Short-range Inertia Prediction for Power Networks with Penetration of RES

Low synchronous generator inertia is a challenge faced by modern networks with high penetration of renewable energy sources (RESs). The increased level of RESs into networks is leading to variability and unpredictability of system inertia due to the intermittent nature of RESs. As a result of reduced inertia, high rates of change of frequency (RoCoF) are obvious, which further lead to frequency control challenges in case of power imbalances. Consequently, time ahead prediction of network inertia values is essential for planning and operating the network for frequency control. This paper, therefore, presents a short-range inertia prediction method that uses an ARIMA model with improved characteristics. Actual data from the New Zealand network are used to evaluate the applicability of the method based on symmetric mean absolute error (SMAE) metric. The proposed method gives the best minimum cumulative SMAE of 0.151 compared to other two prediction methods with higher cumulative SMAE values.