Journal of Water Resources Planning and Management

A screening model called the South Florida Systems Analysis Model (SFSAM) was developed to support the Central and South Florida Project Comprehensive Review Study (Restudy). The objective of the Restudy, performed by the Jacksonville District of the U.S. Army Corps of Engineers and the South Florida Water Management District, was to recommend a plan for improving environmental quality and urban and agricultural water supply reliability affected by the Central and South Florida water management project. As a “modeler’s model,” SFSAM was limited in scope and was used primarily to assist analysts in the development of alternatives. Additionally, model results provided some insights to operating strategies that could be tested and refined through more detailed modeling of the alternatives. This review of SFSAM includes a brief overview of the Central and South Florida project, a description of the model formulation, some interesting technical challenges faced in modeling the system, and representative and interesting results from the model application.

Booster disinfection is the addition of disinfectant at some critical locations of a water distribution network such that disinfectant residuals are maintained at a level greater than the minimum for public health. Compared to conventional methods that apply disinfectant only at the source, booster disinfection can reduce the total disinfectant dose. The present work investigates the booster facility location and injection scheduling problem in water distribution networks. The problem is formulated as a multiobjective optimization model. The objectives are minimization of the total disinfectant dose and maximization of the volumetric demand within specified residual limits. Multiobjective genetic algorithms are used for solving the problem. The model utilizes the theory of linear superposition in water quality modeling for calculating concentration profiles at network nodes. Unlike previous models, the present multiobjective approach does not require pruning of monitoring nodes to find feasible solutions; all demand nodes are considered as monitoring nodes. Application of the model to an example problem reveals that there is a critical point in the level of constraint satisfaction, after which the disinfectant dosage rate increases significantly in order to satisfy a few remaining constraints.

Deliberate contamination is generally viewed as the most serious potential terrorist threat to water systems. Chemical or biological agents could spread throughout a distribution system and result in sickness or death among the people drinking the water. Since September 11, 2001 the U.S. Environmental Protection Agency’s water protection task force and regional offices have initiated massive actions to improve the security of the drinking water infrastructure. A methodology is presented for finding the optimal layout of an early warning detection system (EWDS). The detection system is comprised of a set of monitoring stations aimed at capturing deliberate external terrorist hazard intrusions through water distribution system nodes—sources, tanks, and consumers. The optimization considers extended period unsteady hydraulics and water quality conditions for a given defensive level of service to the public, defined as a maximum volume of polluted water exposure at a concentration higher than a minimum hazard level. Such a scheme provides an EWDS for a deliberate terrorist external hazard intrusion, as well as for accidental contamination entries under unsteady conditions—a problem that currently has not been solved. The methodology is cast in a genetic algorithm framework for integration with EPANET and is demonstrated through two example applications.

The analysis and modeling of water distribution networks has been a well established engineering field for many years. However, important questions remain concerning the correct assessment of the spatial and temporal distribution of network user demands. To contribute to better knowledge and understanding of consumption patterns in an urban network, a stochastic model for residential water demand simulation is developed. The model is based on a rectangular pulse point process of residential consumption of given duration and intensity. Both variables are considered as statistically independent variables, with a nonhomogeneous point process used to describe pulse occurrences over time. The model includes a total of nine free parameters that define five different statistical functions. The parameters were calibrated from known demands in residential areas located in Milford, Ohio, and in Valencia, Spain. The model is also applied to the simulation of longer periods, with satisfactory agreement generally found between synthetic and historical series for most significant variables used in practical applications.

A methodology based on the range of variability approach (RVA) is presented for determining the feasible combinations of flow diversion and instream flow release for a projected diversion weir. The RVA is designed to support efforts to manage water system operations in a manner that minimizes impacts on natural hydrologic variability, and thereby minimizes ecological impacts. This approach is used to evaluate the prediversion flows and establish the riverine management targets in terms of 32 hydrologic parameters called indicators of hydrologic alteration (IHAs). The goal is to make the postdiversion flows attain the target ranges at the same frequency as that which occurred in the prediversion flow regime. A weir-operation simulation approach is employed to compute the postdiversion flows. Based on the simulation results, the degree of hydrologic alteration under various combinations of flow diversion and release is evaluated and plotted as a contour diagram for each IHA. Overlapping the contour diagrams of the 32 IHAs, three overall hydrologic-alteration regions are constructed. The feasible region, i.e., the overall low-alteration region, is defined by the combinations of flow diversion and instream flow release for which none of the 32 IHAs is significantly altered. The feasible combinations of flow diversion and release are further evaluated with their corresponding water-supply shortage indices. The proposed methodology allows for the incorporation of both water-supply and environmental protection concerns in water resources planning and management. The merits of this methodology are demonstrated with an application to the proposed Taitung diversion weir in Taiwan.

An approach is presented for short-term (i.e., daily and monthly) forecasting of municipal water use that utilizes a deterministic smoothing algorithm to predict monthly water use. The smoothing algorithm considers level, trend, and seasonality components of the time series. Daily deviations from the monthly average are then forecasted for up to six days using autocorrelation and weather dependence. While providing accurate operational forecasts, the approach required about six years of daily data to develop and validate the models. The approach is applied and evaluated for a number of municipalities near Tampa, Fla. Results show that the approach provides accurate daily forecasts as measured using a validation period of about three years.

A methodology for evaluation of risk for a river water quality management problem is presented. A fuzzy waste load allocation model is solved with a simulation–optimization approach for obtaining optimum fractional removal levels for the dischargers to the river system. With the optimal fractional removal levels, sensitivity analysis and first-order reliability analysis are applied to identify key random variables which influence the water quality simulation model output, and key checkpoints where the model output is more likely to be affected (i.e., has high variability). Frequency distributions of the output variable are obtained at the key checkpoints using Monte Carle simulation with the key random variables as input variables. The event of low water quality at a checkpoint in a river system is considered as a fuzzy event, with appropriate membership functions defined for the fuzzy risk of low water quality. With the help of fuzzy membership functions and frequency distributions, fuzzy risk levels are computed at the key checkpoints. The proposed methodology is demonstrated through a case study of Tunga–Bhadra River in southern India.