STUDY OF ELECTRICITY TARIFF OF PHOTO VOLTAIC NO BATTERY USING MACHINE LEARNING PREDICTION METHOD

• Main Author: Prahastono, Iswan
• Format: Dissertations
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/79080
• Institution: Institut Teknologi Bandung

The evolution in the electricity sector have resulted in electricity tariffs not being set using a simple transaction method, where electricity tariffs on the consumer side are determined by the costs incurred by the electricity producer plus profits. The electricity market system is growing due to the emergence of renewable energy connect to the grid system. In Indonesia, electricity tariffs for renewable energy, which in this study are Solar Power Plants (PLTS) or Photovoltaic (PV), are determined by the government without any incentives for PV power producers or for grid operators (PLN) who utilize PV. Based on literature study show that there is still a possibility to conduct a study on electricity tariffs, especially the FIT type for PV generators, one of which is the Dynamic FIT for PV connected to the grid system within a certain period of time. This research formulates Dynamic FIT for PV connected to a grid system that has several power plants as base load, followers and back-up/peakers, and influences the stability of the grid system. . In the future where there are already many PLTS and electricity market mechanisms have been formed, this research can be useful for determining which PLTS will enter the grid system. In this study the authors propose a Dynamic FIT for PV that enters a grid system which contains a Flat FIT element as a base rate calculated using the Long Run Marginal Cost (LRMC) method and there is a dynamic component that arises due to the intermittent nature of PV based on to the prediction of energy generated by PV with machine learning methods. This research use case study in the North Sulawesi and Gorontalo (Sulutgo) Grid Systems in Indonesia that interconnected with PLTS (PV) Likupang. The formulation of this Dynamic FIT in this research requires a PV supply energy prediction, called a PV unit commitment, as a reference for PLN to operate other interconnected power plants. The realization of PV energy will be compared with the PV unit commitment and the differentiation will be calculated based on the proposed formula that proven by simulating using the conditions in the Sulutgo Grid Systems in Indonesia. The result is a comparison of the costs that must be incurred by PLN between the Flat FIT and the formula proposed by the author of the Dynamic FIT. There are 3 types of Flat FIT used, i.e. contractual rates, regulatory rates and Long Run Marginal Cost (LRMC) rates The results of this study indicate that the Dynamic FIT indicates a sharing risk between PV owners and PLN, to maintain system stability due to the intermittancy of PV. The results of this study indicate that the Dynamic FIT tariff provides accurate planning for grid operators (PLN) so that PLTS can enter the grid system without disturbing system stability and provides balanced risk sharing between PV owners and PLN, to maintain system stability due to the intermittent nature of PV. The application of Dynamic FIT from the results of this study can have a positive impact on the development of a more equitable electricity market in the future which accommodates PV growth in particular and other renewable energy which has intermittent properties in general