Formatted Title

Designing a Treatment Solution Using Quantitative High Resolution Site Characterization Data to Determine Life Cycle Mass Flux and Discharge

Background/Objectives

The 1140 Broadway site in Ann Arbor, Michigan, has long been known to be contaminated with tetrachloroethylene (PCE) with little to no natural degradation to depths of 40 feet below grade.  The PCE plume traverses the site and migrates off site at the eastern boundary, towards the Huron River.  A new developer purchased the site to construct a mixed-use development.  As a condition to receive tax incentives from the City and the County, the developer agreed to install a permeable reactive barrier (PRB) at the eastern boundary with the goal of reducing the contaminant mass flux migrating off site by a minimum of 80%.  This goal was challenged by several constraints, chief among them that the site was not fully characterized due to data gaps in the existing historical data set and the complexity of the hydrogeologic conditions made it difficult to estimate mass flux/discharge.

Approach/Activities

The design of a PRB at the eastern boundary was challenged by the lack of adequate soil, groundwater, and contaminant data; non-continuous and variable glacial sediments; and the presence of partially separated upper and lower saturated units.  Additional challenges included a constrained construction schedule that would be active concurrent with the installation of the PRB.  The available area for the PRB was limited by the proposed building footprint to a narrow strip along the eastern boundary.  The project team evaluated moving forward with the PRB design based on mass flux estimates; however, the risk of failure was too high do to the highly variable hydraulic conductivity of the glacial sediments. The decision was not to estimate the rate mass would enter the PRB but to determine the total mass the PRB would manage based on what is distributed within the source and mid-plume upgradient of the PRB. To do so, a quantitative high-resolution site characterization (qHRSC) program was conducted at the site, comprised of high density remedial design characterization soil and groundwater sampling within the source and along the groundwater plume to create a quantitative 3D conceptual site model.  The sampling scope included 79 soil borings and 46 nested groundwater well clusters.  From the sample points, 1,120 soil and 185 groundwater samples were analyzed using 8260b analysis in a remedial design focused laboratory.  The 3D model was used to determine the total mass and distribution of mass in the formation that requires treatment by the PRB. The total mass provided definitive data that there was insufficient space to construct an effective, narrow PRB along the eastern boundary without source treatment to reduce the contaminant load on the PRB.  In response, the developer worked with the Michigan Department of Environment, Great Lakes, and Energy (EGLE) to secure a cleanup grant to treat the source area concurrent with the PRB installation.  The PRB was also split into two portions: one section (PRB1) was placed upgradient of the building under construction and the second (PRB2) was placed downgradient along the eastern boundary. 

Results/Lessons Learned

The results of the qHRSC program determined the total PCE present was greater than historically estimated, in total approximately 4,125 pounds (lbs) of PCE was present in a 60 foot-wide band across the site.  Concentrations of PCE were as high as 4,640,000 µg/kg in soil and 137,000 µg/L in groundwater.  Approximately 99% of the contaminant mass was PCE, indicating little natural degradation occurred over time. It was determined that treating the entirety of the 36,000 square foot on-site plume was not possible with the allocated tax incentives and grant dollars; therefore, the remedial approach focused on maximizing available funds to decrease the contaminant mass.  The quantitative 3D model allowed for the development of an optimized remedial design based on spatial mass loading as opposed to modelling contaminant mass flux using a sparse soil and hydrogeological data set.  The upfront remedial design characterization sampling resulted in a successful remedial approach as demonstrated by more than four years of post-remediation monitoring. The final remedial approach and post-remediation data will be shared during the presentation.