Progress 07/01/99 to 06/30/05
Outputs
We developed and applied several simulation/optimization (S/O) modeling procedures for improving water management. In previous periods we reported development of a variety of optimization techniques, including hybrid heuristic optimizers that couple genetic algorithms, tabu search, simulated annealing, or artificial neural networks. Some papers have been published on these, and others will be published soon. In the final six months of the project we finalized a procedure for increasing the robustness of an optimal groundwater pumping strategy developed for field application (journal paper has been accepted for publication). The procedure maximizes the likelihood that the implemented strategy will achieve management goals in the field, without degrading the primary objective function value (for example, without increasing the cost of a minimum-cost strategy, or without increasing the total pumping of a minimum-pumping strategy. This differs from normal stochastic
optimization approaches in which increasing strategy field reliability usually requires harming the primary objective function value. We demonstrated the method by remediating Umatilla Army Depot trinitrotoluene (TNT) and Royal Demolition Explosive (RDX) plumes in Oregon. We reported the developed optimization methods and application results at conferences in California, and Alaska.
Impacts
The new techniques we developed should be very beneficial and will significantly increase application of mathematical optimization for water and other fluid management. Applying mathematical optimization methods when developing management strategies usually provides strategies that are at least 20 percent better, in terms of the primary objective function, than those developed using simulation modeling techniques alone. The new procedure finished during the final six months of the project addresses the concern that mathematical optimization drives a solution towards having tight constraints. Mathematically tight constraints are more likely to be violated in the field than loose constraints, if a simulation prediction model differs significantly from the true field physical system. Thus, the newly-developed robustness-enhancing optimization approach overcomes an application weakness of normal stochastic optimization approaches. It also avoids the stochastic optimization
need for parametric probability density functions (these are difficult and costly to develop). During the final six months of project life, we published one journal article on hybrid simulated annealing and genetic algorithm optimization. We wrote several other papers, two of which have been accepted, but are not yet published, and so are not listed in Section 43. Graduate Student Years: Six (6).
Publications
- Aly, A.H. and R. Peralta. 1999. Comparison of a genetic algorithm and mathematical programming to the design of groundwater cleanup systems. Water Resources Research, 35(8):2415-2425.
- Aly, A.H. and R. Peralta. 1999. Optimal design of aquifer clean up systems under uncertainty using a neural network and a genetic algorithm. Water Resources Research, 15(8):2523-2532.
- Peralta, R. C. 2001. Remediation simulation/optimization demonstrations. In Proceedings of MODFLOW and Other Modeling Odysseys. 2001. Eds, Seo, Poeter, Zheng and Poeter, Pub. IGWMC. p. 651-657.
- Peralta, R. C. 2001. Demonstrations in remediation. In Modeling and Management of Emerging Environmental Issues , Eds, Chien, Medina, Pinder, Reible, Sleep and Zheng. DuPont Corporation. p. 157-166
- Peralta, R. C., Ierardi, M. and J. Santillan. 2001. DoD view of optimization. In Modeling and Management of Emerging Environmental Issues , Eds, Chien, Medina, Pinder, Reible, Sleep and Zheng. DuPont Corporation. p. 118-123
- Shieh, H., and R. Peralta. 2005. Optimal in-situ bioremediation design by hybrid genetic algorithm-simulated annealing. ASCE Journal of Water Resources Planning and Management 131(1):67-78.
- Peralta, R. 2005. Optimizing integrated water resources management: data, tools, and examples. In Proc., 3rd Int. Conf. on Irrig. and Drainage-Water District Man. and Governance, US Committee on Irrig. and Drainage. San Diego, CA, 1 Apr 2005. p 591-601.
- Peralta, R. 2005. Software for optimal integrated water resources management. In Proceedings, World Water and Environmental Resources Congress, 2005, ASCE. Presented in Anchorage, Alaska, 17 May 2005. 11p.
- Peralta, R. C. 2001. Simulation/optimization applications and software for optimal ground-water and conjunctive water management In Proceedings of MODFLOW and Other Modeling Odysseys. 2001. Eds, Seo, Poeter, Zheng and Poeter, Pub. IGWMC. p. 691-694.
- Peralta, R. C. 2001. Manejo optimo conjunctivo de aguas bajo distintos regimenes legales e hidrologicos; and Optimal conjunctive water management in different legal and hydrologic regimes. Proceedings, Fourth Inter-American Diagogue on Water Management. Iguacu Falls, Parana, Brazil, 4 Sep 2001. p 100-124.
|
Progress 01/01/04 to 12/31/04
Outputs
Developed alternative optimization problem formulations for Tooele Army Depot, Utah. These formulations address weaknesses of the formulations developed previously for the facility. Developed optimal pumping strategies for these formulations. Developed improved pumping strategies for Cache Valley, Utah, Umatilla Army Depot, Oregon, and Blaine Naval Ammunition Depot, Nebraska. Reported the developed optimization methods and application results at conferences on three continents. Presented workshops teaching others about optimizing groundwater management in North and South America.
Impacts
The optimization techniques usually provide management strategies that are about 20 percent better than those developed using simulation modeling techniques alone. These techniques will be very beneficial and will save millions of dollars as they are applied worldwide for contamination remediation. They will enhance the long term use of groundwater and surface water resources for water supply. Papers are being prepared for submission to journals. Graduate students completed 1 PhD and 1 MS degrees. The MS student is now beginning a PhD.
Publications
- Kalwij, I. M., and R. C. Peralta. 2004. Effect of optimization problem constraints on pump and treat designs for Tooele Army Depot. In Proceedings of EWRI 2004 World Congress. American Society of Civil Engineers.
- Peralta, R. C. and I. M. Kalwij. 2004. Mathematically optimizing water management. In Proceedings of EWRI 2004 Groundwater Symposium. ASCE.
- Peralta, R. C. and S. Wu. 2004. Software for Optimizing International Water Resources Management. In Proceedings of EWRI 2004 World Congress. ASCE.
- Fayad, H. C., and R. C. Peralta. 2004. Multi-objective conjunctive use optimization. In Proceedings of EWRI 2004 World Congress. ASCE.
- Peralta, R. C. 2004. Optimizing management of nonlinear flow and transport in groundwater and surface water systems. In Proceedings of FEM-MODFLOW International Conference, Karlovy Vary, Czechoslovakia, 13-16 Sep, 2004.
- Peralta, R. C. and R. Shulstad. 2004. Optimization modeling for groundwater and conjunctive use water policy development. In Proceedings of FEM-MODFLOW International Conference, Karlovy Vary, Czechoslovakia, 13-16 Sep, 2004.
- Peralta, R. C. 2004. Optimizacion para el normado, planeamiento, diseno y manejo del agua subterranea. In website Proceedings of XIII Congreso Brasileiro de Aguas Subterraneas. Cuiaba, Matto Grosso, Brazil, 23 Oct, 2004.
- Peralta, R. C., Kalwij, I. M. and B. Timani. 2004. Optimizing complex plume pump and treat systems for Blaine Naval Ammunition Depot, Nebraska. In Proceedings of EWRI 2004 World Congress. ASCE.
- Das, R., Peralta, R. C., and B. Timani. 2004. Cache Valley: optimizing sustainable water use and ecosystems while considering water rights. In Proceedings of EWRI 2004 World Congress. ASCE.
|
Progress 01/01/03 to 12/31/03
Outputs
We have refined simulation/optimization (S/O) modeling techniques for optimizing deterministic and stochastic groundwater contamination management and for deterministic conjunctive water management. We have made significant improvements to genetic algorithm (GA) heuristic optimization approach and to a linked artificial neural network(ANN)/GA approach. We have demonstrated its use for groundwater contamination remediation, and for managing intrusion of salt water from a salt lake or ocean. We have improved the GA for application to three additional types of management problems: conjunctive water management (coordinating groundwater and surface water), optimizing groundwater management while constraining stream depletion, and maximizing sustained yield groundwater planning. We have developed methods for maximizing reliability of groundwater remediation strategies, and for increasing robustness while optimizing. We have also demonstrated how to create more valid 3D
spatially distributed contaminant plume representations than by using geostatistical methods alone. We applied that technique to a trichloroethylene plume in San Bernardino, California (but the procedure needs to be automated). We have given presentations on the optimization techniques at a national Department of Defense conference in San Antonio, and at international professional society conferences in Philadelphia and Golden, Colorado.
Impacts
Our optimization techniques have been generally demonstrated to yield improvements of 20% over management strategies developed using simulation modeling techniques alone. If applied to the Kennecott Copper Mine plume, remediation expenses would be reduced by millions of dollars. The long term national benefits of our project will be significant reductions in remediation costs and in residual contamination in aquifers nationwide. Currently, two graduate students work on the project (1 PhD and 1 MS).
Publications
- Peralta, R. C., Kalwij, I. M. and S. *Wu. 2003. Practical simulation/optimization modeling for groundwater quality and quantity management. In Proceedings, MODFLOW and More 2003: Understanding through Modeling, Golden Colorado. International Groundwater Modeling Center, p 784-788.
- Peralta, R. and B. Timani. 2003. Developing initial TCE arrays for predicting concentrations downgradient of Norton AFB. Interim report prepared for EarthTech Corporation. 31 Oct 2003.
- Peralta, R. C. and I. Kalwij. 2003. Simulation/optimization tools for robust pumping strategy design. Published electronically by Air Force Center for Environmental Excellence in Proceedings, AFCEE Technology Transfer Workshop, 23 - 27 February 2003, San Antonio, TX.
- Peralta, R. C. 2003. SOMOS Simulation/Optimization Modeling System. In Proceedings, MODFLOW and More 2003: Understanding through Modeling. International Groundwater Modeling Center, Golden, CO. p 819-823.
|
Progress 01/01/02 to 12/31/02
Outputs
The objective is to develop improved simulation/optimization (S/O) modeling techniques, as opposed to normal simulation (S) modeling techniques, for optimizing groundwater management. We are to do this for representative situations. We applied heuristic optimization (HO) methods for computing optimal water management strategies for groundwater contamination remediation. Included HO techniques are genetic algorithm (GA), tabu search (TS), and simulated annealing (SA). We applied these to hypothetical and complex real-world contaminant plumes. The largest plume is over 4 by 1.5 miles in size, and posed a very computationally intensive and challenging nonlinear optimization problem. To speed processing we sometimes used artificial neural networks (ANN) as simulators to predict groundwater head, contamination concentration, and contaminant mass. We developed optimal pumping strategies and reports for plume management problems at three sites: (1) Umatilla Ammunition Depot,
Oregon; (2) Tooele Army Depot, Utah; and (3) Blaine Naval Ammunition Depot, Nebraska. We are currently fine-tuning groundwater contaminant transport predictions of a solvent plume in San Bernardino, CA.
Impacts
If installed, our optimal design for the Umatilla site will cost over 55% less than the current design--developed by another party using a normal simulation (S) model. Our design would achieve cleanup of royal demolition explosive (RDX) and tri-nitrotoluene (TNT) within 4 years. Our optimal design for the Tooele site will cost about $500,000 less than a design created by a consulting firm using a normal S model. Our optimal design for the Nebraska site would cost about $10,000,000 less than the design by a consulting firm using a normal simulation model. Benefits at the San Bernardino depend on the plume predicted after we refine the model transport prediction capabilities. The long term national benefits of our project will be significant reductions in remediation costs and in residual contamination in aquifers nationwide. Currently, two graduate students work on the project (1 PhD and 1 MS).
Publications
- No publications reported this period
|
Progress 01/01/01 to 12/31/01
Outputs
We have developed improved techniques for optimizing management of complex nonlinear groundwater and stream-aquifer systems. We have tested two techniques at a site in Umatilla, Oregon. There we showed how to minimize the cost of remediating plumes of dissolved TNT and RDX, common explosives. We are applying it to a trichloroethylene plume at Tooele, Utah. We have applied a different optimization technique to maximize sustainable groundwater extraction rates for Cache Valley, Utah, without unacceptably harming the ecosystem or senior water rights.
Impacts
We are significantly reducing the cost of groundwater contamination remediation. If we applied our technology at the Kennecott plume in Salt Lake Valley, we would save the polluters many millions of dollars. The cost reduction at a Umatilla, Oregon site was over 50 percent. We are showing how to safely increase the sustained availability of groundwater.
Publications
- Peralta, R. C. 2001. Remediation simulation/optimization demonstrations. In Proceedings of MODFLOW and Other Modeling Odysseys. 2001. Eds, Seo, Poeter, Zheng and Poeter, Pub. IGWMC. p. 651-657.
- Peralta, R. C. 2001. Simulation/optimization applications and software for optimal ground-water and conjunctive water management In Proceedings of MODFLOW and Other Modeling Odysseys. 2001. Eds, Seo, Poeter, Zheng and Poeter, Pub. IGWMC. p. 691-694.
- Peralta, R. C. 2001. Optimal conjunctive water management in different legal and hydrologic regimes. Proceedings, Fourth Inter-American Dialogue on Water Management. Invited paper and presentation, at Iguacu Falls, Parana, Brazil, 4 Sep 2001. 8 p.
- Peralta, R. C. 2001. El proyecto Chavi-Mochic en el Peru. Presentation at Fourth Inter-American Diagogue on Water Management. Iguacu Falls, Parana, Brazil, 4 Sep 2001.
- Peralta, R. C., Kalwij, I. and S. Wu. 2001. Flow and transport optimization end-of-simulation results: Umatilla Chemical Depot. Interim project completion report and Errata submitted to U. S. Navy. July 2001. 5 p.
|
Progress 01/01/00 to 12/31/00
Outputs
We have emphasized developing algorithms for computing optimal water management strategies for groundwater contamination remediation. I designed a representative problem consisting of two aquifer layers with a different contaminant and distribution in each layer. Unless restrained by a pump and treat system, groundwater and contamination move to the north toward public supply wells and a river. The management goal is to determine how best to prevent the contamination from reaching the supply wells or river, and to remediate the plume within one year. We are to determine those combinations of additional extraction and injection well locations that satisfy those containment and clean-up goals while: minimizing cost or minimizing residual contaminant mass. This problem is representative of many found in the U.S. This is a computationally intensive and challenging nonlinear optimization problem. To speed processing we use artificial neural networks (ANN) as simulators to
predict groundwater head, contamination concentration, and contaminant mass. We use evolutionary optimization (EO) techniques to perform optimization, Example EO techniques include genetic algorithm (GA) and simulated annealing (SA). We have significantly improved the ANN, which now usually predicts within about 1-2 percent of values predicted by detailed finite difference model. We have significantly improved the GA used in screening optimization. The intent of the screening optimization step is to determine which pumping wells, from among a group of candidate wells, are most likely to yield the best pumping strategy for user-specified management goal and constraints. We have developed a C version of a SA algorithm we developed previously. We are currently performing comparisons of the SA and GA algorithm for screening optimization. We completed an optimal pumping strategy for a plume of solvent-contaminated groundwater on Cape Cod, Massachusetts. Our recommendations concerning well
placement have been implemented, i.e. the wells and pipelines were constructed and the system should soon be operating. We are currently calibrating a groundwater flow model for part of San Bernardino, CA underlain by a zone of solvent-contaminated groundwater. Two graduate students work on the project (1 PhD and 1 MS).
Impacts
The long term benefits of our project will be significant reductions in remediation costs and in residual contamination in aquifers nationwide. Ultimately, the more distribution our algorithms have, the greater the national benefit.
Publications
- Aly, A.H. and R.C. Peralta. 1999. Comparison of a genetic algorithm and mathematical programming to the design of groundwater cleanup systems. Water Resources Research, 35(8):2415-2425.
- Aly, A.H. and R.C. Peralta. 1999. Optimal design of aquifer clean up systems under uncertainty using a neural network and a genetic algorithm. Water Resources Research, 15(8):2523-2532.
- Peralta, R.C. and A.H. Aly. 2000. Computing optimal pumping strategies for groundwater contaminant plume remediation. Section 8.8 in Handbook of Environmental Science, Health and Technology, Ed. Jay Lehr, McGraw Hill. p. 8.106-8.121.
- Peralta, R.C. 2000. Contrasting pump and treat system design by simulation model versus simulation/optimization model. In Proceedings, AICHE (in press). 2000 Spring National Meeting: Advancing New Technologies in Industry. Invited and presented at Remediation Process Optimization Session, 8 March 2000, Atlanta, GA. 7 p.
- Peralta, R. C., Wu, S., and Y. Huang. 2000. Optimal pumping strategies proposed for Massachusetts Military Reservation CS-10 TCE plume. Report submitted to EnviroTech Center, 20 p.
|
|