You are here

Optimization of Natural Attenuation with Active Remediation under Uncertainty Using a Multi Objective Genetic Algorithm

Title: Optimization of Natural Attenuation with Active Remediation under Uncertainty Using a Multi Objective Genetic Algorithm.
Name(s): Iyer, Satyajeet K., author
Hilton, Amy Chan, professor directing thesis
Dzurik, Andrew, committee member
Leszczynska, Danuta, committee member
Simpson, James, committee member
Department of Civil and Environmental Engineering, degree granting department
Florida State University, degree granting institution
Type of Resource: text
Genre: Text
Issuance: monographic
Date Issued: 2004
Publisher: Florida State University
Place of Publication: Tallahassee, Florida
Physical Form: computer
online resource
Extent: 1 online resource
Language(s): English
Abstract/Description: Natural attenuation (NA) has recently emerged as a viable groundwater remediation technology at various petroleum contaminated sites in the United States. NA is a passive remedial approach that depends upon natural processes to degrade and dissipate petroleum constituents in soil and groundwater. Such natural processes include advection, sorption, diffusion, dispersion and biodegradation. Due to heterogeneous nature of most contaminated groundwater (GW) sites there exists uncertainty in subsurface system parameters. This study evaluates sensitivities of parameter uncertainty on the performance and design of remediation plans that use natural attenuation with active remediation. This analysis is completed by using an optimization tool combined a GW flow and contaminant transport simulation model. The Enhanced multi-objective Robust Genetic Algorithm (EMRGA) is the optimization tool used here for the simultaneous optimization of multiple conflicting objectives under parameter uncertainty. The multi-objective optimization problem is to minimize the cost of the natural attenuation-active remediation system and minimize the maximum contaminant concentration at the end of the five-year remediation period under parameter uncertainty and heterogeneity. The optimization model is applied to a problem based on a field site, contaminated with benzene located in Eglin Air Force Base, Florida. The uncertain parameters considered in this study are hydraulic conductivity (K), hydraulic gradient (dH/dx) and first-order benzene decay rate (k) benzene degradation. The optimization problem is solved using fifteen cases with different combinations of uncertain parameters and degrees of uncertainty. In addition, selected designs from the evolved Pareto-optimal sets (trade-off curves) were further evaluated by Monte Carlo analysis. Results show that as uncertainty in hydraulic conductivity increased there was increased difficulty in lowering contamination levels as fewer wells were used at lower pumping rates. For uncertain parameters hydraulic gradient and decay rate the highly uncertain scenarios produced designs employing more wells at higher pumping rates, thus achieving minimum concentration values. Cases with less uncertainty in hydraulic conductivity produced high performing remedial designs with higher remediation reliability and higher clean up levels. On the other hand, the designs evolved by the EMRGA had lower reliabilities and lower clean up levels for cases having low variations of hydraulic gradient and first-order decay rate. Also, active remediation in the initial stages of the total remediation period emerged as a feasible and most cost-effective solution for the multi-objective optimization problem. Overall, uncertainty in hydraulic conductivity had the most significant impact on remediation reliability of the designs. Results indicate a threshold pumping index value of 430 m3/day/well over which a remediation design was almost certain in achieving 100% reliability. Effects of multiple parameter uncertainty were highly pronounced for cases involving a wider range of hydraulic conductivity values. For these cases the remediation costs dropped to 5.7% and 31.2% with increasing range of hydraulic gradient and heterogeneous decay rate respectively while the Cmax values increased by 217% and 307% for increasing range of hydraulic gradient and heterogeneous decay rate respectively. In general the EMRGA successfully identified Pareto-optimal remedial designs having a wide range of objective values, which satisfied both the conflicting objectives focused in this study. Based on these conflicting objectives (Remediation cost and Maximum residual concentration) seven highly reliable remedial options are identified. In general these designs used just two extraction wells (located just down-gradient from the contaminant plume) at pumping indices between 350 to 450 m3/day/well. These seven remediation plans are embedded in a decision tree to aid the remediation designer in getting an overview of possible groundwater remediation design requirements at the OU-1 site.
Identifier: FSU_migr_etd-3842 (IID)
Submitted Note: A Thesis submitted to the Department of Civil and Environmental Engineering in partial fulfillment of the requirements for the degree of Master of Science.
Degree Awarded: Spring Semester, 2004.
Date of Defense: December 3, 2003.
Keywords: Genetic Algorithm, Multi-Objective, Uncertainty, Natural Attenuation
Bibliography Note: Includes bibliographical references.
Advisory Committee: Amy Chan Hilton, Professor Directing Thesis; Andrew Dzurik, Committee Member; Danuta Leszczynska, Committee Member; James Simpson, Committee Member.
Subject(s): Civil engineering
Environmental engineering
Persistent Link to This Record:
Owner Institution: FSU

Choose the citation style.
Iyer, S. K. (2004). Optimization of Natural Attenuation with Active Remediation under Uncertainty Using a Multi Objective Genetic Algorithm. Retrieved from