Grid Integration of Photovoltaic Plants – a Generic Description of Pv Plants for Grid Studies

This article presents an approach to modelling of threephase photovoltaic (PV) inverters for RMS-based grid integration studies under balanced and unbalanced conditions. The inverter model was rigorously derived from accurate instantaneous value models that were simultaneously used for the development of inverter control and operations software. This approach leads to a precise generic phasor model in the dq reference frame that can be used both for certification purposes and for dynamic interconnection studies of PV power plants and allows implementation on several simulation platforms. The new model describes the basic functionality of inverter front-ended generators, and can be modularly extended with common features like voltage control or frequency control. Thereby grid operators are enabled to account for the grid supporting features of PV power plants in grid simulations. INTRODUCTION It is well known that by means of an increasing portion of installed PV power, simultaneously its influence on the existing electrical grid structure grows. In new revisions of national as well as international grid codes and terms of technical installations, requirements are currently being defined which illustrate a paradigm shift regarding the grid behaviour of photovoltaic plants and other decentralised generators of electrical energy. Adjacent to the resulting requirements to inverter manufacturers for the implementation of functions like feed-in management or dynamic grid support, the demand for simulation of photovoltaic plant properties for the purpose of planning and design of distribution and transmission systems increases. Grid operator knowledge of decentralised energy generator behaviour is frequently not established sufficiently. In contrast the inverter manufacturer is challenged to provide models for the simulation of the plant. More precisely, the certification of devices and plants according to the current grid codes, e.g. [1], has to be mentioned. In addition to the verification of the electrical characteristics [2], requirements for the modelling are defined [3]. Simulation of wind farms as elements of power systems started with vendor-specific generator models provided by the manufacturers of wind generators, as was pointed out in [4]. These dynamic models were successfully used for transient stability studies but were also lacking easy handling capabilities in simulation environments due to the fact that they contained proprietary information. For that reason the development of generic publicly available models was started where all respective parameters can be modified according to any practical implementation. It soon evolved that the various types of wind generator technologies could not be represented by only one generic model and so the now well-known four types of wind generator models were suggested. In accordance to this development a similar strategy is currently being explored for PV generators that make use of the “type 4” wind generator model – the full converter interface ([5], [6], [7]). DISTRIBUTED GENERATION While wind generator models are mainly used for transmission network studies, PV generator models also play a significant role in the context of distribution network simulation. The high penetration of distributed generation (DG), primarily PV systems, leads to changes in grid operation, power flow direction and voltage profiles in distribution networks. Moreover the natural variation of solar irradiation causes fluctuations of the voltage V. Accordingly, PV systems should be able to participate in grid support and, when possible, in grid control. Active DG operation can be realised by implementation of primary control algorithms: e.g. Q(V) control, PF(P) control, P(V) control, dynamic grid support (DGS) to actively influence the grid voltage and increase the hosting capacity of the network, or P(f) control in order to impact the grid frequency f and support the grid if possible (here, Q is the reactive power, P is the active power, and PF represents the power factor). Additionally, a secondary or coordinated control can be realised. Especially for studies of the potential of grid support through PV DG, accurate inverter models are required that represent all grid supporting functions of the equipment. 1 also known as fault ride-through (FRT) C I R E D 21st International Conference on Electricity Distribution Frankfurt, 6-9 June 2011