Surrogate model application to the identification of an optimal surfactant-enhanced aquifer remediation strategy for DNAPL-contaminated sites

Abstract

A surrogate model is introduced for identifying the optimal remediation strategy for Dense Non-Aqueous Phase Liquids (DNAPL)-contaminated aquifers. A Latin hypercube sampling (LHS) method was used to collect data in the feasible region for input variables. A surrogate model of the multi-phase flow simulation model was developed using a radial basis function artificial neural network (RBFANN). The developed model was applied to a perchloroethylene (PCE)-contaminated aquifer remediation optimization problem. The relative errors of the average PCE removal rates between the surrogate model and simulation model for 10 validation samples were lower than 5%, which is high approximation accuracy. A comparison of the surrogate-based simulation optimization model and a conventional simulation optimization model indicated that RBFANN surrogate model developed in this paper considerably reduced the computational burden of simulation optimization processes.

Keywords

DNAPL / Latin hypercube sampling / radial basis function artificial neural network / simulation optimization / surrogate model

References

[1] Abriola L M. Modeling Multiphase Migration of Organic Chemicals in Groundwater Systems—A Review and Assessment. Environ. Health Perspect., 1989, 83: 117-143.

[2] Ahlfeld D P, Mulvey J M, Pinder G F. Contaminated Groundwater Remediation Design Using Simulation, Optimization, and Sensitivity Theory: 2. Analysis of a Field Site. Water Resour. Res., 1988, 24(3): 443-452.

[3] Baddari K, Aïfa T, Djarfour N, . Application of a Radial Basis Function Artificial Neural Network to Seismic Data Inversion. Computat. Geosci., 2009, 35(12): 2338-2344.

[4] Bear J. Hydraulics of Groundwater, 2007 New York: Dover Publications, 67.

[5] Carnicer J M. Interpolation and Reconstruction of Curves and Surfaces. Rev. Real Academia de Ciencias. Zaragoza., 2008, 63: 7-40.

[6] Chatterjee K, Fang K T, Qin H. A Lower Bound for the Centered L 2-Discrepancy on Asymmetric Factorials and Its Application. Metrika, 2006, 63(2): 243-255.

[7] Chen S, Cowan C F N, Grant P M. Orthogonal Least Squares Learning Algorithm for Radial Basis Function Networks. Proceedings of IEEE Transactions on Neural Networks, 1991, 2: 302-309.

[8] Ciocoiu I B. RBF Networks Training Using a Dual Extended Kalman Filter. Neurocomputing, 2002, 48(1–4): 609-622.

[9] Delshad M, Pope G A, Sepehrnoori K. A Compositional Simulator for Modeling Surfactant Enhanced Aquifer Remediation, 1 Formulation. J. Contam. Hydrol., 1996, 23(4): 303-327.

[10] Fen C S, Chan C, Cheng H C. Assessing a Response Surface-Based Optimization Approach for Soil Vapor Extraction System Design. Journal of Water Resources Planning and Management, 2009, 135(3): 198-207.

[11] Fernandez-Garcia D, Bolster D, Sanchez-Vila X, . A Bayesian Approach to Integrate Temporal Data into Probabilistic Risk Analysis of Monitored NAPL Remediation. Advances in Water Resources, 2012, 36: 108-120.

[12] Fetter C W. Contaminant Hydrogeology, 1999 New York: Macmillan Publishing Company, 208-262.

[13] Guan J, Aral M. Optimal Remediation with Well Locations and Pumping Rates Selected as Continuous Decision Variables. J. Hydrol., 1999, 221(1–2): 20-42.

[14] He L, Huang G H, Zeng G M, . An Integrated Simulation, Inference, and Optimization Method for Identifying Groundwater Remediation Strategies at Petroleum-Contaminated Aquifers in Western Canada. Water Res., 2008, 42(10–11): 2629-2639.

[15] Helton J C, Davis F J. Latin Hypercube Sampling and the Propagation of Uncertainty in Analyses of Complex Systems. Reliab. Eng. Syst. Saf., 2003, 81(1): 23-69.

[16] Helton J C, Davis F J, Johnson J D. A Comparison of Uncertainty and Sensitivity Analysis Results Obtained with Random and Latin Hypercube Sampling. Reliab. Eng. Syst. Saf., 2005, 89(3): 305-330.

[17] Hora S C, Helton J C. A Distribution-Free Test for the Relationship between Model Input and Output when Using Latin Hypercube Sampling. Reliab. Eng. Syst. Saf., 2003, 79(3): 333-339.

[18] Huang Y, Li J, Huang G, . Integrated Simulation-Optimization Approach for Real-Time Dynamic Modeling and Process Control of Surfactant-Enhanced Remediation at Petroleum-Contaminated Sites. Pract. Period Hazard Toxic Radioact. Waste Manag. (ASCE), 2003, 7(2): 95-105.

[19] Johnson V M, Rogers L L. Accuracy of Neural Network Approximators in Simulation-Optimization. Journal of Water Resources Planning and Management, 2000, 126(2): 48-65.

[20] Kegl B, Krzyak A, Niemann H. Radial Basis Function Networks and Complexity Regularization in Function Learning and Classification. Proceedings of the 5th International Conference on Pattern Recognition. IEEE, 2000, 2: 81-86.

[21] Kuiper L K, Illangasekare T K. Numerical Simulation of NAPL Flow in the Subsurface. Computat. Geosci., 1998, 2(3): 171-189.

[22] Liu L. Modeling for Surfactant-Enhanced Groundwater Remediation Processes at DNAPLs-Contaminated Sites. J. Environ. Inform., 2005, 5(2): 42-52.

[23] Liu W H, Medina M A Jr., Thomann W, . Optimization of Intermittent Pumping Schedules for Aquifer Remediation Using a Genetic Algorithm 1. J. Am. Leather Chem. As., 2000, 36(6): 1335-1348.

[24] McKay M D, Beckman R J, Conover W. A Comparison of Three Methods for Selecting Values of Input Variables in the Analysis of Output from a Computer Code. Technometrics, 1979, 21(2): 239-245.

[25] Moradkhani H, Hsu K, Gupta H V, . Improved Streamflow Forecasting Using Self-Organizing Radial Basis Function Artificial Neural Networks. J. Hydrol., 2004, 295(1–4): 246-262.

[26] NRC Alternatives for Groundwater Clean up, 1994 Washington DC: National Academy Press, 1-316.

[27] Olsson A, Sandberg G, Dahlblom O. On Latin Hypercube Sampling for Structural Reliability Analysis. Struct. Saf., 2003, 25(1): 47-68.

[28] Pan W. The Research and Application of the Online Algorithms: [Dissertation], 2003 Changchun: Jilin University, 1-61.

[29] Pennell K D, Jin M, Abriola L M, . Surfactant Enhanced Remediation of Soil Columns Contaminated by Residual Tetrachloroethylene. J. Contam. Hydrol., 1994, 16(1): 35-53.

[30] Petelet M, Iooss B, Asserin O, . Latin Hypercube Sampling with Inequality Constraints. Asta. Adv. Stat. Anal., 2010, 94(4): 325-339.

[31] Powell M J D. Radial Basis Functions for Multivariable Interpolation: A Review. Algorithms for Approximation, 1987, 143-167.

[32] Qin X S, Huang G H, Chakma A, . Simulation-Based Process Optimization for Surfactant-Enhanced Aquifer Remediation at Heterogeneous DNAPL-Contaminated Sites. Sci. Total Environ., 2007, 381(1–3): 17-37.

[33] Rathfelder K M, Abriola L M, Taylor T P, . Surfactant Enhanced Recovery of Tetrachloroethylene from a Porous Medium Containing Low Permeability Lenses. 2. Numerical Simulation. J. Contam. Hydrol., 2001, 48(3–4): 351-374.

[34] Rogers L L, Dowla F U, Johnson V M. Optimal Field-Scale Groundwater Remediation Using Neural Networks and the Genetic Algorithm. Environ. Sci. Techno., 1995, 29(5): 1145-1155.

[35] Schaerlaekens J, Mertens J, Van Linden J, . A Multi-Objective Optimization Framework for Surfactant-Enhanced Remediation of DNAPL Contaminations. J. Contam. Hydrol., 2006, 86(3–4): 176-194.

[36] Schumaker M F, Kramer D M. Comparison of Monte Carlo Simulations of Cytochrome B6f with Experiment Using Latin Hypercube Sampling. Bull. Math. Biol., 2011, 73(9): 2152-2174.

[37] Shen W, Guo X, Wu C, . Forecasting Stock Indices Using Radial Basis Function Neural Networks Optimized by Artificial Fish Swarm Algorithm. Knowl-Based Syst., 2010, 3(24): 378-385.

[38] Sreekanth J, Datta B. Multi-Objective Management of Saltwater Intrusion in Coastal Aquifers Using Genetic Programming and Modular Neural Network Based Surrogate Models. Journal of Hydrology, 2010, 393(3–4): 245-256.

[39] Van Camp M, Walraevens K. Pumping Test Interpretation by Combination of Latin Hypercube Parameter Sampling and Analytical Models. Computat. Geosci., 2009, 35(10): 2065-2073.

[40] Yan S, Minsker B. Optimal Groundwater Remediation Design Using an Adaptive Neural Network Genetic Algorithm. Water Resour. Res., 2006, 42(5): 1145-1155.

[41] Yan S, Minsker B. Applying Dynamic Surrogate Models in Noisy Genetic Algorithms to Optimize Groundwater Remediation Designs. Journal of Water Resources Planning and Management, 2011, 137: 284-292.

[42] Zhong L R, Mayer A S, Pope G A. The Effects of Surfactant Formulation on Nonequilibrium NAPL Solubilization. J. Contam. Hydrol., 2003, 60(1–2): 55-75.