Investıgatıon of the Impact of the Solar Power Generation Forecast By Using Big Data Analytics On the Local Electrıcity Market
Özet
Integration of Distributed Renewable Energy Sources (RES) into the existing energy system becomes more challenging as the number of RES increases due to their intermittent and variable nature. One way to address this issue is to use Local Electricity Markets (LEM) where consumers and producers can actively participate in trading locally produced electricity within their own Local Energy Communities (LEC). However, knowing the production value in advance (usually for a short period) is crucial for the formation of prices, evaluation of bids, and creating offers in local energy markets. Therefore, short-term load forecasting, which is an important parameter that helps electricity grid operators make decisions such as purchasing and selling electricity, load balancing, and maintenance planning, plays a significant role in system operations.
The aim of this study is to examine the possible role of a solar power plant whose short-term production value is estimated in advance through simulation in the local electricity day-ahead market and the effects it may have on electricity prices. Additionally, this study aims to pre-shape pricing by obtaining bids before the day electricity will be supplied in this market.
In the first stage of the study, a high-capacity solar power plant was selected, and day-ahead electricity generation was estimated for this plant using past electricity production data and meteorological data from the plant's region. Due to the high variety and volume of the data, Big Data Analytics method was used in this analysis, and the analysis was carried out using machine learning techniques in Python programming. Three different models were examined, and the Light GBM model provided the best result for the day's electricity generation estimation In the second stage of the study, the forecasted electricity generation values for the modelled day were used in the local electricity market simulation model. Grid Singularity, an open-source and online software, was used to verify simulated scalable scenarios and evaluate LEMs economically. Firstly, a community was identified under Grid Singularity, and local market players were added for this community. Then, three different scenarios were developed to examine price formation, profitability, and the community's self-sufficiency thoroughly. In the first scenario, a solar power plant was not included in the community, and local market players were forced to meet all their electricity needs from the grid. In the second scenario, a solar power plant with high installed capacity was added to the system, and the simulation was run in this way. Finally, in the third scenario, two batteries with separate capacities of 10 kWh and 30 kWh were added to the system, unlike the second scenario, and the simulation was run again. In situations where solar energy could not be provided, local consumers purchased electricity from the battery, and it was observed that this increased the self-sufficiency of the community. When the results of all scenarios were evaluated, self-sufficiency rates were obtained as 0%, 65.0%, 69.0% & 77.0% (by depending on the battery power) respectively. The values indicates that the community can utilize the green electricity generated in the local market at the stated percentages. However, achieving these percentages fully is not possible due to the fact that solar energy is the primary renewable energy source in the community, and the production of the solar power plant is subject to fluctuations in meteorological values throughout the day. Moreover, it was achieved that penetration of substantial quantity of renewable energy into the system resulted in a decrease of 26.7% and 30% in the average market price of electricity in the second and third scenarios, respectively, as compared to the first scenario. As a result, it has been observed that the integration of a high-capacity solar power plant into the local electricity market lowers market prices. Additionally, it has been emphasized that knowing the production that this plant will generate one day in advance allows market participants to take effective positions in the market.
