Tool for the Optimal Operation of DVPPs under Uncertainty of Non-dispatchable RES

H. Nemati, A. Ortega, P. Sánchez, L. Sigrist, L. Rouco

Resumen:

In Task 5.1, the core of a novel Renewable Energy Source (RES)-based Dynamic Virtual Power Plant (DVPP) model that optimizes its operation and configuration when participating in European energy and ancillary service markets was proposed. Such a Dynamic Virtual Power Plant (DVPP) model comprises different business cases by properly defining the objective function of the optimization model such that the DVPP selectively participates in one or more of the three marketsconsidered, namelyDay-AheadMarket(DAM), Secondary Reserve Market (SRM) and Intra-Day Markets (IDMs).
One limitation of the DVPP model core proposed in D5.1 is that it assumes that the forecasts of a priory uncertain data, such as market prices and stochastic energy sources (wind speed and solar irradiation), are exact. However, this is an strong assumption that generally does not correspond to reality. In fact, in the majority of cases, the longer the time horizon considered to determine the optimal schedule of the DVPP, the lower the accuracy of such forecasts.
This document presents an update of the DVPP model core, where uncertainties are duly taken into account for determiningtheoptimal participation of the DVPP inthedifferent markets considered. This is achieved by implementing the so-called Robust Optimization, which solves the optimization problem for the worst-case realization of the uncertain range.
Compared to other existing techniques that tackle the uncertainty problem, the Robust Optimization shows significant advantages such as the possibility to define, from the user’s side, the boundaries of the uncertain range based on the strategy tofollow, whether optimistic, neutral, or pessimistic. Other advantages include the relatively simple format of the uncertain input data required by the Robust Optimization algorithm, where only the median and upwards/downwards deviations of the uncertain distributions need to be defined.
This document also describes in detail the tool developed to parameterize, execute and visualize results from the optimization algorithm. The operation and performance of the tool is discussed by means of a comprehensive set of illustrative examples that serve to validate the tool proposed.

Proyecto del IIT: H2020POSYTYF (H2020POSYTYF)

Entidad financiadora: Comisión Europea. H2020

Fecha de publicación: 31-jul-2022