Statistical Modeling of 15-Min Changes of Production from Renewable Sources

In this paper, 15-min production data from renewable energy sources (RES), aggregated by technology (onshore wind, offshore wind, solar) and by country (Germany, Austria, Hungary), are analyzed. The concept of a confidence interval is introduced as a parameter for practical use in power-system management. In probabilistic dimensioning of the FRR (Frequency Restoration Reserve; sometimes also called “secondary reserve”), it is necessary to ensure reserve sufficiency for a high percentage of time p, in the order of 99.9%. The confidence interval is specified by upper and lower deviation limits, expressed as percentages of the total installed capacity of the observed RES system, not to be exceeded with a probability greater than $1-p$. The concept of the “regulation multiplier” is also considered, which essentially indicates how many additional megawatts of RES capacity can be installed for each added megawatt of FRR capacity, ceteris paribus. Finally, a previously experimentally developed regulation-multiplier model is verified by replicating the original research on a new dataset used in this paper.