Detection of Critical Events in Renewable Energy Production Time Series

TL;DR

The paper addresses the challenge of assessing grid reliability under increasing renewable variability by applying the Maximally Divergent Intervals (MDI) outlier-detection method to ERA5-derived wind and solar production time series. It compares divergence measures (Cross Entropy and unbiased KL) and demonstrates how MDI can identify high-impact, weather-driven intervals without predefining events, while revealing no clear historic trend in outlier intensity. The work provides practical tuning guidance, shows the added value of multivariate analysis over univariate approaches, and discusses how detected outliers can inform scenario-based power system simulations and future climate assessments. The approach offers a scalable pathway to incorporate extreme-weather conditions into reliability analyses and supports planning for climate-aware grid investments.

Abstract

The introduction of more renewable energy sources into the energy system increases the variability and weather dependence of electricity generation. Power system simulations are used to assess the adequacy and reliability of the electricity grid over decades, but often become computational intractable for such long simulation periods with high technical detail. To alleviate this computational burden, we investigate the use of outlier detection algorithms to find periods of extreme renewable energy generation which enables detailed modelling of the performance of power systems under these circumstances. Specifically, we apply the Maximum Divergent Intervals (MDI) algorithm to power generation time series that have been derived from ERA5 historical climate reanalysis covering the period from 1950 through 2019. By applying the MDI algorithm on these time series, we identified intervals of extreme low and high energy production. To determine the outlierness of an interval different divergence measures can be used. Where the cross-entropy measure results in shorter and strongly peaking outliers, the unbiased Kullback-Leibler divergence tends to detect longer and more persistent intervals. These intervals are regarded as potential risks for the electricity grid by domain experts, showcasing the capability of the MDI algorithm to detect critical events in these time series. For the historical period analysed, we found no trend in outlier intensity, or shift and lengthening of the outliers that could be attributed to climate change. By applying MDI on climate model output, power system modellers can investigate the adequacy and possible changes of risk for the current and future electricity grid under a wider range of scenarios.

Figures (10)

• Figure 1: The autocorrelation of variables at different time lags, using the Yule-Walker method with sample size adjustment.
• Figure 2: Figures depicting the outlier with the highest score using the Cross Entropy measure. The top figures show the generation of each technology and the temporal context in which the outlier (indicated by red lines) was found. The bottom images provide histograms of the generation (in MWh) of each of the three technologies during the interval (in their respective colour) and the remaining data (in purple).
• Figure 3: Boxplot of the average hourly Total Energy Generation during the top 50 outlier events per decade based on the Cross Entropy measure.
• Figure 4: Figures depicting the outlier with the highest score using the unbiased Kullback-Leibler divergence measure. As shown in Figure \ref{['fig:Full-region-CE_norm_1']}.
• Figure S1: Histograms of the Solar Photovoltaic, Wind-Onshore and Wind-Offshore energy generation time series, plotted together with a fitted Normal Distribution and a fitted KDE using Gaussian Kernels.