Optimization of selective withdrawal systems in hydropower reservoir considering water quality and quantity aspects

Highlights

• Adaptive surrogate based simulation–optimization framework is used in this study.

• The optimal selective withdrawal system (SWS) in water reservoir is designed.

• Optimal reservoir operation rule in SWS is derived in this research.

• The hydropower energy and water quality issues are focused in SWS operation rule.

• Parametric reservoir operation rules in SWS are derived focusing on sustainability.

Abstract

The selective withdrawal system (SWS) has often been used to alleviate downstream temperature pollutions and in-reservoir unacceptable oxygen concentrations whilst increasing hydropower energy generation. The aim of this study is to a) design the optimal locations of intakes in SWS, and, b) derive optimal parameterized selective withdrawal operation rules. The integration of water quality and quantity aspects to meet long-term environmental services and economic benefits is challenging. The main challenge lies in coupling search-based optimization algorithms with a numerical hydrodynamic and water quality simulation model (i.e., CE-QUAL-W2). To cope with the computational burdens of CE-QUAL-W2, surrogate models have been developed to represent the dynamics of temperature and water quality according to various SWSs. The surrogate models and the hydropower energy computation module were coupled with a multi-objective particle swarm optimization algorithm in an adaptive surrogate-based simulation–optimization framework (ASBSOF). The ASBSOF is applied in Karkheh reservoir, Andimeshk, Khuzestan, Iran, to address problems a and b. The results indicate that the optimal SWSs provide the advantage of sustainable management to balance the energy generation whilst minimizing environmental damages.

Introduction

Water reservoirs are the main elements of water supply systems that serve multiple functions, including municipal water supply, hydropower generation, irrigation, and navigation. Nowadays, many reservoirs worldwide are suffering from excess nutrient inflows and water quality deterioration. These human-made water infrastructures affect the river ecological environment in many ways, including changes in the cycles and rhythms of river natural flow and alteration of thermal and water quality patterns in the water bodies. The aforementioned processes, have generally been threatening water security on both local and global scales. Under the risk of water scarcity, appropriate design and operation practices emphasizing the environmental aspects are crucial in the efficient management of reservoirs as valuable freshwater resources. These new and modern reservoir management practices include a new paradigm which tries to balance economic and social objectives against water quality conservation requirements.

Water reservoirs could be equipped with selective withdrawal structures, consisting of multiple intakes drawing water from multiple depths. In this technique, multiple intakes could be opened simultaneously to mix water from various elevations to help achieve in-reservoir and downstream temperature and other water quality objectives. The design of SWS (vertical locations and withdrawal ratio) and adjustment of withdrawal patterns have been known as new alternatives to enhance in-reservoir and downstream river water quality. Although the application of SWS could help fulfill desirable water quality conditions in water bodies, it could negatively affect the energy generation objectives in reservoirs with hydropower objectives. Consequently, the design and operation of SWSs in reservoirs should be studied in integrated water resources management with quality and quantity aspects in mind. Saito et al. (2001) studied the effects of installing a temperature control device on phytoplankton production in a dam at Shasta Lake, California. CE-QUAL-W2 model was applied to simulate and predict the impacts of new operation strategies on the lake water quality responses. Gelda and Effler (2007) developed a simulation model framework to evaluate the benefits of multi-level intake configurations to improve reservoir outflow temperature and turbidity. A 2D hydrodynamic and water quality simulation model (WQSM), CE-QUAL-W2, was coupled with a heuristic search algorithm to support the automated determination of intake levels through evaluating the efficacy of various combinations of the number and position of intake. The proposed methodology was applied in Schoharie Reservoir, New York, to enhance the downstream temperature and water quality under the conditions of a critical historic year. The solutions serving as the best intake levels to alleviate the adverse effects of reservoir thermal stratification and turbidity issues do not necessarily guarantee the overall optimum due to the short-time frame in planning. In other words, the long-term hydrological and meteorological variability and different initial storage conditions have to be considered in an optimal design of SWSs. Çalışkan and Elçi (2009) examined the effects of SWS to control released water temperature. A 3D hydrodynamic model (EFDC) for Tahtali Reservoir in Turkey was calibrated and validated to evaluate the effects of SWSs on water mixing and anoxia reduction in the reservoir. Rheinheimer et al. (2015) developed a mathematical linear programming model to determine the water released from different thermal pools in Lake Spaulding. The developed model aimed to mitigate the deviations from target downstream temperatures in South Fork Yuba River in California with climate warming. A dynamic thermal model with simplified assumptions was applied to simulate downstream water temperature. Rounds and Buccola (2015) modified the algorithm in the CE-QUAL-W2 model (3.7) for blending the outflow leaving multiple reservoir outlets (up to 10 outlets) to meet the outflow target temperature on a time series scale. The revised version of the CE-QUAL-W2 model showed the capability to simulate the effects of floating outlets in conjunction with lower fixed-elevation outlets ranking the priority designations and criteria for head and flows into account. The new model features were implemented in Detroit Lake under various environmental conditions. Li and Qiu (2015) developed a simulation–optimization (SO) model to maximize hydropower energy generation and sediment outflow fluxes in Three Gorges Reservoir (TGR), China. In order to manage the mismatch between increased power generation and enlarged sediment scouring, a multi-objective optimization framework was applied to balance the water and sediment content in the reservoir. The TGR was equipped with turbine intake, upper and lower spillways used for power generation, and sediment fluxes. Soleimani et al. (2016) extracted optimal operating rules in selective withdrawal systems considering downstream thermal pollutions as the only objective. LIBSVM model as the surrogate model of CE-QUAL-W2 was coupled with a genetic algorithm to derive monthly operating rules in the selective withdrawal scheme to mitigate reservoir outflow thermal pollutions. He et al. (2017) studied the effects of the temperature-control curtain (TCC) as an effective selective withdrawal facility on both the reservoir outflow temperature and the thermal structure in Sanbanxi Reservoir. CE-QUAL-W2 model of Sanbanxi Reservoir was calibrated to simulate the flow and temperature under different TCC scenarios. Weber et al. (2017) developed a sustainable management model of drinking water reservoir considering water supply and environmental impacts. They derived an optimal withdrawal strategy with the aim of achieving the desirable downstream water temperature and avoiding unacceptable low oxygen concentration within the reservoir. GLM-AED2 model was calibrated and validated to evaluate various withdrawal strategies in Grosse Dhuenn Reservoir to select the optimal scenario. Yu et al. (2018) developed optimal reservoir operation strategies in the TGR system to manage the eutrophication challenges in downstream estuarine and maximize hydropower energy generation. The effects of TCC on reservoir water quality and ecology were studied by He et al. (2019). They considered multiple typical water-retaining proportions (the ratio of the water-retaining area of the TCC to the cross-sectional area of the TCC) and future climatic changes to analyze water quality and ecology. Kim and Choi (2021) studied the effects of the selective withdrawal framework (SWF) on reservoir outflow water temperature and downstream fish habitat in Soyang-gang Dam, Korea. The physical habitat simulations were carried out using habitat suitability curves (HSCs) for adult and spawning Zacco platypus in the study area. The results indicated that using selective withdrawal operations, the effects of water impounding on fish habitat could be reduced. Duka et al. (2021) studied the effects of SWF and upstream reach control through the vertical curtain on the thermal stratification responses of reservoir with 3D WQSM.

As highlighted, most of the recent researches on SWS design and reservoir operation studies has used lumped WQSMs with simplifying assumptions (Rheinheimer et al., 2015, Li and Qiu, 2015). Disregarding the long-term hydrological and meteorological variabilities and the sensitivity to the initial conditions (Gelda and Effler, 2007, Li and Qiu, 2015, Yu et al., 2018), they have either identified the best solutions according to trial and error methods (Saito et al., 2001, Çalışkan and Elçi, 2009, Rounds and Buccola, 2015, He et al., 2017, He et al., 2019, Weber et al., 2017, Duka et al., 2021, Kim and Choi, 2021), and/or focused solely on quality objectives (Gelda and Effler, 2007, Soleimani et al., 2016). Very limited studies on SWSs optimization (Saadatpour et al., 2020) have considered hydropower energy generation and water quality issues under an extensive hydrological period featuring wide ranges of reservoir drawdown and refill. Also, limited researches have focused on SWS design and /or have embedded parametric reservoir operation rules focusing on both quality and quantity aspects in a selective withdrawal scheme. This research attempts to overcome the aforementioned limitations.

In specific, this study introduces an integrated modeling framework for an optimal SWS design followed by the derivation of the optimal operation rules emphasizing the selective withdrawal framework to increase hydropower peak energy generation and enhance in-reservoir water quality and downstream temperature. This study intends to integrate and resolve the challenges associated with applying metaheuristic algorithms and analytical tools, spatial and temporal level of potential water quality and quantity impacts, cognitive complexity of 2D hydrodynamic and WQSM, and adaptive modifications of surrogate models in the emulation processes of waterbody’s responses. The paper is organized as follows. Section 2 elaborates the case study and research problems. Section 3 introduces the applied algorithm and analytical tools, examines how the problem definitions affect the decision vectors, and presents the problem formulations. Section 4 discusses the results and Section 5 presents the concluding remarks of this study.