טכניון מכון טכנולוגי לישראל
הטכניון מכון טכנולוגי לישראל - בית הספר ללימודי מוסמכים  
M.Sc Thesis
M.Sc StudentBen- Asher Lea
SubjectOptimization of in Sitn Bioremediation of a Contaminated
Aquifer
DepartmentDepartment of Civil and Environmental Engineering
Supervisor Professor Emeritus Uri Shamir


Abstract

Groundwater pollution is an ever-increasing problem. In recent years, bioremediation methods for treating contaminants have been proposed, which treat the contaminant in-situ. The objective is to reduce the toxicity of the pollutant and change it into a “harmless” material.


The objective of this thesis is to present a management model that finds the optimal layout of wells that will enable bioremediation of the hydrocarbons by bacteria located in the ground (endogenous bacteria) and reduce the maximum concentration of the pollutant to below the allowed level at the end of the project period, at minimal cost.


The case study analyzed is based on a confined aquifer polluted with a hydrocarbon pollutant. Ten wells were located around the hydrocarbon plume. Each well can be either operational or idle. An operational well is capable of either pumping or injecting oxygenated water.


The problem, as presented by flow equations in the aquifer and by equations that describe the biological process, is non-convex (Shoemaker 1997).


The simulation-optimization management model in this work is based on the Genetic Algorithm method as an optimization model and Artificial Neural Networks (ANN) as an approximate simulation model that replace the full simulation model Bioplume II. Bioplume II is a model that simulates in-situ bioremediation of hydrocarbons influenced by oxygen limited biodegradation.


From the study we conclude:

·                    The ANN model can provide a good approximation for simulation in-situ bioremediation. Its results are “close enough” to the results from the full simulation model Bioplume II.

·                    The use of ANN, as an alternative method, makes the management model (which combines optimization and simulation models) quicker and more user friendly.

·                    The Genetic Algorithm is a suitable method for solving management problems of bioremediation.