Role of long-duration energy storage in solar-wind hybrid systems for energy drought mitigation: case study of Polish energy system

This dataset supports the analysis of renewable energy droughts and the role of long-duration energy storage (LDES) in a hypothetical Polish power system based on solar PV and onshore wind turbines (WT), using a 40-year (1980-2019) dataset of renewable energy generation and electrical load. INPUT DATA The input data were structured according to the modelling workflow. First, a set of historical electricity demand data and population-weighted meteorological variables were used to develop and train an ANN-based electricity demand model (Table 1). The model output consists of a long-term hourly electricity demand time series, which serves as an intermediate data product. This generated demand time series is subsequently used as an input to the energy system model, together with time series of solar PV and wind generation (Table 2). Table 1. Datasets used to develop and train an ANN model to generate long-term electricity demand time series Varaibles Unit File name Resolution Temporal coverage Source Population-weighted meteorological variables for the electricity demand model (cloud cover, humidity, irradiance, precipitation, snowfall, snow mass, temperature, wind speed) Various (consistent with the source dataset) Weather_data_1980_2019_pop_wtd.csv hourly 1980-2019 renewables.ninja (based on MERRA-2 reanalysis data) Historical electrical load MW Load_PL_entsoe_2015_2019_ANN_input.csv hourly 2015-2019 Energy-charts (based on ENTSO-E data) Table 2. Nationally aggregated input data used for modeling of the power system based on PV, WT, and LDES Varaibles Unit File name Resolution Temporal coverage Source PV capacity factor - PV_capacity_factor_1980_2019.csv hourly 1980-2019 renewables.ninja (based on MERRA-2 reanalysis data) WT capacity factor - WT_capacity_factor_1980_2019.csv hourly 1980-2019 renewables.ninja (based on MERRA-2 reanalysis data) Load MW Load_PL_1980_2019_ANN_output.csv hourly 1980-2019 ANN-based electricity demand model OUTPUT DATA The time series of power system simulations for a representative scenario is in the Model_output_timeseries.csv file. This representative scenario assumes that: The LDES energy capacity is sufficient to meet 7 days of average national energy demand (3.23 TWh), PV and WT installed capacities are sized to achieve a long-term average generation equal to 120% of electricity demand. The PV-to-WT ratio for long-term average generation is 35:65. The RE_drought_characteristics.rar dataset contains 16 CSV files with simulation results for four scenarios, where PV and WT installed capacities are sized to achieve a long-term average generation equal to 100, 120, 140, and 160% of the mean electricity demand (e.g. RES100 means that PV and WT average generation is equal to 100% of average demand). For each scenario, we provided four files presenting: (1) the number of drought hours per year, defined as hours with positive unserved energy, (2) the annual energy deficit expressed in TWh; (3) the duration of the longest single event in a given year, expressed in hours, (4) the largest energy deficit associated with a single event in a given year, expressed in GWh. In all files, rows correspond to years reported in a summer-to-summer format (e.g. 1980/81), while columns represent the PV-to-WT ratio in electricity generation (e.g. "0.35PV : 0.65WT").