8    MODELING STRATEGY FOR PREDICTING
BIOREMEDIATION RATES FOR OIL CONTAMINATED
BEACHES AT PRINCE WILLIAM SOUND, ALASKA
Alan F. Rozich, Program Manager
ERM, Inc.
Exton, Pennsylvania  19341
Anthony F. Gaudy, Jr.
Professor Emeritus
University of Delaware
Newark, Delaware   19711
INTRODUCTION
i
The purpose of this paper is to provide the results of work that was performed in connection with
the Environmental Protection Agency's (EPA) Oil Spill Bioremediation Demonstration Efforts that
were conducted at the Prince William Sound area near Valdez, Alaska. The bioremediation activities
were conducted in connection with the clean-up of beaches which were contaminated with crude oil
from Prudhoe Bay as a result of the 24 March 1989 oil spill from the Exxon Valdez. A complimentary
effort consisted of developing a structured protocol and predictive methodology to predict bioremediation rates. The goal of this effort is to use the predictive methodology to determine the time needed
to remediate oil contamination at the beaches using bioremediation. A corollary objective involved
predicting the impact on clean-up time realized by various amendments or treatments for beach
material (e.g., the use of fertilizers or other products that are commercially available). These additives
are commonly employed in bioremediation applications for enhancing degradation rates.
The overall strategy is to formulate a predictive algorithm that is calibrated using data collected
using bench-scale laboratory microcosm systems. The primary indicators of microbial activity in these
systems are oxygen uptake or carbon dioxide evolution, i.e., respirometric data. The reason for
employing this strategy is that it is more economical and less time consuming to evaluate the effectiveness of various treatments for enhancing bioremediation rates. The basic thrust of the effort described
in this paper involved developing a predictive model for bioremediation which was calibrated using 02
uptake or C02 evolution data.
The effort involving the development of methodology for predicting bioremediation rates at the oil
contaminated beaches involved the following:
• Devising a model utilizing accepted procedures for biological systems that can predict
clean-up times.
• Identifying the key modeling parameters.
• Analyzing existing laboratory and field data using the modeling approach, i.e., attempt
to use the laboratory data to predict field results.
• Determining what would be needed to improve the model.
• Suggesting additional field and laboratory efforts for improving predictive capability
and streamlining batch data collection and analysis.
BACKGROUND
The physicalilties of the working environment on the oil contaminated beaches at Prince William
Sound warrant discussion. The main problem with the spill is that the oil travelled to the section of the
Sound which has hundreds of small islands. The oil contamination was thus distributed over a large
surface area which was spread among many discrete islands. Additionally, getting to the islands is very
difficult. The remote location of the islands means that the only access is by boat or helicopter. This
feature makes implementing any clean-up effort or performing any field tests logistically difficult.
46th Purdue Industrial Waste Conference Proceedings, 1992 Lewis Publishers, Inc., Chelsea, Michigan 48118.
Printed in U.S.A.
65