Architecture, persistence and dissolution of a 20 to 45 year old trichloroethene DNAPL source zone

Michael O Rivett,Rachel A Dearden,Gary P Wealthall

doi:10.1016/j.jconhyd.2014.09.008

Abstract

A detailed field-scale investigation of processes controlling the architecture, persistence and dissolution of a 20 to 45year old trichloroethene (TCE) dense non-aqueous phase liquid (DNAPL) source zone located within a heterogeneous sand/gravel aquifer at a UK industrial site is presented. The source zone was partially enclosed by a 3-sided cell that allowed detailed longitudinal/fence transect monitoring along/across a controlled streamtube of flow induced by an extraction well positioned at the cell closed end. Integrated analysis of high-resolution DNAPL saturation (Sn) (from cores), dissolved-phase plume concentration (from multilevel samplers), tracer test and permeability datasets was undertaken. DNAPL architecture was determined from soil concentration data using partitioning calculations. DNAPL threshold soil concentrations and low Sn values calculated were sensitive to sorption assumptions. An outcome of this was the uncertainty in demarcation of secondary source zone diffused and sorbed mass that is distinct from trace amounts of low Sn DNAPL mass. The majority of source mass occurred within discrete lenses or pools of DNAPL associated with low permeability geological units. High residual saturation (Sn>10–20%) and pools (Sn>20%) together accounted for almost 40% of the DNAPL mass, but only 3% of the sampled source volume. High-saturation DNAPL lenses/pools were supported by lower permeability layers, but with DNAPL still primarily present within slightly more permeable overlying units. These lenses/pools exhibited approximately linearly declining Sn profiles with increasing elevation ascribed to preferential dissolution of the uppermost DNAPL. Bi-component partitioning calculations on soil samples confirmed that the dechlorination product cDCE (cis-dichloroethene) was accumulating in the TCE DNAPL. Estimated cDCE mole fractions in the DNAPL increased towards the DNAPL interface with the uppermost mole fraction of 0.04 comparable to literature laboratory data. DNAPL dissolution yielded heterogeneous dissolved-phase plumes of TCE and its dechlorination products that exhibited orders of magnitude local concentration variation. TCE solubility concentrations were relatively localised, but coincident with high saturation DNAPL lens source areas. Biotic dechlorination in the source zone area, however, caused cDCE to be the dominant dissolved-phase plume. The conservative tracer test usefully confirmed the continuity of a permeable gravel unit at depth through the source zone. Although this unit offered significant opportunity for DNAPL bypassing and decreased timeframes for dechlorination, it still transmitted a significant proportion of the contaminant flux. This was attributed to dissolution of DNAPL–mudstone aquitard associated sources at the base of the continuous gravel as well as contaminated groundwater from surrounding less permeable sand and gravel horizons draining into this permeable conduit. The cell extraction well provided an integrated metric of source zone dissolution yielding a mean concentration of around 45% TCE solubility (taking into account dechlorination) that was equivalent to a DNAPL mass removal rate of 0.4tonnes per annum over a 16m2 cell cross sectional area of flow. This is a significant flux considering the source age and observed occurrence of much of the source mass within discrete lenses/pools. We advocate the need for further detailed field-scale studies on old DNAPL source zones that better resolve persistent pool/lens features and are of prolonged duration to assess the ageing of source zones. Such studies would further underpin the application of more surgical remediation technologies.

Highlights

Industrial use of trichloroethene (TCE) dates back to the early 1930s, some half a century prior to its recognition as a groundwater contaminant of major concern (Pankow et al, 1996; Rivett et al, 2006)
Pooled dense non-aqueous phase liquid (DNAPL), at c. 1% occurrence, accounted for 14% of the mass for a porosity of 0.28, but increased to 23% of the mass for the lower porosity of 0.2
Due to the combination of these processes occurring, uncertainty in sorption coefficients and DNAPL thresholds (Section 3.3) and the moderate resolution of the thin lower K intermediate layers within the cell as well as coarser resolution of the aqueous phase, we provide a simple exponential fit rather than attempt a diffusion-based simulation of these soil concentration profiles

Summary

Introduction

Industrial use of trichloroethene (TCE) dates back to the early 1930s, some half a century prior to its recognition as a groundwater contaminant of major concern (Pankow et al, 1996; Rivett et al, 2006). Chlorinated solvents such as TCE may penetrate as a dense non-aqueous phase liquid (DNAPL) far below the water table to subsequently dissolve over decadal timeframes. They pose long-term threats to groundwater quality (Rivett et al, 2012; Rowe et al, 2007). It is important to discern the field-scale nature of such old source zones to underpin predictions of their continuing longevity and remedial decision making

Objectives

Results

Conclusion