Barriers fashioned from artificially frozen ground may be useful for isolating heavy-metal contaminated soils. We tested the likely effectiveness of a frozen ground barrier in containing heavy-metal laden liquid during remediation of a contaminated soil. To simulate the in-situ clean up of the contaminated soil we used a 0.1 M EDTA solution to chelate the metals and remove them from the soil. A local uncontaminated sandy-silt soil was used in both the frozen barrier and the contaminated section of our experimental chambers. A subsample of this soil was artificially contaminated by spiking with a mixed metal solution containing 1340 mg/L (21.08 mM) Cu, 222 mg/L (1.97 mM) Cd, 778 mg/L (13.25 mM) Ni, and 2513 mg/L (38.44 mM) Zn. This spiking solution yielded a soil with extractable metal contents considered to be above the "natural" range of these elements found in most soils. Soil used in the barrier was saturated with double-deionized water. Both the spiked soil and the barrier soil were sealed while the solution and soil equilibrated. We carried out experiments in four chambers: three replicates and a control. The clean barrier soil was compacted into the bottom half of a cylindrical column 15.2 cm high by 7.6 cm diameter to a wet bulk density of 1.9 g/cm3. A copper coil connected to a recirculating cooling bath was fitted around the bottom half of the chamber and was used to maintain the frozen barrier. The temperature of the barrier soil was held constant at -3 + or - 0.1 oC for a few weeks, until frozen completely, before the upper half of each column was filled with approximately 565 g of spiked soil. We introduced the EDTA solution at the top surface of the soil and allowed it to percolate down through the chamber. A hand pump was used to remove liquid from about 7 cm deep in the soil (at the boundary between the frozen and unfrozen soil). Chelated solution was removed and fresh EDTA added daily. Metal concentrations in the collected solutions were measured using atomic absorption spectroscopy. After 30 days, approximately 80% of the contaminants can be accounted for in the removed EDTA solutions. Cd was removed most efficiently, Ni least efficiently, by this method. Although the bottom halves of the chambers remained at -3 oC throughout the experiment, we believe that the barrier was sometimes incompletely frozen. We periodically collected small amounts of liquid that accumulated in tubing below the frozen barrier. When analyzed, this liquid contained measurable concentrations of Cu, Cd, Ni, and Zn. We believe that ground freezing can be an effective means of containing contaminated soils and liquids during metal removal, but the temperature of the barrier must be maintained below -3 oC.
Ginger E. Boitnott USACRREL 72 Lyme Rd. Hanover, NH 03755 Phone: 603 646 4512 Fax: 603 646 4561 e-mail: boitnott@crrel.usace.army.mil