Applying a Correlation Analysis Method to Long-Term Forecasting of Power Production at Small Hydropower Plants

Gang Li,Chen-Xi Liu,Chun-Tian Cheng,Sheng-Li Liao

doi:10.3390/w7094806

Abstract

Forecasting long-term power production of small hydropower (SHP) plants is of great significance for coordinating with large-medium hydropower (LHP) plants. Accurate forecasting can solve the problems of waste-water and abandoned electricity and ensure the safe operation of the power system. However, it faces a series of challenges, such as lack of sufficient data, uncertainty of power generation, no regularity of a single station and poor forecasting models. It is difficult to establish a forecasting model based on classical and mature prediction models. Therefore, this paper introduces a correlation analysis method for forecasting power production of SHP plants. By analyzing the correlation between SHP and LHP plants, a safe conclusion can be drawn that the power production of SHP plants show similar interval inflow to LHP plants in the same region. So a regression model is developed to forecast power production of SHP plants by using the forecasting inflow values of LHP plants. Taking the SHP plants in Yunnan province as an example, the correlation between SHP and LHP plants in a district or county are analyzed respectively. The results show that this correlation method is feasible. The proposed forecasting method has been successfully applied to forecast long-term power production of SHP plants in the 13 districts of the Yunnan Power Grid. From the results, the rationality, accuracy and generality of this method have been verified.

Highlights

As a generally accepted renewable energy, small hydropower (SHP) has been greatly developed in the past few decades because of its small scale, lower investment, quick returns, lack of pollution and the promotion of local economic development [1–6]
Different from the forecasting problems of large-medium hydropower (LHP), the long-term forecasting of SHP faces a series of challenges: (1) Problems arise from the weakness in management, difficulty of information collection, and lack of available data accumulation; (2) SHP is typically derived from run-off river plants with little regulating capacity, so their power production is determined to a great extent by the reservoir inflow
The SHP plants in Puer are selected as an example

Summary

Introduction

As a generally accepted renewable energy, small hydropower (SHP) has been greatly developed in the past few decades because of its small scale, lower investment, quick returns, lack of pollution and the promotion of local economic development [1–6]. With large-scale SHP plants accessing the power grid, problems of wasting water resources and abandoning electricity have been increasing, and the safe operation of the power grid is threatened [7,8]. It is necessary to forecast the power production of SHP plants to solve the above problems by means of coordination and dispatching of SHP and large-medium hydropower (LHP) plants. Different from the forecasting problems of LHP, the long-term forecasting of SHP faces a series of challenges: (1) Problems arise from the weakness in management, difficulty of information collection, and lack of available data accumulation; (2) SHP is typically derived from run-off river plants with little regulating capacity, so their power production is determined to a great extent by the reservoir inflow. Due to the challenges, forecasting power production of SHP plants is a complex task; a few researchers have explored this issue and have obtained some useful results [22,23]

Methods

Results

Conclusion