PV inverter dynamics for reduced order model development

Climate change concerns have prompted significant interest in generating electricity from renewable resources. Due to decreasing photovoltaic (PV) system costs, PV generation is now becoming more widely deployed throughout the distribution network in Ireland. To study the potential impacts resulting from high penetration levels of residential PV installations, this paper presents the dynamic response of a photovoltaic converter to large changes in active power demand that could form part of a smart grid frequency control strategy. In this scenario, the distributed generation (DG) PV units would be required to collaborate autonomously in a short time scale and with centralized commands over longer time scales to achieve active power network stabilisation. The dynamics of a PV converter are extracted from a detailed, component level, 5.5 kWp PV system model taking account of the dynamic of each process involved. In this investigation a self-commutated standalone system is modelled with unity power factor load changes from 20 % – 90 % of rated capacity, capturing the converter dynamics decoupled from network reactance factors. The extracted system dynamics are required to develop a reduced order model that will reduce simulation times when considering a multi-agent distributed network whilst maintaining model accuracy. The use of this model will allow for advanced analysis of distributed control strategies to optimize distribution grid stiffness, stability and security. The reduction in the overall complexity of the network model, by employing reduced order agent models, is key to the development of detailed PV impact studies.