Journal of Water Resources Planning and Management

A knowledge‐based system for automating the calibration of the runoff model of the widely used storm water management model (SWMM) is described. The paper focuses on the key components of the knowledge‐based system: (1) The results of a sensitivity study of calibration parameters; (2) the strategy used for parameter selection and value assignment so as to match simulated and measured hydrographs; and (3) the knowledge representation scheme used to implement the calibration strategies. An example of the application of the developed system on a catchment in Singapore is also presented and discussed.

Data on rainfall, runoff, water table fluctuation, and subsurface geology are obtained in an open‐cast limestone mine in Kerala, India. These data are analyzed to identify the most critical condition during the full life period of the mine in respect to inundation within the mine pit and to estimate the associated maximum possible inundation rate. The use of design inundation rates less than the maximum value is also studied, and the storage space required to take care of actual inundation rates exceeding the design rate are estimated.

This paper presents an application of a system analysis tool to evaluate the design and operation of a water treatment plant in Taiwan. The optimal design and operation problem is posted as a nonlinear programming problem in which the decision variables are the capacities of the treatment components and the alum and chlorine feed rates used in the treatment process. The solutions to this nonlinear programming problem are obtained from a modified Hestenes' method. Application of the model to the reevaluation of an existing treatment plant in Taiwan revealed that, with the exception of the rapid mixing basin, the optimal capacities determined by the proposed model came fairly close to the original design. Optimal alum and chlorine feed rates and the optimal construction and operation costs for the plant can be estimated from the model. The results of this study suggest that the proposed tool can be used to evaluate water treatment plant design and operation in the future.

The Delta‐O'Neill simulation model (DELMOD) combines the results of the three models used for operating the California State Water Project (SWP) facilities south, north, and west of O'Neill Forebay. Network flow programming (NFP) is used to model a portion of the California Aqueduct containing 30 aqueduct pools, two reservoirs, four pumping plants, and one pump‐generating plant in the system. DELMOD allows modification of the periodic schedule at the Dos Amigos and Banks pumping plants and the San Luis pump‐generating plant to ensure that O'Neill Forebay storage is maintained within allowable limits. For the daily model of hourly detail, the network representation of the system has 4,212 arcs and 1,395 nodes. To improve the model computational efficiency, a condensed network of 534 arcs and 123 nodes is used to find the initial solution. The final solution is then found by use of the expanded network.

The application of extended linear quadratic Gaussian (ELQG) control, a nonlinear stochastic optimization method, to the Salt River Project reservoir system is described. The optimization problem was posed as the maximization of total hydropower avoided cost, subject to constraints that represented system mass balance and physical or operating limitations on releases. The six surface reservoirs, ground‐water storage, and planned cyclic ground‐water storage were modeled as six equivalent reservoirs. The ELQG control algorithm was applied to the historic reservoir inflow for the 51‐yr period 1931–1981 using a 12‐month sliding window. Operating policies for each of the 600 months in the test period were developed and compared with the present operating policy for two reservoir configurations: the existing configuration, and a reduced storage configuration that mimicked temporary system constraints during the construction period for the planned alteration of Roosevelt Dam. The ELQG control results showed that significant increases in hydropower avoided costs could be achieved while maintaining water supply reliability at current levels or higher.

Assessment of the reliability of a water distribution system through evaluation of all possible combinations of mechanical (pipe) failure is computationally intensive. It is nonpolynomial (NP) hard. An approach is proposed in which the probability of varying levels of simultaneous pipe breaks and the probability that the various levels of simultaneous breaks will cause network failure are combined to obtain estimates of overall network reliability. The probability Of a fixed number of simultaneous failures occurring is derived from analyses of historical data. The probability that a particular number of simultaneous failures will cause a system failure is estimated from simulation analyses. In analyzing the historical data, both the maximum number of simultaneous failures and the distribution of the varying levels of simultaneous pipe failures are analyzed for different sizes of a distribution network. A sample application of the procedure is presented, using the pipe failure data from the city of Winnipeg and a smaller scale sample network taken from the literature.