<p>As the solar PV technology continues to evolve as the most common distributed generation (DG) coupled with increasing interconnection requests, accurate modelling of the potential operational impacts of this game-changer is pivotal in order to maintain the reliability of the electric grid. The overall goal of this research is to conduct an interconnection impact analysis of solar PV systems at increasing penetration levels subject to the feeder constraints within the distribution network. This is carried out with a time series power flow analysis method to capture the time-varying nature of solar PV and load with their interactions with the distribution network device operations. Also, this thesis analyses multiple PV systems scenarios and a wide range of possible impacts to enable distribution system planners and operators understand and characterize grid operations with the integration of PV systems. An evaluation of the operational and reliability performance of a grid-connected PV system based on IEC standards and industry guides is performed to detect design failures and avoid unnecessary delays to PV penetration. The performance analysis metrics in this research allow cross-comparison between PV systems operating under different climatic conditions. This thesis shows the significant impact of temperature on the overall performance of the PV system. This research conducts an interconnection study for spatially distributed single-phase grid-tied PV systems with a five minute-resolution load and solar irradiance data on a typical distribution feeder. Also, this research compares the performance of generator models, PQ and P |V |, for connecting PV-DG with the distribution feeder with their respective computational costs for a converged power flow solution. More so, a method capable of computing the incremental capacity additions, measuring risks and upgrade deferral provided by PV systems deployments is investigated in this research. This thesis proposes surrogate metrics, energy exceeding normal rating and unserved energy, for evaluating system reliability and capacity usage which can be a very useful visualization tool for utilities. Also, sensitivity analysis is performed for optimal location of the PV system on the distribution network. This is important because optimal integration of PV systems is often near-optimal for network capacity relief issues as well. This thesis models the impact of centralized PV variability on the electric grid using the wavelet variability model (WVM) which considers the key factors that affect PV variability such as PV footprint, density and cloud movement over the entire PV plant. The upscaling advantage from a single module and point irradiance sensor to geographic smoothing over the entire PV footprint in WVM is used to simulate effects of a utility-interactive PV system on the distribution feeder. Further, the PV interconnection scenarios presented in this thesis have been modelled with different time scales ranging from seconds to hours in order to accurately capture and represent various impacts. The analysis and advancements presented in this thesis will help utilities and other stakeholders to develop realistic projections of PV systems impacts on the grid. Also, this research will assist in understanding and full characterization of PV integration with the grid to avoid unnecessary delays.</p>
Distribution Feeder Parameter Estimation without Synchronized Phasor Measurement by using Radial Basis Function Neural Networks and Multi-run Optimization Method
