The Savage Municipal Water Supply Well Superfund site is located in the town of Milford approximately two miles west of the center of town. The contaminated plume at the site extends approximately 6,000 feet eastward from the intersection of Route 101 and Elm Street. The site lies within the floodplain of the Souhegan River.
The Savage Municipal Water Supply Well was a water supply for Milford from 1960 to 1983. In February 1983, the New Hampshire Department of Environmental Services (DES) detected concentrations of volatile organic compounds (VOCs) in the Savage Well that were above drinking water standards, and the well was shut down. The site received Superfund emergency monies to temporarily provide bottled water to residents of a nearby trailer park whose water supply well was also affected. The trailer park was subsequently connected to the town's public water supply. The site was added to the National Priorities List in September 1984.
In 1985, DES completed a hydrogeological evaluation of the site, identifying four manufacturing plants surrounding the well as potentially responsible parties (PRPs).
In August 1987, the Environmental Protection Agency (EPA) and the PRPs signed a Consent Decree in which the PRPs agreed to conduct the Remedial Investigation/Feasibility Study (RI/FS). The PRPs began fieldwork in the summer of 1988. EPA accepted the RI/FS in 1991, and EPA's Record of Decision (ROD) was issued in September 1991.
In subsequent negotiations, the site was divided into two areas of remedial responsibility. EPA and the State are undertaking the cleanup of the OK Tool Source Area (OU-1) where VOCs were released into the subsurface from the now-closed OK Tool manufacturing facility. The PRPs are responsible for the cleanup of the second area (OU-2), which includes the extended plume of contaminated groundwater in the Milford-Souhegan aquifer.
The remedial action for OU-1, the OK Tool source area includes a physical barrier to contain the source area, in conjunction with pumping and treating the groundwater. Also, air sparging and soil vapor extraction were included in the remedy.
Construction of the 1,500-foot circumference slurry wall/barrier (containment area) and the treatment plant began in 1997 and was completed in March of 1999.
Since plant operations began in 1999, monitoring well sampling of the overburden aquifer indicate contaminant reductions both inside and outside of the slurry wall containment area. This trend demonstrates that the remediation system is functioning as intended.
After extensive investigations to characterize and delineate source areas inside the containment area, EPA and DES performed an in-situ chemical oxidation (ISCO) pilot test in September of 2003 and 2004. The ISCO test included injection of potassium permanganate into the subsurface. Potassium permanganate destroys dissolved chlorinated solvents such as PCE using oxidizing reactions.
Additional investigations to characterize and delineate potential source areas, and their relationship to former waste handling facilities in the western part of the containment area were conducted in 2007 and 2008. The investigations included vertical profiling of suspected DNAPL release areas and included groundwater, soil gas and soil sampling. These investigations aided in designing follow-up ISCO injections and vadose zone excavation and treatment as described below.
In 2008 the treatment system at OU-1 was modified. Modifications consisted of the removal of the SVE blower and air sparge compressor, and replacement of two low pressure steam boilers with two high efficiency water boilers. The small energy efficient boilers have significantly reduced fuel consumption at the OU-1 treatment plant.
During 2008/2009, approximately 2,000 cubic yards of vadose zone soils from the former leach field area at OU-1 were excavated and treated on-site using soil vapor extraction and chemical oxidation using ozone and peroxide. Following post treatment soil sampling which indicted the soils met cleanup standards, the soil was backfilled.
After evaluating the effectiveness of the ISCO pilot injections, three follow-up ISCO injections were performed at OU-1 in the saturated zone to reduce the mass of contamination inside the wall. These ISCO injections used sodium permanganate. The injections occurred in fall 2008, fall 2009 and late winter 2010. A total of approximately 300,000 pounds of 40 percent sodium permanganate were diluted and injected during the three injection events.
At OU-2, in accordance with the requirements of the Consent Decree, the PRPs completed construction of a 450-gallon per minute groundwater treatment plant in late fall 2004 to cleanup the extended plume of contaminated groundwater in the Milford-Souhegan aquifer. The OU-2 groundwater remediation system continues to operate at an average rate of approximately 400 gpm with treated water being discharged to the subsurface and Souhegan River. The metals treatment system is effectively treating metals, being compliant with both surface water effluent discharge limits established by DES and design requirements of the system.
The first Five Year Review was completed for the site in September 2011. Recommended actions include performing a shallow groundwater investigation to evaluate potential vapor intrusion risk at residential receptors in OU-2. Conduct evaluation of the OU-1 extraction system performance due to fouling from residual permanganate. Implement institutional controls; establish GMZ, Town Ordinance, Zoning and/or Deed Restriction. Establish interim cleanup levels and prepare decision document for 1,4-dioxane, vinyl chloride, cis-1,2-DCE and manganese, and modify arsenic to 10 ppb. The vapor intrusion assessment of residential receptors in OU-2 has been completed. The OU-1 residual permanganate fouling issues have been addressed.
Additional deep bedrock investigations are currently underway at the site. This deep bedrock investigation is designated as OU-3. Current analytical data from deep bedrock monitoring wells, located beneath the OK Tool Source Area, indicate high levels of chlorinated hydrocarbons indicating the possible presence of DNAPL in bedrock. A remedial investigation for OU-3 was initiated in 2012 and included pump testing to evaluate hydrogeology in the deep bedrock, potentially identify inter-fracture flow, and obtain a better understanding of the interaction between deep overburden and bedrock. The Remedial Investigation Report is expected to be completed in late 2013.
A Feasibility Study which looks at various technologies and alternatives available to address the risks posed by the deep bedrock contamination has been initiated. The FS report for OU-3 is expected to be completed in 2014.