OPTIMIZATION OF COAL PROCUREMENT NETWORK USING COAL BLENDING METHOD TO MEET THE NEEDS OF POWER PLANT (PLTU)

• Main Author: Imaduddin, Muhammad
• Format: Theses
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/36209
• Institution: Institut Teknologi Bandung

Cost of procuring coal is the biggest variable that can affect the economic value of a power plant, so optimization needs to be done. One method that can be used is by mixing coal (coal blending) and compiling a coal distribution network. This problem can be solved within the scope of the location routing problem (LRP) study. Research on determining the location of CBF (coal blending facility), currently does not take into account the influence of coal distribution routes. While the existing research on LRP (location routing problem) can not be directly applied to this problem due to the application of the coal blending method. Therefore, a new LRP model is needed to facilitate the problem. This research develops the LRP model for coal procurement network problems that use the coal blending method which aims to determine the location of the CBF (strategic problems) and the coal distribution routes (operational problems) simultaneously. Both types of decisions influence each other, by optimizing it simultaneously, the coal procurement costs can be obtained optimally. This optimization problem is formulated in the form of MILP, by considering multiple supply allocations, capacitated facilities, and fixed costs. The LRP model (integrated model) that has been compiled in this study is only able to solve small-scale data because it is included in the NP-hard category. The problem with large-scale data need to be solved by decomposition method using the network design model. The model is used to reduce the number of nodes involved, then processed using the LRP model (integrated model), so that the duration of the processing data becomes faster. Both models have been verified. Based on the results of small-scale hypothetical data processing, the solutions obtained from this method has approached the results of data processing using the optimal method. The value of the objective function obtained has a gap of < 1%. The model is tested on an example of the optimization problem of the PLTU coal procurement network involving 19 potential suppliers, 4 potential CBF, and 7 PLTU. The problem can be solved using this method with the duration of data processing is 913.02 seconds.