MIXED-INTEGER LINEAR PROGRAMMING (MILP) APPROACH TO SOLVE DERATING IN OPTIMIZATION OF THERMAL POWER PLANTS OPERATION UNDER PRIMARY ENERGY UNCERTAINTY

• Main Author: Fauziyah, Nur
• Format: Theses
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/70084
• Institution: Institut Teknologi Bandung

Electrical has an in important role in economic development of a region. Various problems often often occur in meeting the needs electrical of electrical energy on of which is derating. Derating is a condition that occurs when the output power (MW) of a unit is lower than the Net Capable Power (DMN), which means that the load demand is not fulfilled by the generator which results in a shortage of supply or loss of supply of electrical energy. This problem is closely related to the uncertainty of the primary energy availability of coal as one of the primary energy sources in thermal power plants. Various factors affecting the availability of primary energy are coal transshipment problems related to inventory and coal quantity problems, while factors affecting coal quality are determining the calorific value of coal according to the needs of the power plants, both from the design of the engine and capacity. One of the methods used is to carry out fuel blending, this process is used to obtain the calorific value of coal as needed with the constraints of moisture, sulfur and ash. Of these various factors, the researchers are only oriented to the boundaries of the operating system, so that other factors that turn out to have a big influence are often ignored. Meanwhile, in Indonesia, as an archipelagic country, most of coal shipments are carried out by sea from suppliers to power plants in each region where most coal suppliers have mines in different locations which affect the calorie quality and price of coal. So that power plants experience problems in determining coal suppliers that suit their needs. The various problems that have been described can lead to non-optimal scheduling of power plants operations and will have a significant impact on the overall system operation resulting in very expensive operational costs. This makes the research basis for the authors to be developed in detail and specifically by combining the problems of optimizing the generation scheduling with the constraints of the operating system, transshipment, fuel blending and inventory of coal by combining them as a problem of optimization of thermal power plants operation under primary energy uncertainty. In this study, the authors used the Mixed-Integer Linear Programming (MILP) approach in solving the derating problems described earlier. The optimization algorithm is implemented using Pyomo software as an open source library based on Python programming, where GNU Linear programming Kit (GLPK) solver is used to solve the linear programming (LP) dan mixed-integr programming (MIP) algorithm. The optimization algorithm include unit commitment, economic dispatch coal transshipment, fuel blending and inventory algorithms. The combination of algorithms results in a new algorithm that functions in optimization of thermal power plants operation to determine the optimal time for power plants to request coal delivery and system maintenance according to the coal supply in storage of power plants. Algorithm applied on IEEE 39-bus system within a one-week periods, so the total of simulation was 168 hours. The addition of this new algorithm provides 5,57% cheaper generation cost than without using the new algorithm and more optimal power plants operation.