INVESTIGATING THE IMPACT OF VARYING THE NUMBER OF DISTRIBUTED ENERGY RESOURCES ON CONTROLLING THE POWER FLOW WITHIN A MICROGRID Amir Fazeli Substation Automation Solutions, Smart Grid Research Centre Alstom Grid Ltd, Stafford, St Leonards Avenue, ST17 4LX, UK amir.fazeli@alstom.com Mark Sumner, C. Mark Johnson, Edward Christopher Power Electronics Machines and Control Group University of Nottingham Nottingham, University Park, NG7 2RD, UK Abstract-- The electrification of heat and transport in addition to integration of intermittent renewable resources into the existing electricity network is expected to occur in near future. Such a transformation is expected to force the operation of the electricity power system at different levels to its limits and would require reinforcement of the network assets at different levels. The incorporation of active management and control within microgrids and across the low voltage distribution network is thought as a cost effective solution which would facilitate wide scale integration of the emerging distributed energy resources. However since increasing the microgrid size at a certain DER penetration level would increase the total dispatchable power it is expected to affect the effectiveness of any control algorithm that operates at that level. This paper presents the findings obtained from of an investigation into the relationship between microgrid size and the effectiveness of a deterministic control algorithm implemented at that level. Index Terms—Microgrids, Distributed Energy Resources, Power Flow Management, Demand Side Management I. INTRODUCTION he electrification of heat and transport is expected to occur in the near future, which would consequently result in a substantial increase in electricity demand [1]. The increasing uptake of electric vehicles is expected to have a number of negative effects for the radial low voltage distribution network[2]. In particular a significant increase of voltage drop at the consumer end of LV feeders is expected. The integration of different forms of Distributed Generation (DG) to the low voltage network is also on the rise as different countries have set various targets to increase their share of renewable generation capacity. Large scale integration of Distributed Generation (DG) units requires a different approach to the planning and operation of the distribution grid. The foreseen adverse effects for the distribution network include voltage control, power quality, protection system and grid losses [3]. In order to investigate the impact of increasing the penetration level of the aforementioned Distributed Energy Resources (DER) within the low voltage network, the following stochastic DER, load and DG models have been created in Matlab/Simulink with a temporal resolution of one second:  Domestic electricity load [4]  Electric vehicle charging  Domestic heating load [5]  Ground source heat pump [6]  Photovoltaic generation [7]  Wind power generation [8]  Microgrid battery energy storage [9] A novel modelling technique for accurate quantification of electric vehicle charging requirements in terms of charging energy, power and duration has been developed by the authors. This modelling approach will be described in a future publication. All of these models are then used to develop a novel deterministic control algorithm (CPFC) for regulating the power flow through real time dispatch of the DERs within a microgrid [5]. The CPFC algorithm receives a power flow target ( T PF ) from an intermediate level of a smart grid control framework and determines the correct quantity and combination of DERS to dispatch at every moment. The structure and the logic behind the CPFC algorithm will be presented in a future publication. As the CPFC’s operation is based on real time dispatch of DERs, the “total available DER power” is an important parameter which could determine the effectiveness of this algorithm for controlling CPF at its target value. Therefore this project has investigated the effectiveness of the CPFC algorithm with respect to the microgrid size (i.e. in terms of the number of dwellings with different forms of T