Operating Reservoirs in Changing Conditions

The California State Water Project (SWP) is one of the largest water and power systems in the world. It conveys, on average, 2.4 million acre-feet of water annually to 29 long-term water contractors through 20 pumping plants, 8 hydroelectric power plants (includes 3 pumping-generating plants), 25 dams, 34 reservoirs, and more than 700 miles of aqueducts and pipelines. The SWP is also the largest consumer of energy within California. The California Department of Water Resources (DWR) is charged with managing, operating, and maintaining the SWP. With over 150 large hydro units and associated supporting equipment to operate and maintain, DWR recently adopted an Asset Management Plan to permit high-level proactive budgeting of all future SWP projects. Fundamental to DWR's Asset Management Plan was development and subsequent implementation of a Condition Assessment Program (CAP). Consistent and regular condition assessment of SWP plant facilities ensures equipment health for an aging system (35+ years) and enables DWR to operate and maintain the SWP at high levels of operational availability (OA). High OA is critical to DWR since California operates in the deregulated (in 2001) energy market. Unscheduled or prolonged scheduled outages can result in lost opportunities for generation revenue and/or increased costs for daytime pumping. DWR's CAP has three primary planning objectives: 1) long-term — objectively identify future SWP projects for proactive budgeting (1- to 5-year budget cycle), 2) medium-term — enable maintenance resources to embrace and move toward a Reliability Centered Maintenance philosophy, and 3) short-term — identify projects/equipment that may require immediate maintenance. The DWR CAP divides a hydro facility into five equipment groups (Unit, Transformer & Switchyard, Balance of Plant, Communications & Controls, and Waterways), and each equipment group undergoes rigorous inspections (visual, functional, reliability, and test) to determine the overall condition of a facility or plant. Detailed rating guides ensure that equipment condition scores are objective and repeatable. The rating guides detail how to score all equipment regardless who is conducting the CAP inspection. The end result is a plant condition score that enables management to determine where to allocate maintenance resources and future funding. Collected data that is not current (five or more years old) will result in a low confidence score, which indicates that maintenance is not being performed, not being documented, or a combination of both.

Seventy years ago, the Bureau of Reclamation (Reclamation), using monthly time step mass balance modeling and a flow surrogate for Delta water quality, reviewed the California Department of Public Work's water resources plan for California's Central Valley. Today, Reclamation's Central Valley Operations Office (CVOO) of the Mid-Pacific Regional Office in Sacramento, California, continues to use similar technology, albeit with a modern spreadsheet, to allocate water in one of the most complex water delivery systems in the world.

Human beings have delineated the land created by God into countries and states. But the natural resources which have been bestowed upon the human beings have no regard, whatsoever, for the international boundaries. In no other natural resource is this fact more evident than in the case of water. Rivers know no boundaries, be they national or international. The problem of sharing the waters of transboundary watercourses has traditionally been viewed as one lying exclusively within the domain of legal doctrines. However, lack of information about the futuristic scenarios of possible water allocations acts as a deterrent in resolving the conflicts pertaining to transboundary watercourses. It is increasingly being felt that technology, in the form of simulation modelling, has a very significant role to play in this context. However, law and technology so far stand isolated in their respective spheres. This paper is an attempt to bridge the gap between law and technology. The evolution of international laws concerning conflict resolution in the international watercourses has been discussed with special emphasis on underlying philosophies and principles. Furthermore, the role of hydrological modelling in bringing about an effective resolution of the problems pertaining to the sharing of international waters is analysed. A conceptual framework is proposed which illustrates how simulation modelling can help in ensuring community based participatory decision-making.

The U.S. Bureau of Reclamation (Reclamation) and CH2M HILL are developing a water resources operations and planning computer model to simulate the management of water in the American River watershed on a daily time step. The model is being developed using the daily time step version of the CALSIM II software, developed by Reclamation and the California Department of Water Resources. The project includes the development of a new 82-year historical daily and monthly hydrology that is being computed using historical flow, storage, and evaporation data from the USGS and from each water agency managing a project in the basin. The model includes operating rules that have been developed to characterize the operations of facilities owned by each water agency within the watershed that affect, and are affected by, the operations of Folsom Lake, which is a key component of Reclamation's Central Valley Project. The model completed under this project will provide a flexible foundation for potential future analyses including ensemble forecasting, evaluation of climate change scenarios, assistance with weekly operations forecasts, assistance with lower American River temperature forecasts, the simulation of water transfers, and the simulation of hydropower operations.

Efforts in previous decades, largely culminating in the release of the World Commission on Dams Report, have engendered a more circumspect approach to dam construction and management—one which incorporates consideration for the environment, health, equity, stakeholders, and livelihoods. Such integration nevertheless often remains at a qualitative level, preventing tangible incorporation of these factors into Decision Support Tools (DST) for water management at a basin or sub-basin level. This paper uses the experience of the Senegal River Basin (SRB) to generate suggestions for how public health and smallholder livelihood concerns can be explicitly and quantitatively incorporated into operations decisions of the Koka dam in Ethiopia. The study examines the operational tradeoffs made among livelihoods, health, and more conventional water needs such as irrigation and hydropower in SRB water management strategies over the last two decades. The examination of these tradeoffs is used to produce general lessons and specific suggestions that are applied to the operations of the Koka reservoir in Ethiopia.

Water resources management in Arizona is critical to the development and sustainability of local communities. Several regions in the state are facing immediate challenges to meet water demands. However, tools to assist the wide range of water interests understand the impact of decisions are lacking. This paper describes the development of a dynamic simulation based decision support system for the Upper San Pedro Partnership (USPP). This paper briefly presents the planning that they USPP has undertaken to meet the challenges. The emphasis of the paper is the DSS description. The Upper San Pedro is the one of the last uncontrolled rivers in Arizona and has a thriving riparian zone that also serves as a bird passageway. However, groundwater demands of the growing population and industry exceed natural recharge. The USPP was formed by interested parties to develop a strategy to maintain the riparian zone health while allowing for continued water use. The DSS will support the USPP in identifying water management options that eliminate groundwater overdraft, allow for growth and development within the Upper San Pedro basin, and maintain/raise groundwater levels at desired locations along the San Pedro River. In addition to a mass balance, a groundwater model has been represented through response functions to provide information on the impacts of decisions on river levels. Plans for including measures of riparian zone health are underway. Model development was completed over two years with near monthly interactions with the USPP to verify assumptions and determine structure of decisions to incorporate and output to present to users. These meetings were critical to ensure the DSS properly represented the system and was acceptable to all USPP members.

Dez reservoir dam is one of the famous dams in Khuzestan province in southwestern part of Iran. Operation of Dez reservoir dam and it's strategic position has an important role regard to providing country's net power, it's long precedence in operation, Dam dimension, multidisciplinary uses and existence of essential land use in downstream. Thus inflow forecasting could have a great role in effective operation of dam, flood non-structural and risk management by applying an effective flood warning. The more accurate the forecasting of inflow to reservoir will be results the less peak outflow and less risk to downstream. This paper presents a new Conjunctive nonlinear model using Wavelet transforms and Artificial Neural Network. It plays an important role in improving the accuracy of 15-day and monthly prediction inflow time series. Calibration and verification of wavelet network model for prediction of inflow to Dez reservoir as the case study have shown that the method is applicable. For the particular case of the Dez reservoir, the overall forecast accuracy of the conjunction model, as measured by R², RMSE (Root Mean Square Error) and NMSE (Normal Mean Square Error) was better than conventional ANN prediction model.

This paper explores the effects of lake evaporation and precipitation, reliability levels, and streamflow probability distribution on reservoir capacity and average release. A series of reservoir-analysis deterministic and stochastic models of increasing complexity are employed to examine these effects, which are illustrated using a mid-sized reservoir in Santa Barbara County, California. The deterministic reservoir models show evaporation dominating reservoir water balance, and, for the most part determining optimal reservoir capacity and annual releases, whereby comparatively larger reservoirs are required to compensate for the water lost to evaporation. The more complex stochastic models use chance constraints and bootstrapped reservoir inflow probability distributions, and suggest that optimal capacity and annual releases are dominated by reliability levels, with very large optimal reservoir capacities and annual releases required when the reliability levels approach one.

Alamo Dam is a Corps of Engineers multi-purpose project located on the Bill Williams River, a major tributary of the lower Colorado River, in western Arizona. The project was completed in 1968, and its 1 million acre-feet of storage space was allocated among the congressionally authorized purposes of flood control, water conservation, and recreation. The operation of the project has changed significantly since its completion in an effort to increase project benefits and preserve the natural environment affected by the operation of the dam. Competing goals and objectives of a number of water resource and natural resource agencies coupled with the hydrologic characteristics of the watershed (particularly the high variability in annual runoff) were the source of conflict with respect to operation of the project in the 1980's. In addition, the original water control plan for the dam, formulated prior to the National Environmental Policy Act and the Endangered Species Act, had made no provision for the ecological needs of the Bill Williams River. Natural resource agency concerns stemmed from the fact that the Bill Williams River harbors several endangered species and contains the best intact southwestern riparian habitat (primarily cottonwood and willow) in the lower Colorado River basin. Beginning in 1990, a multi-agency collaborative effort among Arizona Department of Game and Fish, Arizona State Parks, Arizona Department of Water Resources, Bureau of Land Management, Bureau of Reclamation, Fish and Wildlife Service, and the Corps of Engineers, along with the 2002 addition of The Nature Conservancy, has successfully worked to improve the operation of Alamo Dam. This paper describes the evolution of the water control plan for Alamo Dam; the approaches taken to address the needs and objectives of the various stakeholders; and the importance of continuous multi-agency engagement and collaboration.

The Municipal Subdistrict of the Northern Colorado Water Conservancy District (Subdistrict) operates the Windy Gap Project that diverts water from the Colorado River to municipal water users on the Northern Front Range of Colorado. A majority of the water diverted by the Windy Gap Project is derived from the Fraser River which is a tributary of the Colorado River. The water is pumped from a forebay reservoir located on the Colorado River, near the confluence with the Fraser River, through a six-mile pipeline to Lake Granby. Stream flow forecasts are vital to this Project due to the intensive pumping energy required and the timing of upstream water diversions by senior water rights. In order to optimize the energy intensive water pumping, the Subdistrict implemented the RiverTrak hydrologic modeling application (Riverside Technology, inc.) to perform extended runoff predictions of the Subdistrict's allocation of the Fraser River. The RiverTrak software employs snow and soil moisture accounting models, initialized to current conditions and driven by historic climatologic inputs, to predict the runoff generated in the Fraser river basin. The Subdistrict uses these probabilistic predictions to help manage an optimal pumping schedule.

Large public works projects cannot be implemented without input from the public The public will not tolerate being left out of the planning. More than ever, public involvement is an essential tool that public agencies must utilize in order to gain consent for public projects. The Freeport Regional Water Project (FRWP) in Sacramento, California, is a case study about the importance and necessity of involving the public at the outset of a project and continuing to work with them until the project's completion. The FRWP is a water supply solution for the Sacramento County Water Agency (SCWA) and the East Bay Municipal Utility District (EBMUD) of Oakland. The FRWP will divert water from the Sacramento River and transport it via pipelines through Sacramento and San Joaquin Counties to SCWA and EBMUD facilities. SCWA will use the water to offset groundwater usage. EBMUD will use the water in dry years only as a supplemental water source. The public introduction of the FRWP in 2003 was met with boisterous local opposition. Residents of the neighborhood adjacent to the proposed water intake site on the Sacramento River opposed the location of the future structure. They were concerned about construction inconveniences and long-term operational nuisances including noise, aesthetics, chemicals, security and potentially declining property values, and they mobilized together to demonstrate resistance to the construction and operation of the facility near their neighborhood. The Freeport Regional Water Authority (FRWA) responded with a comprehensive public outreach strategy that included numerous small group meetings and community presentations. Based on input from the community, FRWA made commitments to minimize operational noise, maintain a secure facility and site, and redesign the layout of the facility. Through its public involvement program FRWA turned the community's opposition to the project into consent. FRWA continues to collaborate with the residents on the design of the project as well as construction and operations planning.

A numerical modeling study of the diversion structure gate operations is presented for the proposed diversion structure on the Rio Grande at Albuquerque, New Mexico. The Rio Grande diversion structure is composed of independently controlled adjustable height overshot gates across the channel, a fish passage through the diversion structure, and an intake channel for 3.7 m³/s of maximum river diversion. Steady state flow patterns induced from the gate opening were simulated and compared with physical model measurements to evaluate the numerical modeling accuracy. The numerical model shows positive agreements with the 1:24 scale distorted movable bed physical model measurements. For the diversion gate operation test, 4.25 days of a snowmelt period hydrograph data was simulated. It is found that the numerical model for the diversion structure, including the overshot gate system, can be a useful prediction tool for the diversion gate operation. An analysis of the results of the snowmelt period dynamic simulation show that the gate opening can significantly affect the Rio Grande flow condition.

A 1998 meeting of a small group of scientists and engineers from The Nature Conservancy (TNC) and Louisville District led to a cooperative effort to modify regulation and operation of Green River Lake. This initial cooperative effort led to the national Sustainable Rivers Project between TNC and US Army Corps of Engineers. Green River is home to 71 mussel species and 151 fish species contributing to its rank as the fourth-most biologically diverse river in the nation. Improved passage of water through the Corps dam to more closely mimic naturally occurring flows, as would occur without the lake, benefits the life cycles of many aquatic species. The two most critical physical factors are water volume and temperature. The successful documentation of environmental benefits to the Green River, especially reproduction of endangered species of mussels, led to the District Engineer signing a Finding of No Significant Impact on March 13, 2006, recommending the continuation of the three-year experimental regulation of Green River Lake as the permanent method of operating this multi-purpose reservoir. Division Commander's approval was received on May 17, 2006. Continued monitoring will determine the long-term success of this effort and/or the need to make further changes.

Through recent technological advances, hydraulically operated valves, optionally combined with programmable logic controllers, are now available for any size dam. The system is composed of the valve-cylinder assembly, the controller cabinet containing the hydraulic pump, 12-volt gel cell battery bank, logic processor, and solar panels to charge the batteries. The programmable controller can respond to a wide range of data such as rate of rise, water level, rainfall, inflow, outflow, time of year, and remote command. Because it is battery powered, it can function during an electrical outage, or be used at a site where there is no electrical power. The controller is capable of cascading logic sequences, allowing the system to become an integral, autonomous part of the dam's Emergency Action Plan. Communication to and from the system can occur by telephone, cellular phone, satellite, internet, or radio, or it can be tied in to an existing SCADA system. In the event of an emergency, having automated controls and redundant warning systems shows that the dam owner made every effort to insure public safety. While this system offers many advantages at the single reservoir level, employing it on multiple installations allows for precise control of water levels in normal operation, and can serve to buffer the effects of a high flow event.

A new Truckee River Operating Agreement (TROA) has been proposed and could replace the current river and reservoir operating policy as soon as late 2008. Operating policy specified by TROA adds complex layers to existing policy. The policy layers provide new and complex water management options for stakeholders. Applying these options will require substantial interagency cooperation and coordination. Decision support tools are being developed, by an interagency group, to support required operating procedures and to analyze benefits and risks of voluntary water operations permitted by TROA.

The Kaskaskia Watershed's management approach relies on forging healthy partnerships among the Corps of Engineers and diverse watershed interests, ultimately creating and sustaining comprehensive solutions within the watershed. The approach serves as a potential model approach for other watersheds. The Kaskaskia River and its watershed cover more than ten percent of Illinois. The Kaskaskia River rises in Champaign County, Illinois, and flows southwest for approximately 310 miles, where it empties into the Mississippi River above Chester, Illinois, approximately 118 miles north of the Ohio River mouth. The basin includes two multi-purpose reservoirs (Carlyle Lake and Lake Shelbyville) and a multi-purpose Navigation Project on the lower end. The multi-purpose reservoir projects—along with the navigation project—have become renowned for recreational opportunity. Additionally, the reservoirs are efficient, reliable providers of flood damage reduction. In 1998, the Kaskaskia River was nominated for the American Heritage Rivers program. This development garnered attention, and highlighted the potential for future water use conflicts and problems. The need for maintenance of navigation channels and a desire to generate interest in recreation rallied some existing groups comprised of business personnel, agency personnel and interested private parties. Their collective effort formed the Kaskaskia Watershed Association (KWA). Concerned about the watershed's condition and immediate future, the KWA, working as a consensus-driven board of directors, protects the watershed and balances navigation, recreation, water supply, conservation, sediment management and other diverse interests. The Corps of Engineers is one of six federal agencies involved with the KWA. Other KWA partners include seven state agencies, fourteen regional interests and many local interests spread across the four reaches of the watershed. The formation of the KWA has resulted in increased collaboration among stakeholders and agencies, providing a forum for grassroots partners to address watershed issues.

Lake Okeechobee is the second largest freshwater lake in the conterminous United States of America and is the heart of the water resources system in south Florida. The Herbert Hoover dike and several water control structures allow management of the Lake to meet different objectives, including flood control, water supply, and environmental enhancement. The regulation of the Lake water levels is performed by the U.S. Army Corps of Engineers (USACE) in consultation with the South Florida Water Management District (SFWMD). Flood control releases from Lake Okeechobee are made through the Caloosahatchee River, through the St. Lucie Canal, and southward to the Everglades. Since the early part of the 20^th century and until the middle of 2000, the Lake was operated using a variety of calendar-based regulation schedules. During the 1990's, the SFWMD and USACE conducted a study to develop and implement a more robust regulation schedule. The Water Supply and Environment (WSE) regulation schedule for Lake Okeechobee was created and it became the official operating schedule in July of 2000. Several innovative factors are introduced in WSE: Use of decision trees to guide water managers, use of indicators for current meteorological and hydrologic conditions, inclusion of seasonal and multi-seasonal hydrologic forecasting (outlooks) in the decision trees, consideration of ecological, and environmental conditions along and downstream of the release canals or rivers, etc. While still being essentially a flood release schedule, WSE was designed to improve the balance of the multiple Lake management objectives. This paper describes WSE and its weekly implementation effort. The different components and elements of WSE are presented, with the climatic and hydrologic data used in the process. Methods used to produce hydrologic outlooks are described. Additionally, the last six years of Lake Okeechobee operations under WSE are discussed. Some of the problems encountered and temporary solutions are presented.

The Lower Colorado River Multi-Species Conservation Program (LCR MSCP) is a comprehensive conservation effort to ensure long-term compliance (50 year) with the Endangered Species Act (ESA) and California Endangered Species Act (CESA) for operation and maintenance of lower Colorado River dams, river modifications, and water delivery infrastructure. Through the LCR MSCP federal entities have compliance under Section 7 of the ESA and non-federal entities through Section 10 of the ESA and through Section 2081 of the CESA, where applicable. Twenty-six species are covered by the Program, including the federally listed Yuma clapper rail (Rallus longirostris yumanensis), Southwest willow flycatcher (Empidonax traillii extimus), desert tortoise (Gopherus agassizii), bonytail (Gila elegans), humpback chub (Gila cypha), and razorback sucker (Xyrauchen texanus). The conservation plan includes: establishment of a $25m fund to support projects to maintain existing habitat values; creation of 5, 940 acres (2403 ha) of cottonwood-willow habitat, 1320 acres (534 ha) of mesquite habitat, 512 acres (207 ha) of marsh, and 360 acres (146 ha) of backwater to provide habitat for covered species. In addition, 660,000 razorback sucker and 640,000 bonytail chub will be released over a 50-year period. The program structure of the LCR MSCP is based on research, monitoring and adaptive management in order to efficiently and effectively develop techniques to accomplish habitat objectives. Implementation of the LCR MSCP will be through the Bureau of Reclamation (Reclamation) with oversight by a Steering Committee that includes permittees and other stakeholders, including a Federal Participant Group, an Arizona Participant Group, a California Participant Group, a Nevada Participant Group, a Native American Participant Group, a Conservation Participant Group, and an Other Interested Parties Participant Group. The Steering Committee operates by consensus, with a dispute resolution process, through a Funding and Management Agreement (FMA)(LCR MSCP 2005). Through the FMA, funding is 50% from the federal government and 50% from the States of Arizona, California, and Nevada. The state share is split 50% California, 25% Arizona, and 25% Nevada.

An error analysis was conducted on an updated San Joaquin River (SJR) representation in CalSim II, a surface water management model of federal, state, and local operations in California's Central Valley. The first stage involves sensitivity analysis of output response to single-parameter adjustment. Results provide information for strategizing model improvements and reveal which inputs are most influential. The latter results feed a second stage involving Monte Carlo uncertainty analysis of how outputs respond to joint changes among the influential inputs. Key challenges involve computational load and data handling across multiple applications in multiple formats, necessitating automation development. This paper discusses project methods. Results will be presented at the conference.

Water management in California is increasingly driven by local water agencies which have designed integrated portfolios of water supply and demand management activities to economically provide reliable water deliveries. Often these portfolios involve other agencies through water market and conjunctive use schemes. Efforts of some major urban, agricultural, and environmental water suppliers are reviewed. Some analytical, data management, and public policy implications of these locally-driven statewide water management portfolios are discussed. In contrast with the traditional centrally-engineered water projects, California has become a system of decentralized engineering and coordination of spatially extensive and integrated water management. Some successes and problems are discussed.

Contemporary reservoir systems often require operators to meet a variety of objectives, which frequently complicate water management decisions. In addition, many reservoir objectives have non-linear relationships and are difficult to implement using traditional optimization techniques. A practical application of multi-objective optimization is developed for Folsom Reservoir near Sacramento, California where water delivery, hydropower and downstream temperature control are desired. Fishery habitat is maintained in the summer and early fall by regulating river temperatures downstream of Folsom reservoir, at a potential cost to hydropower generation. In this application, optimal seasonal reservoir release decisions are found using a multi-objective evolutionary algorithm and a one-dimensional hydrodynamic reservoir temperature model. Seasonal reservoir release policies are examined to evaluate tradeoffs between objectives.

HEC-PRM ("Prescriptive Reservoir Model") is a generalized computer program that performs deterministic network flow optimization of multireservoir systems. PRM "prescribes" optimal values of flow and storage over time by minimizing penalty functions at selected locations in the water resource network. Penalty functions associate a penalty or reward (benefit function) with designated levels of flow or storage. Because a given location (flow reach or reservoir) can have several competing objectives, a separate penalty curve for each objective can be included in the model. Optimization with a Network Row Programming with gains algorithm enables relatively fast runs times for large systems but restricts constraints to capacity limits and mass balance at nodes. PRM has been applied to several large multi-reservoir river systems, such as the Columbia system, the Missouri system and the entire state-wide water resource network of California. HEC-ResPRM is new graphical user interface for PRM that has been developed to facilitate the implementation of large multi-objective reservoir operation studies. The GUI is in large part shared with HEC-ResSim — a multi-reservoir simulation modeling system. The integration of PRM into the HEC-RES modeling platform was made to facilitate the joint development and use of simulation and optimization models. The multi-objective nature of water resource problems is considered and any number of penalty functions with differing units can be added at any network location. A penalty-weight editor facilitates building trade-off curves between objectives by allowing the user to easily change the weight of each penalty type for all locations. Selected preliminary results of a multi-objective reservoir operation study of the Upper Mississippi Reservoir System are presented. This optimization study is part of a "shared vision" planning study, which uses a simulation model to refine the new operation policies suggested by use of the optimization model.

The recent collapse of the Georgia-Florida-Alabama water compact negotiations in the Apalachicola-Chattahoochee-Flint (ACF) and Alabama-Coosa-Tallapoosa (ACT) River Basins discloses significant shortcomings in traditional approaches to resolution of transjurisdictional water conflicts involving multiple competing objectives. Without full recognition of the broader external issues fueling the conflict and without collaboration by the parties to eliminate extraneous sources of intractability, the core dispute is unlikely to be correctly framed, and the negotiations may be ill-informed as a consequence. Poor framing may prevent consensus on core objectives and constraints and misdirect the formulation, analysis and evaluation of water management alternatives. Consensus will be elusive, the diligent efforts of the parties notwithstanding. At the core of the ACT/ACF negotiations were satisficing models for simulation of operational alternatives, which, while sophisticated, addressed primarily symptoms, e.g. flow deliveries, water consumption, reservoir operations, drought response, etc., as opposed to causes of the conflict. The eventual collapse of the negotiations implies that incomplete characterization of the parties, issues, social system, and processes framing the conflict limits the utility of planning decision support systems, and more importantly obscures potentially satisfactory solutions around which consensus is possible. The author proposes a framework for water conflict management designed to systematically disclose and ameliorate many of the most common causes of failure. The procedure consists of the following four steps: (1) identification of sources of intractability in the parties, issues, social system, and negotiation process, (2) re-framing of conflict (parties, issues, social system, process) to eliminate or minimize sources of intractability, (3) achievement of consensus on core problem definition, objectives and constraints, and (4) formulation and parameterization of satisficing models for plan formulation, analysis and evaluation; application of models and consensus-building on management alternatives.

The goal of the rainfall-based management plan is to improve the amount, timing and distribution of flow to the Shark River Slough, the main waterway of the Everglades National Park, South Florida. An essential component of this plan is a Rainfall-Driven Formula, RDF, which provides flow target estimates in response to recent and current weather conditions. In this study, an improvement to the existing plan is provided by introducing a new RDF to predict a more natural flow target and hence delivery. An Artificial Neural Network with one hidden layer and one output layer was trained using a concise set of rainfall, evapotranspiration, and flow data. A South Florida Water Management Model (SFWMM) simulation is used to demonstrate the applicability of the developed formula to manage the water resource infrastructure proposed by the USACE's Combined Structural and Operational Plan (CSOP).

The California Department of Water Resources (DWR) is completing construction of a new reservoir east of the bifurcation of the East Branch and West Branch of the California Aqueduct. Tehachapi East Afterbay provides afterbay storage for the "Valley String" pumping plants (Dos Amigos, Buena Vista, Teerink, Chrisman, and A.D. Edmonston) — the largest power load in California. The project consists of a 910 AF reservoir, 3,150 cfs inlet channel, 3,150 cfs outlet channel, 3,150 cfs bypass and appurtenant structures. The reservoir pool was constructed completely in excavation and lined with compacted soil and hydraulic asphalt concrete. Flexible operational storage enables DWR to reduce pumping operations of the Valley String Pumping Plants during peak energy demand periods; provide ancillary services; and provide other benefits of operational flexibility. The majority of benefits occur early in the project life span, thus DWR collaborated with water customers to expedite planning, design and construction. This process was complicated by competing requirements of early project delivery, rigorous outage requirements, multi-agency coordination and operating criteria of the California Aqueduct.

In transient rain-snow and snowmelt dominant river basins in the western U.S. global climate change is expected to produce systematic changes in winter climate and mountain snowpack resulting in higher flows in winter, earlier spring peak flows, and lower flows in spring and summer. The magnitude of these hydrologic changes is largely dependent on mid-winter temperature regimes, and the impacts will vary with elevation, latitude, and proximity to the ocean. In the context of managed flood control, the associated seasonal streamflow timing shifts are likely to disrupt the balance between flood control objectives and reservoir refill for water supply and other system objectives. To adapt to these hydrologic changes, evacuation of reservoirs for flood control may need to move earlier in the year as warming progresses, and refill may also need to begin earlier. In large rivers such as the Columbia River in the Pacific Northwest, where system-wide flood control storage is distributed between a number of reservoirs in different parts of the basin, rebalancing flood control and refill objectives in response to hydrologic changes poses a significant systems engineering problem. In this study we use the Variable Infiltration Capacity (VIC) macro-scale hydrology model implemented over the Columbia River basin coupled to a monthly time step reservoir optimization model (employing the HEC-PRM solver) to produce dynamic flood rule curves for a group of projects on the Columbia River main stem and tributaries. To test the methods a simple climate change streamflow scenario consistent with projections of warming for the second half of the 21^st century is constructed. The scenario is based on a hydrologic simulation using observed precipitation and detrended historic temperatures consistent with a year 2000 temperature regime, with an additional 2 degrees C added. To calibrate the optimization model for the 20^th century flow regime, the objective function is tuned to reproduce (using the ColSim reservoir simulation model) the current reliability of reservoir refill in the basin while providing comparable levels of flood control (at monthly time scales) to those in the historic record. The objective function is then held fixed, and an ensemble (88 years) of optimal flood rule curves for the future flow scenario are produced. These ensemble results are then summarized by constructing new flood control evacuation and refill schedules as a function of summer streamflow volumes, a framework which is consistent with current operational practice. These procedures provide an objective and well-defined method for rebalancing flood control and refill objectives in complex water resources systems in response to global wanning.

A Truckee-Carson Basin RiverWare modeling system has been developed by the Lahontan Basin Area Office of the United States Bureau of Reclamation. This modeling system does water accounting, forecasting, and simulates operations within the Truckee and Carson basins according to all current basin policy. An overview of the models and the overall modeling process will be presented, and examples of specific operations within the system will be explored. The system consists of two linked simulation models and a comprehensive Hydrologic Database (HDB). The first model is a forecasting model and the second is an accounting/operations model. Each includes a physical model of the basin comprised of objects and links, as well as a ruleset that prescribes the accounting policy, the operational policy and physical constraints on the system. The models run on a daily timestep from the current time through the end of the calendar year. The models were developed in RiverWare. RiverWare's simulation process will be described, and its rule writing interface and language will be demonstrated. Several representative objects from the physical model will be highlighted, and sample rules will be reviewed. The forecasting method will also be discussed. Currently, the accounting/operations model is being substantially revamped to simulate the Truckee River Operating Agreement (TROA). TROA is an innovative, flexible operating agreement between the significant stakeholders in the Truckee-Carson Basin that allows for the establishment and extensive exchanging and trading within the basin reservoirs of many types of credit water. A brief survey of some of the modeling intricacies that TROA requires will be given.

Most newer regional water supply system simulation models employ optimization methods to allocate water and operations according to fixed operational priorities for each time-step. For extensive complex networks with return flows, loops arising from pumping, and proportional delivery reductions for equal-priority deliveries, the assignment of unit weights can be a matter of some art and controversy. A generalized method is developed and implemented to automate assignment of unit weights to guarantee priority-preserving behavior for network flow and linear programming-based simulation models. The method is developed and applied to the CalSim II model of California's State and Federal water supply projects. The automated method should facilitate studies in which major changes to the system are made. The examples also illustrate some paradoxes of water management under prioritized deliveries and return flows, such as occasions when priority optimization implies that some lower priorities receive water when intermediate priorities do not.

The Rapid Prevention of Disputes in Public Policy and Planning (RaPD-3P) Process produced resolution among a large number of stakeholders over resources conflicts in a short period. Resolutions may take decades due to the costs, asymmetric benefits, interests, information asymmetries, scientific uncertainty, rights, and politics. The RaPD-3P process is a multi-disciplinary approach that deals with the nature of conflict and aspects involved in face-to-face discussion to lessen the time consuming barriers. It has a systematic focus on stakeholder interests, provides an ideal consensus environment, and allows real-time evaluation of alternatives so that all interests are not opposed. The process integrates the aspects of modern science and politics with the interests of stakeholder. This process was tested with real stakeholder for a groundwater district in central Texas and being scaled up for large group of stakeholder in central Texas and for international water issues of the Rio Grande. The research results show that the film aspect of the process is much faster and productive the previous method, both in terms of gathering interest, communication, model building. The groundwater stakeholders all found that the film component of the process was conducive for discussion and emotional tensions found in larger group setting were not present. All the stakeholders agreed that process achieved a high level of understanding of interest.

The State of Colorado's Stream Simulation Model (StateMod) is an integral component of the Colorado Decision Support Systems effort. The StateMod model is a generic water resource model capable of simulating stream diversions, instream flow demands, well pumping, reservoir operations and river flows on a monthly or daily basis for any stream system. Fully developed and calibrated StateMod models represent all direct flow and storage water rights and reservoir operations in the following river basins: Upper Colorado River and Gunnison River basins, White River and Yampa River basins, San Juan River and Dolores River basins, and Rio Grande basin in Colorado and the Bear River basin in Wyoming. This paper will provide engineers, planners, practitioners, and the legal community with an overview of the StateMod model and how it can be used to make decisions regarding basin water resources management and analysis of "what if" scenarios. Presented within is an overview of how the model has been applied in practice today and the benefits that have been derived. An overview of the decision logic is also presented. A case study of the Upper Colorado River basin will illustrate various model components including multiple stakeholder interests in ditch systems and reservoirs, and river basin operations related to inter-basin exchanges, hydropower, and irrigation scheduling. A demonstration of on-going model enhancements, including incorporation of a Plan structure to represent reusable supplies, terms and conditions associated with a water transfer, or lagged river depletions from well pumping, will be presented. From simple to complex environments and across a wide scale of basins the publicly-available StateMod model has and is being used to negotiate cooperative settlements, develop water resources policy between multiple entities, and optimize benefits of western water resources. The information presented within will apply on the job to assist attendees in the evaluation of sound water resources planning through a demonstration of real-life applications of a water resource model.

The Secretary of the Interior is directed to negotiate an operating agreement for the Truckee River in accordance with P.L. 101-618, Truckee-Carson-Pyramid Lake Water Rights Settlement Act of 1990. The current draft operating agreement is known as the Truckee River Operating Agreement (TROA). The primary purpose of TROA is to modify operation of Federal and selected non-federal reservoirs in the Truckee River basin located in northeastern California and northwestern Nevada, enhance conditions for threatened and endangered fish species, increase drought protection, improve river water quality, and enhance instream flows. Signatories to TROA are U.S. Department of the Interior, States of Nevada and California, Truckee Meadows Water Authority (Reno and Sparks), and Pyramid Lake Paiute Tribe. This is the first of four papers on modeling and communication efforts for reservoir operations on the Truckee-Carson River systems. This paper will present an overview of TROA and the modeling efforts related to the TROA negotiations over the past 15 years. This paper will also include background information relative to the other three papers on interagency coordination, overview of the current and future modeling system using RiverWare, and the tools being used and developed for stakeholders on the Truckee-Carson River systems.

The Lower Colorado River Authority (LCRA) is a water conservation and reclamation district that operates a series of six lakes on the watershed of the Lower Colorado River in Central Texas to provide water supply and flood control to a 33-county area, including the City of Austin and several rice irrigation districts along the Texas Gulf Coast. In addition, the LCRA produces wholesale power for a 53-county service area and provides water resources for lake recreation activities and in-stream flow maintenance. Currently, the LCRA uses beginning-of-year storage levels to determine the amount of water available to meet demands in the coming year. Seasonal and long-term forecasts are not used by the LCRA for a number of reasons, including high seasonal and annual variability of stream flow, the absence of easily measured hydrologic indicators such as snowpack, and a lack of experience with probabilistic planning methods. In this paper, we illustrate and verify a simple approach for adjusting seasonal water availability forecasts based on hydrologic persistence. Predictor variables include historical monthly streamflow observations and simulated soil moisture content from the NCEP North American Regional Reanalysis. Extensions of the approach are discussed, including generation of probabilistic forecasts and consideration of climate indicators such as the El Nino/Southern Oscillation, Pacific Decadal Oscillation, and North Atlantic Oscillation.

A complete joint utilization for a multi-reservoir system should consider two operating stages: one is to investigate how much water should be released to meet the target demand, and the other is to decide the desirable end-of-period storage for each single reservoir. To simulate the joint operating procedure in a large scale water resource system, a simulation model, named WRASIM-M, was developed by this study based on Network Flow Programming. Two primary advantages from utilizing WRASIM-M are the comprehensive water allocation concept and the computational efficiency. WRASIM-M could be regarded as a simulation tool to help establishing joint operating rules, especially suitable for a large scale and complex multi-reservoir system. The results of case study validated the simulating function of WRASIM-M.

The Contra Costa Water District (CCWD) delivers water exclusively from the river delta (Delta) formed by the confluence of the Sacramento and San Joaquin Rivers prior to discharge into the Pacific Ocean. Since source water quality within the Delta varies significantly with season and tide, CCWD employs a multi-faceted operating strategy to optimize cost-effectiveness given a number of operational and water quality related constraints. Two key operational tools are used to achieve this objective: Los Vaqueros Reservoir (LV) and a sophisticated salinity-outflow model (G-Model), which is based on a numerical simulation paradigm developed by in-house staff. Tactical management is also critically dependent upon coordination, not only between key staff members in different departments within CCWD, but also between CCWD and several State and Federal agencies. This strategy allows CCWD to provide a reliable supply of high quality water at the lowest cost possible, in an environmentally responsible manner.

The state of Florida and the U.S. Army Corps of Engineers are designing and building massive water resource infrastructure to store excess water for the restoration of the Everglades and to provide for increasing water supply needs of southern Florida. With each phase of implementation of new storage areas there is a need to revise the operating rules for Lake Okeechobee. Lake Okeechobee is the second largest freshwater lake located wholly-within the continental United States. Previous operating rules have been developed with the South Florida Water Management Model (SFWMM). The SFWMM is a regional-scale computer model that simulates the hydrology and the management of the water resources system from Lake Okeechobee to Florida Bay. Although the SFWMM is the best available tool for performing a comprehensive evaluation, it is not suitable for quickly testing a broad range of ideas for operating the lake. The rapidly growing population of south Florida has led to an increase in stakeholders and a corresponding need to educate them about the capabilities and constraints of the water control system. Some of the stakeholders also have their own ideas for managing the lake that they would like to see tested with the SFWMM. However because the SFWMM is a large and complex model, it cannot be used to effectively test a large number of varied operating strategies. The need exists for a screening model that provides immediate feedback to analysts and stakeholders. The increasing utility and computational power of the Microsoft Excel spreadsheet software made it a logical platform for building a new tool known as the Lake Okeechobee Operations Screening (LOOPS) model. The LOOPS model is a simple model of the hydrology and operations of Lake Okeechobee and its primary outlets. Analysts can use LOOPS to easily test a broad variety of operating strategies and receive instant feedback showing the performance for the primary lake-management objectives. This paper describes the Lake Okeechobee Operations Screening Model structure and some of its capabilities for screening alternative operating schedules.

A modeling and decision support system is currently being developed for the Truckee and Carson River Basins to produce regulated streamflow forecasts, maintain an accounting of water storage categories, and implement proposed changes to basin operating policy. The system will be used to assist basin stakeholders with the coordination of reservoir operations, and is described in detail in related papers presented at this conference. Some of the key elements to the system are the tools used to present the model results to water right owners and stakeholders in the basins. A variety of presentation and communication tools are currently being used to disseminate model results for use by these stakeholders at monthly meetings during the runoff season. These tools are evolving into more innovative communication tools that will assist with the increasingly complicated reservoir operations that will exist in the future under a proposed new basin operating policy. This paper will focus on the communication and presentation tools currently used with the model results to help manage and coordinate operations, and will show some of the tools in development for future use with new basin policy. Experimental tools involving risk and probability associated with model results will be presented, as well as a discussion of the benefits and problems associated with direct model use by stakeholders.

In this study, a fuzzy inference system is developed for reservoir inflow forecasting and reservoir operation. The system consists of two models. In the first model, the seasonal river stream-flow is forecasted with a fuzzy rule based system. The southern oscillated index, rain, snow, and stream-flow are inputs of the model and the seasonal stream-flow is its output. The second model is a reservoir operation model based on an "If-Then" principle, where "If" is a vector of fuzzy premises and "Then" is fuzzy consequences. The reservoir storage capacity, inflow, demand, and year condition factor are used as the premises and monthly release is taken as the consequence. As a case study, the Zayandeh-Rood Reservoir in Iran is studied. To evaluate the performance of the operation model, different performance criteria such as reliability, resiliency, and vulnerability are calculated. Results indicate that use of this method in extracting knowledge from an informative data set having ill-defined and highly nonlinear structures would be helpful and have advantages over traditional operation methods such as standard operating policy or ordinary least-squared regression rules constructed based on the results of optimization models.

Launched in July 2002, the Sustainable Rivers Project (SRP) is a nation-wide partnership between the U.S. Army Corps of Engineers and The Nature Conservancy to improve the integrity and life of rivers by changing the operations of Corps dams. The SRP is working towards this goal through a combination of demonstration projects, training, staff exchanges, and software development, all of which are designed to support innovative water management practices. The SRP currently involves work on nine river systems across the country, with two more under consideration. The nine river systems include the West in Vermont; the Ashuelot in New Hampshire; the Roanoke in North Carolina and Virginia; the Savannah in Georgia and South Carolina; the Big Cypress in Texas and Louisiana; the White, Black, and Little Red Rivers in Arkansas and Missouri; the Green in Kentucky; the Bill Williams in Arizona; and the Willamette in Oregon. This paper introduces the SRP, discussed its history, successes, and future plans. It is written and presented as part of an SRP session which also details ongoing efforts at two of the SRP sites — the Green and Bill Williams Rivers (in these proceedings, see Turner and Byron, and Evelyn and Hautzinger, respectively).

The TDG loading in the Snake River downstream of Lower Monumental Dam depends upon the spillway discharge and pattern, powerhouse operation, and TDG saturation in the river just upstream from the dam. In the past, spill was generally distributed uniformly across 6 to 8 spill bays during voluntary spillway operations for fish passage. During the 2004 fish passage season, an alternative spill policy at Lower Monumental Dam was tested that concentrated spill through only 2 or 3 spill bays. Total dissolved gas (TDG) measurements were made at the Lower Monumental Dam to determine the effects of two operating scenarios. The data showed a consistent pattern of increased TDG levels for the bulk spill pattern compared to the normal spill pattern based on the total spill discharge. The mass loading of TDG, due to bulk spill operation, was also greater than the mass loading for the normal spill pattern. Predictive equations were developed for TDG production with bulk spill and normal spill patterns. These equations are incorporated into a spreadsheet model "SYSTDG" that can be used to estimate TDG in the Ice Harbor forebay.

Operation planning for multiple reservoir systems is a complex and challenging problem because of inherent uncertainties in inflow forecasts. Long-term inflow predictions are required for mid-term planning. However the practice of using several inflow predictions does not fully reflect the various characteristics of the decision in uncertainty analysis, such as non-anticipating decisions or the serial correlations imbedded in the inflow. We therefore applied the stochastic linear programming (SLP) approach to tackle the uncertainties that are inherent in reservoir operation planning due to the inflow uncertainty. A SLP model is developed for coordinating the multi-reservoir operation to determine the efficient monthly target reservoir storage. The model is formulated as a multi-period, two-stage SLP based on the form of "fan of individual scenarios." The inflow scenarios are generated by the multivariate periodic AR(1) model considering the serial and spatial correlations. The model becomes a large-scale, linear programming model (comprising over 120,000 columns and 80,000 rows), which was solved very quickly (in 5 minutes) using the linear programming solver CPLEX. The expected benefit of the stochastic model was analyzed quantitatively based on value of information measure. The results indicated that the solutions of the stochastic model are much more effective than those of the deterministic model with average inflows, and that this effectiveness is also maintained in real-time operation in the presence of uncertainty. The benefit of applying this stochastic model to the Nakdong River basin in Korea was presented. It implies that the use of the stochastic model in real-time operation is more effective in the presence of uncertainty.

This paper focuses on developing sustainable water management strategies for Charlottesville, Virginia and the surrounding Rivanna River watershed and how they are being facilitated through risk management. Risk management is a means by which water systems can accurately quantify and manage risk using inflow forecasts, thus improving sytem reliability while offering potentially large cost savings through more intensive use of existing supplies and water conservation. The paper describes the advantages of risk management by highlighting the limitations of the safe yield approach and how the former is necessary for improved drought management and capacity planning. It also briefly discusses the basis of the instream flow needs and how risk management is applied to best meet them.

This study proposes a methodology that can update a monthly stochastic optimization model in the middle of a month as a new forecast is available and examines the value of this updating in monthly reservoir operations during a 9-month drawdown period. The proposed optimization model is called SSDP/ESP that uses Sampling Stochastic Dynamic Programming with Ensemble Streamflow Prediction. A monthly ESP forecast is updated in the middle of a month using (1) the observed meteorological (single) series from the first day of the month to the forecast day and (2) the historical meteorological (multiple) scenarios from the forecast day to the end of the month. SSDP/ESP runs and updates their policy whenever the ESP forecast is updated. Applied to the Geum river basin in Korea where two reservoirs are located in series and operated primarily for water supply, the proposed model is compared with other optimization models through 2100 cross validation simulation runs. The simulation results show that the proposed model reduces the annual water shortage by 1 million m³ ( = 4.8% of the total shortage) on average and approaches 97% of the ideal model that assumes perfect foresights if the model is updated three times a month. This study also shows the water shortage reduction between a well-forecasted and a poorly-forecasted years can be up to 20% of the total water shortage in the poorly-forecasted year even if a similar amount of the annual inflow is available in the two years.

Multiple reservoir management requires the consideration of competing demands for the water for irrigation, domestic consumption, hydroelectric generation, recreation, transportation, and ecosystem management. Such management requires sophisticated tools along with well educated personnel. Many point to the need for universities and professional societies to partner to assure that newly educated professionals will be available. This paper described water resources planning and management curriculum that incorporates the river basin modeling package RiverWare. RiverWare is a flexible general river basin modeling tool that allows water resources engineers to both simulate and optimize the management of multipurpose reservoir systems for daily operations as well as for planning studies. The curriculum was developed by faculty and staff at Humboldt State University and University Colorado Boulder. In the fall of 2005, ten Environmental Resources Engineering seniors at Humboldt State University experienced the RiverWare curriculum in ENGR 445, Water Resources Planning and Management. RiverWare assignments were developed for the following topics: storage yield analysis, firm yield/firm water, storage allocation zones, rule curves, operation rules, operation of multi-purpose reservoirs, and hydropower concepts. These assignments are further described as well as students' responses to the RiverWare tutorials and assignments. By the end of the semester, students had enough expertise to develop their own models. One student developed a rule curves model of the Shasta and Keswick reservoirs in Northern California. Finally, the paper describes further planned curricular developments. This curriculum is available at no cost for faculty that want to teach Water Resources Planning and Management using RiverWare. The authors anticipate receiving valuable feedback on the content of this curriculum from conference attendees.