| Summary: | 碩士 === 國立中山大學 === 環境工程研究所 === 87 === In this work, a study on simulating in-situ remediation of diesel-contaminated soils using steam injection / vacuum extraction was carried out by a bench-scale soil column. Experimental designs the L9(34) and L4(23) orthogonal arrays were used. Experimental results were analyzed statistically to determine the relative contribution of each experimental factor on treatment efficiency. The optimal operating conditions were further determined. Four experimental factors studied in the L9(34) orthogonal arrays are : soil water content (0 wt %, 5 wt %, and 10 wt %), steam injection pressure (4 Kg/cm2, 6 Kg/cm2, and 8 Kg/cm2), initial contaminant concentration (4 wt %, 12 wt %, and 20 wt %) and soil type (sandy loam and silty loam). Three experimental factors studied in the L4-1(23) orthogonal arrays are : soil water content (5 wt % and 10 wt %), steam injection pressure (4 Kg/cm2 and 6 Kg/cm2) and with vacuum extraction or not. On the other hand, three experimental factors studied in the L4-2 orthogonal arrays are : soil water content (5 wt % and 10 wt %), initial contaminant concentration （0.5 wt % and 20 wt %）, and soil column length (20 cm and 60 cm).
For the case of the L9 orthogonal arrays, the results of the analysis of variance (ANOVA) have shown that the initial contaminant concentration yielded the greatest contribution (i.e., 78.01%) to the treatment efficiency. The contributions of the soil type, soil water content, and steam injection pressure were determined to be 12.86%, 5.95%, and 3.18%, respectively. Accordingly, regardless of the initial contaminant concentration, a steam injection pressure of 6 Kg/cm2 and a soil water content of 5 wt% were determined to be the optimal operating conditions for both the sandy loam and silty loam tested. Under the optimal operating conditions, the treatment efficiencies in the ranges of 56.5-100% and 27.4-92.0% were found for the sandy loam and silty loam, respectively. Furthermore, the results of verification experiments were in good agreement with the theoretically calculated treatment efficiencies in all cases.
For the case of L4-1, the results of ANOVA have indicated that having a vacuum extraction unit in the treatment system would yield the greatest contribution (i.e., 55.04%) to the treatment efficiency. In this regard, the contributions of the steam injection pressure and soil water content were determined to be 32.43% and 12.53%, respectively.
Similarly, for the case of L4-2, the initial contaminant concentration gave rise to the greatest contribution of 81.14%, followed by 12.01% due to the soil water content and 6.85% due to the length of soil column.
To determine the effect of the temperature loss from the soil column on the treatment efficiency, two additional experiments were carried out in this regard. Experimental results have shown that having a thermal insulation material surround the soil column would increase the treatment efficiency by 10-20%.
|
|---|
