Laboratory and controlled field experiments using potassium permanganate to remediate trichloroethylene and perchloroethylene DNAPLs in porous media

Few proven technologies exist that may be used to treat dense non-aqueous phase liquid (DNAPL) contaminants. In-situ chemical flushing is a proposed technology which consists of flushing DNAPL source zones with a reactive solution to degrade the contaminant mass below ground. A laboratory and controlled field experimental program was conducted to assess the potential of potassium permanganate (KMnO4) as a reagent for in-situ DNAPL remediation. The results of laboratory experiments indicated that two common DNAPL contaminants, perchloroethylene (PCE) and trichloroethylene (TCE), were rapidly degraded to chloride and carbon dioxide. Column experiments, using residual PCE flushed with oxidant concentrations as high as 10 g L-1, indicated that chloride could be used as a reaction tracer. From the chloride data, it appeared that the rate of PCE removal from the columns was a complex process dependent upon the kinetics of both dissolution and oxidation. Two experimental applications of in-situ oxidation were conducted in the Borden aquifer isolated within a 7.5 m(3) double sheet-pile cell. The cell was fitted with injection and recovery wells through which aqueous solutions of KMnO4 were flushed to oxidize solvent source zones in situ. In the initial experiment, flushing of a 1 L PCE residual source with 10 g L-1 KMnO4 at total flow rates of up to 100 L per day, completely removed the source within 120 days. A second experiment, using an 8 L mixture of PCE and TCE slowly allowed to infiltrate into the cell, was conducted using a system to recycle the oxidant. The oxidant was added at 10 g L-1 with a flow of approximately 50 L per day. After 290 days of flushing, it was concluded from the monitoring data that 62% of the initial source (as equivalent chloride mass) had been oxidized and it was evident that oxidation was continuing in the upper third of the cell. These experiments have suggested that the effectiveness of in-situ chemical oxidation will depend primarily upon the distribution of the DNAPL in the subsurface and its effects upon dissolution. In both experiments, spatial variability of chloride measurements appeared to reflect both the DNAPL location and distribution. (C) 1998 Elsevier Science B.V.