Journal of Water Resources Planning and Management

Optimal allocation of powerplant releases during peak demand periods carries an economic advantage in the operation of hydropower systems interconnected to large electrical networks. This paper presents an alternative formulation for determining release strategies when the objective is not maximizing total hydroelectric generation per se, but rather the economic benefit stemming from it. The proposed hydropower problem is formulated within the realm of concave programming without sacrificing realism in the formulation, thus yielding a nonlinear‐concave objective function. A nonlinear mathematical programming technique, sequential quadratic programming (SQP), is used to solve the problem by successive solution of quadratic programming problems. The concave characteristic of the nonlinear objective function is fully exploited by SQP, which exhibits a rapid convergence to the global optimum. An expeditious procedure to formulate the hydropower problem under the SQP framework is also presented. The methodology is tested on an existing multi‐reservoir hydropower system in Argentina. SQP proves very efficient to the extent that only a few minutes are necessary to solve the problem on a microcomputer. Furthermore, SQP is found to be superior to sequential linear programming in the rate of convergency toward the optimal solution.

Optimal operation of a water supply system is solved by progressive optimality (PO), an iterative dynamic programming (DP) method. Given the forecasted demands for the coming 24 hr, the initial and final conditions in the reservoirs, the hydraulic properties of all system components, and the variable energy cost over the day, an optimal schedule of pump operation is found. The algorithm cycles iteratively over the time steps (hours of the day) and network subsystems, and converges to the optimum from any (feasible or infeasible) initial solution. The global optimum is guaranteed only under certain conditions; otherwise a local optimum may be reached. The method is developed and demonstrated on a regional water supply system with eight reservoirs and seven pumping stations.

A multiobjective urban land‐use/water‐resource model, appropriate for use at the strategic planning level in less‐developed countries (LDCs), is described and applied to Cali, Colombia. Objectives are to minimize service costs and biochemical oxygen demand (BOD) loadings and maximize service levels while accommodating population growth. Vector optimization based on generating techniques allocates new residents to potential urban‐expansion zones while simultaneously determining the optimal mix of various water‐supply and sanitation systems. Potential flood impacts are then simulated and the results used to constrain development in critical areas. Further flood control and reduction of nonpointsource pollution via detention basins are examined. Postoptimality analysis shows solutions to be quite sensitive to estimated water consumption and preset density limits. Deficiencies due to the optimization‐simulation linkage and the runoff modeling assumptions are discussed and their significance for such modeling in LDC settings is appraised.

Management of water quality in estuarine systems is a multistepped process, ranging from characterization of the hydrological and hydromechanical properties that govern water quality changes through identification of waste loads, regulatory constraints, and water quality attributes to identification of alternative candidate plans and selection by so‐called decision makers‐ of the best plan to be implemented. The ultimate step in this complex process is to operate the physical system within the stipulated performance goals. Recently, with the advent of everincreasing computer power, geographic information systems (GIS) and interactive graphics, supporting integrated user‐friendly data management and processing, have been added to the body of tools intended to support the decision process through its many steps. This paper focuses on implementation of a new generation microcomputer as the sole host of the decision support system, and seeks to demonstrate how it may be used at each step of the decision process. Special attention is given to features like: (1) Object‐oriented data and control handling; (2) the integration of simulation models into the decision support system; and (3) the water quality management process. The case of a small estuary in Southern California is presented as a demonstration of the approach.

The method of residuals was recently used by the U.S. Geological Survey (USGS) to delineate seven flood regions for the state of Kentucky. An alternative approach is to use cluster analysis in conjunction with important statistical properties of the maximum annual flood‐peak series. Applying the FASTCLUS clustering procedure of the Statistical Analysis System, five cluster regions are identified using a similar data base as the USGS study. Flood regions delineated under these two methods of flood regionalization are then compared by examining trends in important hydrological characteristics within each of the regions, and through a discriminant analysis based upon watershed physical attributes. Results show that, although cluster regions are in no way similar to those defined by the method of residuals nor coincident with geographical boundaries, they are more distinguishable and better discriminated in terms of the hydrological characteristics controlling flood response than the USGS regions. Furthermore, standard errors associated with the regression equations relating the log‐Pearson, Type‐III 50‐year flood estimate to watershed physical attributes are comparable under the two methods of regionalization.

The advantage of introducing Landsat‐derived land‐cover information into a flood‐flow forecasting model is examined. Two modeling alternatives, with and without Landsat data, were applied to 10 km by 10 km grid sizes. The landcover information is primarily derived from Landsat imagery and used directly for rainfall‐excess estimation and runoff routing. Runoff is calculated separately for each of six land‐use/land‐cover classifications for each watershed element. Without Landsat data the model is a lumped‐parameter mpdel, while with Landsat data it is a fully distributed model. The advantage of calculating runoff for each landuse/land‐cover class separately is that a watershed element can be substantially larger than a typical homogeneous hydrologic unit. The paper reports that the improvement of predicting flood peaks and total runoff gained by using Landsat data is at the 10% level of significance.

Water customers suffer hidden private costs if complete price information is not provided when the salinity in water increases. Previous attempts to estimate the damages or hidden costs resulting from increased salinity have ignored the impact on the production of goods made with the more saline water. Assuming rational economic behavior and a fixed production budget, this note provides an economic rationale for including these effects in future damage estimates. Since water is most often used as an input combined with other inputs, derived‐demand analysis should be used in future damage‐estimate studies to account properly for shifts in water use and productivity caused by the unannounced increase in salinity. This note shows clearly that damages are underestimated when lost output from productivity shifts is ignored. The note concludes that an informational campaign by the water utility, if clean up subsidies are ignored, may be the best way to eliminate the hidden cost to water users.

A method for extending the available water supply in a region is discussed. Using dilution (mixing) junctions, brackish water is mixed with higher quality water, thus producing the required water quality. For good results, the process should be controlled automatically. Typically, the controlled variables are salinity, outlet flow, water level or outlet pressure. Mixing junctions are classified, possible configurations discussed, and the most promising analyzed in detail. Steadystate equations are derived, and a microcomputer program simulating steady‐state operation is presented. This program helps to determine various design constants of the junction. For a dynamic study, a hybrid computer simulation was made, with dynamics of instrumentation simulated on an analog computer. A conclusion of this simulation is that equal‐percentage rather than linear‐control valves should be used. A cost analysis showed that lowest cost is achieved using a self‐actuated pressure or flow regulator, and single‐speed floating control for salinity. A circuit implementing single‐speed floating control, using a conventional control valve and two small solenoid valves, is described and tested by hybrid simulation.