풍력발전단지 개발 프로젝트의 순익 리스크 범위 예측

Abstract

In Handong on Jeju Island, South Korea, an investigation was carried out which looked at net profit risk ranges in wind farm development. Wind measurement data was collected over a one-year period in Handong, and reference wind data for a fifteen-year period for the same area was collected from a meteorological observatory at Gujwa. The measure-correlate-predict (MCP) method was applied to obtain long-term artificial wind data for Handong, in order to estimate variations in the annual energy production (AEP) and the net present value (NPV) which in turn helped determine the risk ranges. The AEP and the NPV were calculated under the assumption of having installed 15 units of a Vestas 2 MW wind turbine at the near measurement site. Various Probabilities of Exceedance (PoE) levels were predicted for both the AEP and the NPV in order to clarify the range of possible risk ranges. The regression and matrix methods for MCP application were used to the one-year wind data, and variable SMP and weighted average variable Jeju SMP were applied for economic feasibility analysis. The results showed that a maximum NPV deviation was KRW 30,288 (Mil.won) at a probability of exceedance of 75% (P75), which meant there was deviation of 100 million won per year for each wind turbine. Other economic analyses(IRR, B/C ratio, P/P) were also conducted and studied for comparison. A large sensitivity of C.F. and SMP occurred in B/C ratio and NPV. It means the way of estimation for AEP and SMP is very important for obtaining reasonable project economic analysis results. Moreover, the deviation in mean wind speed, the AEP, and the NPV were estimated assuming that the annual average wind speed varies in a cycle of fifteen years. The range of NPV variation was from -16.78% to +26.24% over fifteen years, which was greater than that of the corresponding variations for wind speed and the AEP. Accurate estimation of wind speed is the most essential part for minimizing the risk in wind farm development project.