Economic Scheduling of Thermal Power Plants in the West Sumatra System Using Genetic Algorithms

Abstrak

Fuel cost is the largest component of operating expenses for thermal power plants; consequently, suboptimal scheduling increases system costs. This study determines a minimum-cost 24-hour generation schedule for the West Sumatra electrical system by optimizing four thermal units—specifically at the Teluk Sirih and Ombilin power plants—while assuming the Muara Labuh geothermal unit operates at a capacity of 85 MW. A genetic algorithm is employed to generate candidate schedules, evaluate total fuel costs and power balance constraints, convert feasible solutions into fitness values, and improve the population through selection, crossover, and mutation. The algorithm uses a population size of 20,000, a crossover probability of 0.9, a mutation probability of 0.1, and 100 generations. The genetic algorithm yields a total 24-hour cost of Rp3,202,331,042.71, representing a 2.9% cost reduction compared to the actual schedule. Compared to the lambda iteration method, the cost is lower by Rp25,365.40, or 0.0008%. These findings demonstrate the effectiveness of the genetic algorithm for economic scheduling in a real-world thermal system, although its numerical advantage over the lambda iteration method in this specific case is relatively small