Real-time optimal scheduling for prosumers resilient to regulatory changes

Abstract

The last decade marked an exponential increase in photovoltaic (PV) systems installed on the rooftop of domestic residences within Europe. This situation was basically favored by generous financial schemes such as Feed-in-Tariff and the market of green certificates. However, such governmental incentives drastically reduced, or they were already replaced with netmetering schemes which favor different scenarios, like those based on increase of self-consumption and decrease of grid-back injections. This unstable regulatory environment puts both new and old owners of PV systems under a regulatory financial risk. Recently, a regulatory resilient architecture, called UniRCon, was proposed, to overcome both financial and technical regulation uncertainties, where local battery energy storage system plays a key role. Here we propose a real-time energy management tool that could be used for the daily operation of such UniRCon architectures. The methodology is based on a mixed-integer linear programming energy management tool that considers possible arbitrage benefits due to price difference in the energy purchased from the grid, while explicitly considering the efficiency of the power electronic interfaces (converters) according to the operation point. We prove our approach using a lab-scale experimental setup of a DC residential microgrid. The results are analyzed under realistic operation scenarios derived from one-year load and PV power output measurements.

Adaptive Distributed EMS for Small Clusters of Resilient LVDC Microgrids

“…The physical model assumes a small distance between neighboring microgrids, thus the proposed EMS ignores the transmission/distribution line impedances. The detail implementation and associated algorithms could be consulted in [23].…”

“…PV) power production. These profiles are the output of the first layer of the EMS, pre-processing of data, and they were described in detail in [23]. Note that these profiles were obtained using real measured data for one year with a 10 minutes resolution, from a residential location in Cyprus.…”

“…same temporal shape for all MGs, which is not desired). Each individual MG first runs its local EMS, a scheduling based on mixed integer linear programming optimization as it was described and analyzed in detail in [23]. The output of this local EMS are the local variables of the associated ER (step 1 from the algorithm presented in Table II) of the MG.…”