A Combined Nonlinear Least Squares and Evolutionary Method for Cost Approximation Model of Gas-Fired Power Plants in Indonesia

Abstract

The project cost estimate is one of the crucial stages in the total project's planning activities. It deals with how much the cost would be spent on the project. As the cleanest fossil fuel, natural gas has a strategic role in domestic energy. Therefore, gas-fired power plants (GFPPs), which comprise gas turbine power plant (GTPP), gas engine power plant (GEPP) and combined-cycled power plant (CCPP), are still promising to serve electricity demand. Moreover, the growth of renewable energy is still low. A lack of knowledge of project cost estimates and the database of existing project costs have initiated finding the project cost approximation by a mathematical model. The study aims to present an overview of project cost estimates for several types of GFPPs in Indonesia based on existing contract projects or original equipment manufacturer (OEM) data sets. The cost estimate of GFPP was approximated by a combined nonlinear least square and evolutionary algorithm as a proposed method. The findings of the study show that the approximation function model for each type of GFPP was a nonlinear-power function. The three performance criteria were evaluated. GTCC and CCPP have better performance criteria than GEPP due to more data sets. However, the results can be applied as one reference for budgeting or planning purposes and are not recommended for detailed estimation in procurement or project execution due to the low definition of project.