Maximization of hydropower generation through the application of a linear programming model

Summary

The linear objective function is examined as an alternative to a nonlinear form with the aim of maximizing hydropower energy generation. The Yongdam multipurpose dam, located on the Geum River in South Korea, is selected as the subject of the model application. Inflow data with a reliability of 50% is applied to the model for an operation period of 12 months. This study analyzes the effect and sensitivity of the model’s release and reservoir storage on the maximization of hydropower energy generation based on calculations of optimal values. The operation according to the optimized policy is shown in terms of the given priority. The ratio between two parameters (releases and reservoir storage) is also examined in the context of the objective function of the linear model. The maximum annual energy production of the proposed model was approximately 184 GWH, which represents 86% of the potential energy production level.

Introduction

Due to an increase in energy demand, fossil fuel resources required for the generation of energy are becoming increasingly scarce. Furthermore, the development of atomic power has been regulated because of safety concerns. Consequently, there has been renewed interest in hydropower as an energy source. In comparison to other energy sources, hydropower is clean. It also requires no fossil fuel, thereby offering economic advantages even though the cost of development is higher than thermal power. Hydropower is also appealing because of its fast start-up time and peak load characteristics. For these reasons, hydropower plants are constructed by remote development in multipurpose dam development projects in South Korea and operated as storage plants.

Supplying water is the primary role of multipurpose dams in South Korea, and their hydroelectric power plants are mostly managed alongside the water supply systems. Performance criteria can be established for the operation and optimal stabilization of hydropower systems. The performance criteria generally include installed capacity, firm power and annual energy production (Kwon, 1994). The objective function for the performance is nonlinear because the products of power generation release, storage level and reservoir operation are of a sequential decision-making nature. Dynamic programming has commonly been used as a method to solve the problem (Yakowitz, 1982, Pereira and Pinto, 1985, Hiew, 1987, McLaughlin and Velasco, 1990, Paudyal et al., 1990, Ko et al., 1992, Kim and Palmer, 1997, Tilmant et al., 2002).

Research has been conducted on the application of successive linear programming (SLP) for the implementation of approximate linearization of the nonlinear function, and has been employed as a method for overcoming the nonlinearity of such objective functions (Grygier and Stedinger, 1985, Tao and Lennox, 1991, Ko et al., 1992, Reznicek and Simonovic, 1992, Barros et al., 2003).

This research applies a linear programming method similar to an SLP model. However, this study differs from the SLP model in that a new linear hydropower objective function is formulated and optimized in the model without the iterative algorithm.