Masters Theses

Ting Chow Tan

Date of Award

8-1994

Degree Type

Thesis

Degree Name

Master of Science

Major

Chemical Engineering

Major Professor

Tse-Wei Wang

Committee Members

Robert S. Kerr, Duane A. Graves, Robert T. Ladd

Abstract

When considering choosing natural bioremediation as the means for cleaning up a contaminated site, the decision makers would require strongly supported data that would allow them to be convinced that natural degradation is indeed taking place, at a fast enough pace to eventually returning the site to a level suitable for reuse in a reasonable amount of time. How does one go about analyzing the site monitoring data in order to provide such data is the subject of this thesis.o-xylene, m-xylene, and p-xylene) at 21 wells scattered over three sites, along with geochemical information, such as pH, temperature, alkalinity, phosphates, and methane concentrations.

The objective of analysis is to take all the concentration data in at one time and apply MVS analysis to the whole set and to attempt to arrive at an overall natural degradation rate for all the contaminant compounds taken as a whole. This would be much more superior to the traditional one-component-at-a-time analysis in that all the information would be used in the analysis and any correlation among the various contaminant compounds is taken explicitly into consideration by the MVS approach and redundant information is cast away before final analysis is done. As a result, not only can one overall degradation rate constant be calculated that characterize the degradation kinetics of the entire set of contaminant compounds, but the segregation, if any, of the contaminants into distinct classes with different degradation kinetics between classes, can also be detected.

In analyzing the data, the residence time for the BTEX compounds that corresponds to the compounds sampled at each of the sampling points is first calculated from the hydraulic gradient information. Because an inert compound was not included in the monitoring effort until the last two sampling points, the concentration of benzene was used as the reference for correcting for dilution effect of all other compounds, for ben-zene was found in Wilson's study to have undergone the least amount of bioattenuation over the two years of monitoring time. In preparation for applying the MVS analysis approach, the concentration data is first column mean-centered (i.e. the column mean becomes zero), followed by scaling column wise, by dividing each column by the variance of the corresponding column at the initial reference time point, such that any subse-quent decrease in the concentration of the BTEX compounds can be discerned. Proper Choice of the scaling factor is one of the most important, yet at the same time, the most subjective aspect of applying the MVS analysis approach.

MVS analysis performed on the Sleeping Bear Dunes data showed that indeed only one similarly behaving class of compounds is present. Assuming first-order degradation kinetics, an overall rate constant was also derived to be 0.0133/day for the site that is 9.1 meters down gradient from the spill site. This rate constant was arrived at by fitting ten points after MVS analysis by an exponential curve, representing that of first-order kinetics. Therefore, the calculated rate constant is relatively robust against errors in sampling in any one or several individual concentration values.

Analysis tools such as these that can model the natural degradation rates of the contaminants at a site, can provide valuable information to decision makers as to the feasibility of relying on natural remediation to render the contaminated site suitable for reuse, in a reasonable amount of time. However, it is crucial that suitable monitoring data be taken so that applicable analysis can be done to yield useful information. Among the monitored data, it is imperative that detailed hydraulic gradient information, as well as the concentration of an inert compound that can be used as the reference compound to correct for dilution effect, be available throughout the study period.

Download
Files over 3MB may be slow to open. For best results, right-click and select "save as..."

Share

COinS