, fate and transport of the COC is the primary factor monitored as well as the nature and extent of reducing waters (aquifer geochemistry) present in the areas of attainment. The implementation of LUCs in the NIA in 2014 (see Figure 4 for the area of LUCs) supplemented the LTM in the NIA by eliminating future potential for direct exposure to groundwater in the NIA through prohibiting the use or installation of drinking water or irrigation wells in the impacted area.