The Department of Defense (DoD) needs improved methods for estimating the mass of non-aqueous phase liquids (NAPLs) in subsurface environments contaminated from past releases. In addition, it is a significant challenge to predict the time needed to achieve specified remedial action objectives (RAOs), with or without NAPL depletion. Credible data are required to support management decisions regarding when and to what intensity active remediation efforts should be pursued at these sites. The primary objective of this project was to evaluate a methodology to improve evaluations of the extent of source remediation required at both light non-aqueous phase liquids (LNAPL) and dense non-aqueous phase liquids (DNAPL) impacted sites.

Specific quantitative performance objectives for the methodology evaluated during this project focused on three topics:

1. Accurate simulation of the groundwater flow field through a heterogeneous source zone
2. Accurate predictions of the NAPL architecture and contaminant mass discharge in the source zone
3. Accurate simulation of the measured reductions in contaminant mass discharge, which can result from either a reduction in the total NAPL mass or a change in the NAPL composition.

These quantitative objectives were assessed primarily by comparing the Sequential Electron Acceptor Model, 3D transport (SEAM3D) numerical model results to the observed field data. In addition, two qualitative performance objectives were evaluated— (1) the ease of implementing the field test procedures and (2) the cost to perform the test.

To address deficiencies in field measurements for assessing source zone depletion (SZD) functions, new approaches were field-tested at Site ST012 on the former Williams Air Force Base (WAFB, now known as Williams Gateway Airport), Arizona. In 2001, the Air Force initiated a study to evaluate remedial strategies for NAPL contamination at Site ST012, including field tests and modeling. From 2008 through 2010, the Air Force conducted and evaluated a pilot test of thermally enhanced extraction (TEE) as a suitable technology for reducing the mass and longevity of a multicomponent fuel source (jet fuel) residing in the saturated zone. A rising water table (approximately 4 feet per year) over the last two decades created a submerged smear zone of fuel NAPL spanning a depth of about 75 feet and resulting in a long-term source to groundwater of a number of chemicals of concern (COCs), including benzene and naphthalene.

During the TEE pilot test, a suite of innovative diagnostic techniques were used to evaluate the benefits of partial NAPL source reduction. These techniques included:

• Passive Flux Meters™ (PFMs), which allow simultaneous measurements of groundwater flow velocity and contaminant mass trapped on sorptive resin during the time that the PFM is left in a monitoring well
• Integral pumping tests (IPTs), which measure the contaminant mass in groundwater extracted over time, often with varying extraction rates
• Modeling using the solute transport code SEAM3D with an enhanced input SZD function.

Additionally, tracer tests were performed during the IPTs to characterize preferential and asymmetric groundwater flow paths. The field measurements (IPT and PFMs) were performed both before and after the TEE pilot test within a portion of the NAPL source zone at ST012. The testing provided data related to NAPL architecture and rates of mass transfer from the NAPL to the aqueous phase. The tests measured mass transfer characteristics on length scales varying from a few feet (PFM data) to the 70 foot distance between injection and extraction wells (IPT data) within the TEE cell. Groundwater samples from multiple extraction and monitoring wells provided data on intermediate length scales. The IPT and PFMs have not been combined previously to provide data for a mass transfer analysis with an appropriate model, with the intent of leveraging the advantage of each technique.

The field data were intended to provide input to a calibrated transport model. SEAM3D is an advective-dispersive numerical solute transport model that simulates the full range of natural attenuation processes in groundwater systems. SEAM3D also explicitly simulates the dissolution of a NAPL source zone based on fundamental mass transfer analyses and a calibrated SZD function for purposes of scaling the results from the TEE test scale to the entire NAPL zone at the site.

The data collected on the various scales before and after the TEE pilot test were synthesized into a working quantitative model of the NAPL architecture and mass dissolution rate for the SEAM3D enhanced SZD function. This approach seeks to circumvent the reliance (or at least reduce the emphasis) on long-term source depletion data to calibrate the SZD function associated with a site solute transport model. The aim is to reduce the uncertainty associated with estimates of source and plume longevity through the direct measurement of a bulk mass transfer coefficient, which is then used to produce more accurate modeling results of the estimated time of remediation (TOR) for the site. Because of the unique smear zone at Site ST012, the results are applicable to a broader class of sites than just those impacted with LNAPL, including those contaminated with DNAPLs.

The success criteria for both the quantitative and qualitative performance objectives were achieved, except at three monitoring locations for the third quantitative performance objective. This variance was the result of uneven thermal treatment across the cell, which was not captured by the modeling assumption of uniform NAPL composition across the cell. Although the SEAM3D can account for variability in NAPL residual saturation in space, this level of sophistication was not specified in the Demonstration Plan for this project.

The mass transfer tests (MTTs) and associated modeling were able to measure the bulk mass transfer coefficient directly and to relate the absolute source mass to the mass discharge. The methodology resulted in a more accurate SZD function, allowing a more credible prediction of the source longevity and the impacts of partial source reduction.

The benefits of using such a methodology include reduced uncertainty, as well as a credible basis for establishing remedial objectives and defining the metrics for source treatment. The costs of the methodology depend on the existing infrastructure. The costs represent a small increment of the remediation costs if a pump-and-treat system is active at a facility and monitoring wells exist within the source area, or if the installation of such a system is anticipated as part of the site remediation. However, it may represent considerable additional cost at sites where the needed infrastructure for the field testing must be installed. This methodology can be applied at sites with LNAPL or DNAPL and can significantly improve the quality of management decisions.