Journal of Water Resources Planning and Management

A multi‐objective linear program is developed to study the operation of a metropolitan water supply system during drought. The Indianapolis Water Company is used as a case study. Twelve pressure districts, three reservoirs, and eight supply sources are included in the model. Four noncommensurate and conflicting objectives are identified and built into the linear program. The model is used to analyze operations during the drought of record and to develop trade‐off curves among the objectives. Experience with the model suggests that this methodology can provide information that can aid operational decision making.

In developing management programs to regulate point sources of waterbomewaste, it may be convenient or even necessary to subdivide the dischargers into groups. For simplicity and effectiveness in meeting water quality goals, it is desirable that the regulatory decisions governing one group of dischargers be minimally influenced by those governing other groups. One way to accomplish this is to separate the watercourse into sets of water quality checkpoints such that each set of checkpoints is associated with a group of dischargers and the effect of the dischargers excluded from a given group upon the checkpoints associated with that group is small. This paper presents a quantitative method for effecting such groupings. The method minimizes the impacts of the dischargers included in a group on checkpoints associated with other groups of dischargers. The method is illustrated using data for several river basins, viz., the Lower Fox River in Wisconsin, the Willamette River in Oregon, and the Mohawk River in New York.

A significant amount of research during the past few years has focused on the application of expert systems technology to problems of water resources management. While these investigations have led to speculation as to the benefits of intelligent reasoning applied to real‐time reservoir operation, working systems are nonexistent or in the preliminary stages of development and testing. This research presents a novel perspective on the use of knowledge‐based inferencing techniques applied to real‐time reservoir operation. A hybrid reasoning structure using both Bayesian and rules‐based inferencing is presented. Rules are used to achieve a real‐time simulation that is comparable to other rule‐based systems reflecting expert operations as proposed in the literature. The Bayesian mechanism then provides a judgment about the quality of recommended releases based on prior information and present conditions. An additional feature of this system is its learning capabilities that can be used for further refinement of system recommendations.

CVPOP is a nonlinear programming model for the optimization of the multi‐month operation of the hydropower system of the California Central Valley Project (CVP). CVPOP includes the dependence of energy values within each month on the capacity factor of the generating unit, avoiding the simplification of assuming constant monthly or yearly values as is common in other models. The model also includes contractual energy and capacity constraints which are nonlinear because of the powerplants' variable head performance curves (capability in MW and energy production rate in kWh per unit release versus reservoir storage). An example illustrates the difference between the values of the energy generated with operations schedules obtained using different objective functions.

The spillway design flood is a function of social, moral, economic, and technological restrictions. Given the significant initial cost of a spillway and the very high cost of dam failure and downstream damages, the economic sizing of the spillway becomes a very important decision. This paper presents a model that minimizes the social cost of a spillway. The model incorporates the uncertainty of the flood magnitude estimator in the optimization of spillway capacity, using an estimated flood probability density function and a relationship between spillway capacity and cost. The spillway design flood is explicitly treated as a decision variable, which is useful in the planning phase of a reservoir. To determine the effect of uncertainty in the flood magnitude estimator, the optimum spillway design flood is examined for different flood probability density functions with varying sample sizes. It is shown that with a limited sample size, the optimum spillway design flood may be as high as the probable maximum flood.

A mathematical model has been developed for estimating the design storage capacity of a reservoir by using the expected water demand, storage loss due to sedimentation, and physical and hydrological characteristics of the watershed. Suitable mitigative measures can be incorporated in dam design and reservoir operation to substantially reduce sediment entrapment in the reservoir, and to improve dissolved oxygen levels by releasing hypolimnetic waters from the reservoir. These measures may also greatly reduce streambed degradation downstream of the dam and consequent initiation of a new erosion cycle in the tributaries. Reservoir design based on storage conservation is of prime importance considering the dearth of new feasible sites for reservoirs. Economic analyses for different storage‐maintenance measures (such as undersluices and flushing pipes) have been investigated in terms of reduction in initial reservoir design storage, cost of installing measures, and cost of any dredging operations. These analyses are performed for a site in Illinois for several water‐demand levels and useful lives of the reservoir.

An ASCE Task Committee conducted an opinion survey of 563 engineering professionals on the adequacy of current educational programs in water resources engineering. Where the responses of practicing professionals and educators differed, separate tables present the opinions of both groups. The respondents call for many refinements and changes of emphasis and scope in current undergraduate, graduate, and continuing education programs. Specific recommendations include greater emphasis upon knowledge and skills that are important in practice, improvement of communications skills at all levels, and increasing the breadth of degree programs. Recommendations for strengthening entrance requirements for the profession are included, and a call is made for institutional adjustments that are necessary for those in practice to update their education and training. Respondents want the ASCE to continue to stimulate interaction and discussion that will lead to refinement and improvement of educational and training programs. A list of recommended implementing measures accompanies each of seven recommendations for action.

An evaluation of the hydrometric network of New Brunswick was performed starting from basic principles of hydrometric network design and proceeding through the analysis of the present network and its ability to meet user needs. Discharge stations used to monitor the streamflow component of the regional hydrology are assessed on their ability to contribute to the transfer of information to ungauged sites. A survey was conducted to determine users' needs. The results of the various analyses are combined using an audit approach based on selected rating factors. The final output includes specific network improvements designed to satisfy hydrometric needs. These improvements are presented in the form of a list of network adjustment scenarios that are available as a management guide. A summary of the approach and major findings is presented.