On the Importance of using an AC or DC Network Model in the Multi-Period Secure Stochastic Optimal Power Flow for Settling a Multidimensional Day-Ahead Market

Abstract

As the penetration of renewable energy sources increases, their variability and uncertainty have pushed the development of secure stochastic approaches to solving the secure day-ahead and intra-day multi-period optimal power flow. Some of these approaches, in addition to settling the energy market, also procure other products that generators can offer such as spinning reserves and ramping capability, the latter of which becomes even more important under high penetration of renewable sources. The resulting large scale nonlinear minimization problem makes using the more simple DC flow model of the network appealing, at least in the day-ahead stage. This work focuses on the impact of using an AC vs. a DC model of the network in the results of these multi-dimensional markets, using the Colombian system as a test case. The study implies that although the overall energy allocations may not change much from one model to another, other products may exhibit very different allocations under different network models.