Superfund Research Program

Supercritical Fluid Technology for Remediation of PCB/PAH Contaminated Soils and Sediments

Project Leader: Lawrence L. Tavlarides (Syracuse University)
Grant Number: P42ES004913
Funding Period: 1995 - 2000

Project-Specific Links

Final Progress Reports

Year:   1999 

The goal of this project is to develop a supercritical fluid (SCF) technology to remediate soils contaminated with polychlorinated biphenyls (PCBs) and polycyclic aromatic hydrocarbons (PAHs). A two step process is advanced consisting of extraction of the PCBs/PAHs from the contaminated soil using supercritical (SC) CO2-methanol fluids and destruction of the extracted concentrates using supercritical water oxidation (SCWO). Researchers have demonstrated the technical feasibility and the economic competitiveness of both steps of the process. Also, important design information has been developed for large scale process design.

Results include:
(1) Sorption equilibrium studies have been completed using St. Lawrence River sediment (SLRS) contaminated with PCBs with SC CO2-5 mol% methanol. Models for sorption isotherms have been developed for these data and those for SC CO2 obtained last year. Linear isotherms were observed for the total PCBs and 16 major PCB congeners under all investigated conditions. Preferential desorption of lower chlorinated congeners was also found.

(2) Bench scale desorption experiments (~2 kg sample) conducted confirm the 99.8% (less than 5 ppm of residual PCB concentration) removal rate of PCBs from SLRS at 116 atm, 50 EC and 30-60 min processing time. The CO2 usage per kg SLRS was shown to be less than laboratory values suggesting lower processing costs than $198-318/m3 soil processed calculated last year. The desorption model describes the bench scale studies and confirms model utility for process design calculations.

(3) A linear driving force desorption model was developed, using the above sorption equilibrium isotherms. A two step calculation is employed with the model to describe both laboratory and bench scale PCB desorption data. The simulation describes both the fast desorption (short time) and slow desorption very well, giving the overall average absolute relative derivation of less than 10%.

(4) A congener specific global kinetic study of Aroclor 1248 in supercritical water resulted in the development of a second order rate law which predicts the oxidative destruction of PCBs at 250 atm and 723 to 823EK. The 99.95% high level of destruction at 250 atm and 823EK for 54 s contact time is modeled successfully.

(5) Kinetic data for the overall reaction pathway for 3,3=,4,4=-tetrachlorobiphenyl at 250 atm and 673 to 773 K have been successfully modeled using second order kinetics and suggest a successive dechlorination leading to oxidation.