Exploratory Data Analysis for the Techno-Economic Impact of Renewable Energy and Dynamic Line Rating on Grid Reliability

Between 2015 and 2018, hourly generation, demand, price and weather data for Spain were collected from Kaggle. This dataset encompassed various sources of generation, including fossil fuels such as coal, gas, and oil, as well as renewable sources like wind, solar, hydro, and biomass. Analyzing this dataset, we aimed to uncover the key factors influencing electricity prices in Spain. To achieve this, we employed the Maximum Relevance-Minimum Redundancy Algorithm to identify the most significant features for predicting electricity prices in 2018. Subsequently, we explored the effectiveness of three machine learning methods - Decision Tree, Support Vector Machine (SVM), and Ensemble - in forecasting electricity prices. Evaluation of model performance was conducted using mean absolute error (MAE), mean squared error (MSE), root mean squared error (RMSE), and the coefficient of determination (R2). Our findings revealed that the Ensemble model outperformed Decision Trees and SVM, exhibiting an MAE of 3.07, MSE of 21.47, RMSE of 4.63, and an R2 of 0.90. This suggests a strong fit between predicted and observed electricity prices. Moving forward, future research will delve into exploring the techno-economic impact of Dynamic Line Rating (DLR) on the reliability of electricity supply.