Journal of Water Resources Planning and Management

Transferable discharge permit (TDP) programs show potential as cost-effective methods of pollution control. Nevertheless, there remain uncertainties that, if not adequately addressed, might impair their success. Concerns include modeling difficulties that might cause erroneous predictions of cost savings and environmental performance. This study focuses on environmental modeling associated with the stochastic environment. The mean-value first-order second-moment (MFOSM) method is extended to demonstrate how changes, due to discharge permit trading, in the environmental quality mean and∕or variance of a system will cause the environmental quality reliability of the system to either decrease, increase, or remain unchanged. The Willamette River in Oregon and the Athabasca River in Alberta, Canada, are used as example case studies. These river systems are simulated to predict how they might respond if trading were implemented. In this study, environmental quality reliability is defined as the probability of the minimum dissolved oxygen (DO) concentration over the length of the river meeting a predetermined standard. Generally, for the Willamette River, trading improves the reliability, while for the Athabasca River, trading makes the reliability worse.

Conjunctive management of surface and ground water requires an understanding of the degree to which surface water resources will be impacted by aquifer stresses at different locations. Response functions quantify the rate of depletion or accretion of a surface-water body relative to a unit stress on an aquifer under simplified, but often realistic, conditions. Response functions for four reaches of the Snake River have been determined for each cell of the eastern Snake River Plain aquifer model grid in southern Idaho. Mapping the response functions for each reach creates a visual image of how effects of aquifer stresses at different locations are distributed among the four hydraulically connected river reaches. The mapped distribution of response functions reflects aquifer properties such as the distribution of aquifer transmissivity. Cluster analysis of the response functions for each cell and each river reach has proven useful for subdividing the aquifer into zones for the conjunctive management of ground water and surface water. The response function variance within each zone is minimized in this procedure. In most situations, zone boundaries defined through cluster analysis will likely be modified to better conform to existing political and administrative units. The selection of the number of zones, and the degree to which boundaries based on response functions are modified, reflect a balance between the exactness provided by the science and administrative convenience and efficiency. The approach of managing an aquifer through response function based zones has gained initial acceptance in Idaho, as evidenced by the citation of these zones in Idaho’s draft Water Management Rules. How management zones based on response functions will ultimately be used in Idaho water policy and procedures has not yet been determined.

Robust ecosystem analysis of water resource systems remains elusive. A principle reason is the difficulty in linking engineering models used to simulate physicochemical processes associated with project design or operation with biological models used to simulate biological population attributes. A retrospective shows that each modeling tradition can be generally assigned (with exceptions) into either an Eulerian or Lagrangian reference framework. Eulerian and Lagrangian reference frameworks can be coupled to create a new synthesis, the Coupled Eulerian-Lagrangian Hybrid Ecological Modeling Concept (CEL Hybrid Concept), capable of simulating different ecosystem processes that range widely in spatial and temporal scale. The foundation of the CEL Hybrid Concept is the coupler, a collection of algorithms based on conservation principles that transform and conserve data in a way that allows the two frameworks to share a common information base. The coupling algorithm allows the simulation to aggregate, disaggregate, and translate information, as required by each framework, so that processes that differ substantially in scale can each be adequately simulated. The coupled system is illustrated by linking a fish swim path selection model with a hydrodynamic and water quality model.

Water supply and allocation scenarios for the Klamath River, Ore. and Calif., were evaluated using the Systems Impact Assessment Model (SIAM), a decision support system developed by the U.S. Geological Survey. SIAM is a set of models with a graphical user interface that simulates water supply and delivery in a managed river system, water quality, and fish production. Simulation results are presented for drought conditions, one aspect of Klamath River water operations. The Klamath River Basin has experienced critically dry conditions in 1992, 1994, and 2001. Drought simulations are useful to estimate the impacts of specific legal or institutional flow constraints. In addition, simulations help to identify potential adverse water quality consequences including evaluating the potential for reducing adverse temperature impacts on anadromous fish. In all drought simulations, water supply was insufficient to fully meet upstream and downstream targets for endangered species.

A method incorporating a genetic algorithm tailored to EPANET for the conjunctive optimal design and operation of multiquality water distribution systems under unsteady hydraulics is presented and demonstrated. The objective is to minimize the total cost of designing and operating the system for a selected operational time horizon while delivering to consumers the required quantities at acceptable qualities and pressures. The decision variables for the design are the pipe diameters, tank maximum storage, maximum pumping unit power, and maximum removal ratios at the treatment facilities. For the operation phase, the decision variables are set for each time step of the total operational time horizon. These decisions include the scheduling of the pumping units and the treatment removal ratios at the treatment facilities. The constraints are domain heads and concentrations at consumer nodes, maximum removal ratios at the treatment facilities, maximum allowable amounts of water withdrawals at the sources, and return at the end of the operational time horizon to a prescribed total storage in the water distribution system tanks. The model is explored through two example applications.

We formulate and solve an estimation problem for identifying both the time and location of contamination sources in municipal water networks using concentration measurements from a sparse sensor grid. Previous work showed that the direct sequential approach was insufficient to solve the time-dependent problem. Instead, a direct simultaneous approach is used, converging the network model and optimization problems simultaneously. An origin tracking algorithm is presented to reformulate the pipe expressions and characterize the time delays associated with network pipes. This algorithm removes the need to discretize along the length of the pipes and is efficient for large networks with many source and output nodes. The solution of the resulting nonlinear program provides the complete time-dependent injection profiles, identifying both the time and location of potential contamination sources. The effectiveness of this formulation is demonstrated on a network model for a real municipal water network.

Recent developments in the field of computational intelligence (CI) techniques are helping to solve various problems of water resources modeling and management. These techniques have also shown their potential as an alternative approach to conventional controllers. In this paper, artificial neural networks (ANN) and fuzzy systems (FS) are shown to be efficient alternatives to using optimal control algorithms in real-time control of the polder water system of Overwaard in The Netherlands. The relation between the optimal decision or action and the influencing parameters are learned by ANN and FS and then used to derive the decisions and control actions in real-time. It was possible to reproduce the centralized behavior (in terms of water levels and corresponding discharges) of optimal control action by using easily measurable local information. Moreover, it is demonstrated that model simulation with external intelligent controllers is ten times faster than that with the optimal control.

A conceptual examination of the energy impact of leaks in systems with storage is undertaken. Consideration of how leakage is experienced at the pump is followed by an analysis of how different leakage levels alter energy costs for a rudimentary system with three topological configurations: two with a storage tank located at different points, and one without storage. Additionally, two friction regimes are subsumed in the analysis. EPANET 2 simulations are used to determine system pressures, storage tank levels, energy costs, power consumption, and leakage volumes for all scenarios at five levels of leakage. Leaks increase operating costs in terms of lost water and extra energy consumption for all systems, and when a price pattern is implemented, the financial cost of energy can sometimes be traded off with actual energy consumption. Storage in a system does not guarantee lower energy use relative to direct pumping, and in some cases it may promote higher leakage due to elevated system pressures. The results, though system specific, suggest that the importance of leaks in a system with storage depends on a number of factors, especially the relative locations of system components and the pumping strategy. Thoughtful consideration of the latter can be instrumental in achieving operation that balances financial and energy conservation objectives. To further test key relationships, a representative network is briefly considered. In all cases, the percentage increase in energy cost is greater than percentage leakage when the same pressure requirements are met.