Journal of Water Resources Planning and Management

The management of wet-weather flow (WWF) is necessary to maintain the quality of urban water resources. Throughout history, strategies have been implemented to control WWF for many reasons, including flood and water quality control, aesthetic improvement, waste removal, and others. A comprehensive literature review has been conducted to determine past strategies and to revisit the historical developments of WWF management. Understanding these past strategies and the development of WWF-management systems over time will aid current and future generations in their WWF-management efforts. This paper summarizes the historical literature review, highlighting the development of WWF management from ancient times to the present. The relationship between past developments, the current state, and the future of WWF management is addressed by identifying several lessons learned.

A screening model is developed to support the selection of alternatives for meeting future water demands and protecting sensitive ecosystems in the Edwards aquifer region of central Texas. Important attributes of the model are its basis on existing ground- and surface-water simulation models and its incorporation of uncertainty in future supplies and demands. The model is used to evaluate trade-offs involving environmental and economic risks and to identify water resources alternatives that promise to be robust (i.e., those that can be managed in a flexible manner to meet society's goals under a wide range of hydrologic scenarios). Use of the model to help determine conjunctive surface- and ground-water management policies is also demonstrated. Although limited benefits can be obtained through more coordinated operation policies, results indicate that either large investments in infrastructure or substantial reductions in projected water use will be required to meet environmental goals with a high degree of reliability.

Several alternative formulations of a genetic algorithm for reservoir systems are evaluated using the four-reservoir, deterministic, finite-horizon problem. This has been done with a view to presenting fundamental guidelines for implementation of the approach to practical problems. Alternative representation, selection, crossover, and mutation schemes are considered. It is concluded that the most promising genetic algorithm approach for the four-reservoir problem comprises real-value coding, tournament selection, uniform crossover, and modified uniform mutation. The real-value coding operates significantly faster than binary coding and produces better results. The known global optimum for the four-reservoir problem can be achieved with real-value coding. A nonlinear four-reservoir problem is considered also, along with one with extended time horizons. The results demonstrate that a genetic algorithm could be satisfactorily used in real time operations with stochastically generated inflows. A more complex ten-reservoir problem is also considered, and results produced by a genetic algorithm are compared with previously published results. The genetic algorithm approach is robust and is easily applied to complex systems. It has potential as an alternative to stochastic dynamic programming approaches.

One of the major issues in planning the operation of reservoirs for energy generation is uncertainty in future reservoir inflows and energy demands. A reliability programming model addresses input uncertainty. This paper demonstrates the practical application of this model for the Manitoba Hydro energy generation system. The model evaluates the optimal reliability (probability of meeting system energy demand) by balancing the total benefits accrued from energy generation with the economic losses incurred from the decrease in reliability, through incorporation of a loss function in the model formulation. A four-step simulation algorithm is proposed to derive the loss function, a relationship between the reliability and its associated economic losses. The sensitivity of the operating policy computed by the reliability programming model for different loss functions is presented to demonstrate the model's ability to perform an explicit reliability-benefit analysis in medium-term planning for hydropower systems.

This paper is a summary of the efforts of Working Committee 4 (WC4) for the Great Lakes–St. Lawrence River Basin water levels reference study. A fuzzy relation analysis (FRA) model was used to analyze (including ordering, rating, ranking, and screening) numerous disparate alternatives that had accumulated over the course of WC4's study. Compared with other multicriteria methods for impact assessment under uncertainty, the FRA has advantages in data availability, computational requirements, and results interpretation. Relative strengths and weaknesses of different alternatives were evaluated by relating their impacts to a number of evaluation subcriteria. The results indicated that reasonable solutions have been generated, which were helpful for governments/stakeholders to obtain insights into the interrelationships between different system components. They, with results of other International Joint Commission research projects, also provided useful bases for assessing the relative effectiveness of different governmental actions to reduce the adverse impacts of fluctuating lake levels.

Based on measurements made on the Middle Mississippi River during the 1993 flood, stage-discharge relations for the rating curve are not single-valued. Analyses of these data suggest that the range of flow rates at a given stage is larger than that attributable to measurement replication error alone. Patterns of stage-discharge measurements also imply a tendency toward hysteresis. However, neither changes in stage nor changes in flow rate correlate with short-term scour and fill at gauging sites.

This paper compares computational performance of eight optimization algorithms used to identify the most cost-effective policy for in situ bioremediation of contaminated ground water. Six of these methods have not previously been applied to optimization of groundwater remediation. Numerical results were obtained for bioremediation of three problems based on two aquifers with time-invariant or time-varying pumping rates. Three major classes of algorithms are considered in the comparison: evolutionary algorithms [binary-coded genetic algorithm (BIGA), real-coded genetic algorithm, and derandomized evolution strategy (DES)], direct search methods (Nelder-Mead simplex, modified simplex, and parallel directive search), and derivative-based optimization methods (implicit filtering for constrained optimization and successive approximation linear quadratic regulator). Based on the three problems considered, the successive approximation linear quadratic regulator is the fastest algorithm. No one algorithm was consistently the most accurate on all three problems. The DES displayed an impressive combination of speed and accuracy. The DES has the advantage that it does not require derivative information. The BIGA was much slower and less accurate than all the other algorithms on the two problems BIGA solved.