Experimental comparison of PV-smoothing controllers using distributed generators

Abstract

The power output variability of photovoltaic systems can affect local electrical grids in locations with high renewable energy penetrations or weak distribution or transmission systems. In those rare cases, quick controllable generators (e.g., energy storage systems) or loads can counteract the destabilizing effects by compensating for the power fluctuations. Previously, control algorithms for coordinated and uncoordinated operation of a small natural gas engine-generator (genset) and a battery for smoothing PV plant output were optimized using MATLAB/Simulink simulations. The simulations demonstrated that a traditional generation resource such as a natural gas genset in combination with a battery would smooth the photovoltaic output while using a smaller battery state of charge (SOC) range and extending the life of the battery. This paper reports on the experimental implementation of the coordinated and uncoordinated controllers to verify the simulations and determine the differences in the controllers. The experiments were performed with the PNM PV and energy storage Prosperity site and a gas engine-generator located at the Aperture Center at Mesa Del Sol in Albuquerque, New Mexico. Two field demonstrations were performed to compare the different PV smoothing control algorithms: (1) implementing the coordinated and uncoordinated controls while switching off a subsection of the PV array at precise times on successive clear days, and (2) comparing the results of the battery and genset outputs for the coordinated control on a high variability day with simulations of the coordinated and uncoordinated controls. It was found that for certain PV power profiles the SOC range of the battery may be larger with the coordinated control, but the total amp-hours through the battery-which approximates battery wear-will always be smaller with the coordinated control.

Stochastic optimisation with risk aversion for virtual power plant operations: a rolling horizon control

“…The aggregate capacity of the VPP is ∼1694 kW. Prior work demonstrated the control of the PNM BESS and the MdS BEMS to provide PV smoothing [57, 58] as well as the control of DETL assets [59].…”

Design and Evaluation of a Secure Virtual Power Plant.

“…For instance, PNM research at the Prosperity Site has demonstrated PV smoothing 56 , load shifting 57 , and peak smoothing 58 . This team also conduced PV-smoothing control demonstrations using the prosperity battery and the MdS genset 59,60 .…”

“…The remaining generators and demand response units at MdS, Prosperity and UNM would be controlled with a variety of communication protocols: -Control of the MdS devices will be conducted through a communication link to the Toshibadeveloped Building Energy Management System (BEMS). In a previous U.S.-Japan bilateral project, BEMS-connected DERs at MdS were controlled to provide PV smoothing using an OSIsoft PI historian 72 . -The prosperity battery is controlled using gain adjustments and auxiliary inputs in the control algorithm 73 , programmed into the Battery Energy Storage System (BESS) by Ecolt.…”