Journal of Water Resources Planning and Management

In water supply planning, managers must assess both quantity and quality of sources. These assessments require adequate data, which is not always available, even in developed countries, and much less so in developing countries where the availability of water quality data is especially limited. Yet, most developing countries need to expand water supplies quickly, and methods for rapid assessment, with limited data, are needed. A method for such assessments is discussed and its application to Mexico, a rapidly-growing nation with an ambitious plan for the expansion of its water supply, is shown. The paper concludes with suggestions for how other developing nations can apply the concepts presented.

This study examines the presence of nonconvexity and multiple extrema in the feasible region defined by the concentration response function. The response function is defined in the context of concentration constraints imposed on groundwater quality management models and is constructed assuming that concentration constraints are imposed at every node in a numerical grid. The groundwater system is represented by a homogeneous numerical simulation model. The response function is graphically depicted as a function of pumping rates at two wells. The dependence of the response function on simulation model and management model characteristics is examined by varying dimensionless parameter values. This study shows that the response surface is nonconvex and contains multiple local extrema over a wide range of parameter values. These features of the surface are most prominent at the transitions from extraction to injection pumping, Binding constraints often occur at or near low-velocity zones.

A method for exploiting the sparsity structure of a finite-element simulation model in a linked dynamic optimal control model is developed. The linked optimization/simulation model has been used for the computation of time-varying optimal pumping rates for the pump-and-treat remediation of contaminated ground water. The methodology presented reduces the computational effort involved in the determination of time-varying optimal pumping rates by an order of n, from n³ to n², where n is the number of non-Dirichlet nodes used in the simulation of the aquifer. The method presented uses the characteristic banded structure of finite element model matrices in derivative computations used by the optimal control algorithm, and also within the algorithmic computations of the optimization method itself. Timing results demonstrating the improvement of this method for problems having n= 100 to n= 1,575 (state dimension 2n= 200 to 2n= 3,150) are presented. It is shown that the efficiency of the sparse algorithm is highest when the length of management periods equals the length of simulation time periods. The results indicate that for a problem with 2n= 1,000, the sparse algorithm is as much as 98 faster than an algorithm neglecting sparsity.

The South Florida Water Management District has developed a comprehensive management strategy to assess eutrophication in Lake Okeechobee. Key components of this strategy are research, monitoring, and modeling of the lake. The purpose of modeling is to develop tools that aid in the prediction of lake eutrophication, and the evaluation of alternative management options. The initial modeling effort started a decade ago, using Vollenweider loading plots to predict trophic status. Today a more rigorous modeling framework, WASP/EUTRO, is being configured for Lake Okeechobee. Preliminary model results reproduce the seasonal trends of water quality constituents in the lake. While model refinements designed to better address management questions continue, current modeling analysis and results are reported in this paper.

In India, reservoirs are usually constructed to serve multiple purposes, such as irrigation, municipal and industrial water supply, hydropower generation, and flood control. Because of the high temporal and geographical variability of rainfall in this country, reservoir operation occupies an important place in the utilization of water resources. The operation of the Sabarmati system, consisting of four reservoirs and three diversion structures, was studied. The function of the system is to provide a municipal and industrial water supply, irrigation, and flood control. For conservation regulation of the system, rule curves were derived for the various reservoirs. Using the simulation analysis, the rule curves were fine-tuned to achieve the targets to the maximum possible extent. The operation procedure for flood regulation was developed for the Dharoi reservoir. Since the potential damage center is located far downstream of the dam, the modeling of flow in the Sabarmati River at Ahmedabad was carried out. Software was developed to assist the dam operator in determining the safe release from the Dharoi reservoir during floods.

We have developed the first application of optimization techniques to the design of in-situ bioremediation, and have encountered a number of computational issues that have not arisen in previous efforts to optimize the design of pump-and-treat remediation. In this paper, these computational issues are discussed and a number of strategies for improving performance of the model are presented. The issues addressed include increased computational complexity, difficulties in convergence of the model, and convergence to local minima. The strategies developed will be useful for any application of optimization techniques to nonlinear processes in water resources.

Soil vapor extraction is a cost-effective technique for removing volatile organic compounds (VOCs) immobilized in the unsaturated zone. The design of a soil vapor extraction system (SVES) requires that the number and location of extraction wells and the extraction schedule for each extraction well be determined. In this study, schedule optimization for selected active wells is embedded in location optimization to provide an equal basis for system comparison. A combinatorial two-change local search is employed in both optimizations, but with different user-defined properties to address their distinct problem characteristics. The results indicate that the proposed two-change local search is robust and produces reliable designs. Impacts of soil heterogeneity and initial guess of design are evaluated. A detailed analysis of optimization results indicates that extraction well locations dominate the efficiency of contaminant removal. To reduce additional computation, a screening test based on results at the early stage of schedule optimization is suggested to reject obviously inferior designs.

Objective functions to describe the operating and capital costs of granular activated carbon (GAC) treatment are incorporated into a management model for dynamic optimal pump-and-treat groundwater remediation design. The optimization approach, quasi-Newton differential dynamic programming including a constant shift in the Hessian Matrix, could not reliably handle the nonconvex objective functions for GAC system. To avoid the nonconvexity, linearized iterative approximations of operating costs were used to solve the multiple management period problem, although this approach does not guarantee global optimality for the original nonconvex objective function. Repeated optimizations assuming different size absorbers were required to consider the fixed costs of the GAC system. Basing the cost function on concentrations at the beginning of each management period provided reasonable conservative cost estimates. Dynamic policies were found to be superior to steady-state policies even when capital costs of treatment were included. This work suggests that improvements in the differential dynamic programming algorithm may be needed for it to be a robust technique for realistic nonconvex objective functions.