International Journal of Energy Research, Vol.44, No.9, 7525-7547, 2020
Sustainable personnel scheduling supported by an artificial neural network model in a natural gas combined cycle power plant
Despite the impact of personnel scheduling on the sustainability of production processes, it is a rarely studied problem in the electricity generation sector, whose main purpose is sustainable energy generation. From this point of view, personnel scheduling problem is handled in this study based on the impact of this problem, which focuses on fair and balanced job distribution according to the personnel qualifications, on the sustainable generation in the power plants. With its advantages such as ease of operation and maintenance, fast commissioning, lower greenhouse gas emissions compared to other fossil fuels and etc. one of the biggest Natural Gas Combined Cycle Power Plant (NGCCPP) in Turkey is chosen as application area. The main purpose of the study is to demonstrate the effect of fair, balanced and competency-based personnel scheduling on generation stoppages resulting from personnel planning in power plants, which have great importance in energy supply security of the countries. This study is the first in the literature in terms of three different aspects. Firstly, in the energy sector, it is different in that it is a personnel scheduling study that takes into consideration the job assignment of the personnel according to their abilities and minimizes costs. Secondly, the large size of the problem is different from that of previous studies in the literature. Finally, the number of personnel employed in this generation facility varies according to the season. At this point, the past generation data are analyzed using an Artificial Neural Network (ANN) method and an estimation is made, and personnel scheduling is performed in the light of this information. Consequently, the proposed multi-objective mathematical model supported with multi-criteria decision making and ANN, shutdown rate of the power plant due to operator error is reduced 67.3%, and personnel satisfaction level is increased from 42% to 89%as solution.