Simulation Tools and Techniques for Analyzing the Impacts of Photovoltaic System Integration

Solar photovoltaic (PV) energy integration in distribution networks is one of the fastest growing sectors of distributed energy integration. The growth in solar PV integration is incentivized by various clean power policies, global interest in solar energy, and reduction in manufacturing and installation costs of solar energy systems. The increase in solar PV integration has raised a number of concerns regarding the potential impacts that might arise as a result of high PV penetration. Some impacts have already been recorded in networks with high PV penetration such as in China, Germany, and USA (Hawaii and California). Therefore, network planning is becoming more intricate as new technologies are integrated into the existing electric grid. The integrated new technologies pose certain compatibility concerns regarding the existing electric grid infrastructure. Therefore, PV integration impact studies are becoming more essential in order to have a better understanding of how to advance the solar PV integration efforts without introducing adverse impacts into the network. PV impact studies are important for understanding the nature of the new introduced phenomena. Understanding the nature of the potential impacts is a key factor for mitigating and accommodating for said impacts. Traditionally, electric power utilities relied on phasor-based power flow simulations for planning their electric networks. However, the conventional, commercially available, phasor-based simulation tools do not provide proper visibility across a wide spectrum of electric phenomena. Moreover, different types of simulation approaches are suitable for specific types of studies. For instance, power flow software cannot be used for studying time varying phenomena. At the same time, it is not practical to use electromagnetic transient (EMT) tools to perform power flow solutions. Therefore, some electric phenomena caused by the variability of PV generation are not visible using the conventional utility simulation software. On the other hand, EMT simulation tools provide high accuracy and visibility over a wide bandwidth of frequencies at the expense of larger processing and memory requirements, limited network size, and long simulation time. Therefore, there is a gap in simulation tools and techniques that can efficiently and effectively identify potential PV impact. New planning simulation tools are needed in order to accommodate for the simulation requirements of new integrated technologies in the electric grid. The dissertation at hand starts by identifying some of the potential impacts that are caused by high PV penetration. A phasor-based quasi-static time series (QSTS) analysis tool is developed in order to study the slow dynamics that are caused by the variations in the PV generation that lead to voltage fluctuations. Moreover, some EMT simulations are performed in order to study the impacts of PV systems on the electric network harmonic levels. These studies provide insights into the type and duration of certain impacts, as well as the conditions that may lead to adverse phenomena. In addition these studies present an idea about the type of simulation tools that are sufficient for each type of study. After identifying some of the potential impacts, certain planning tools and techniques are proposed. The potential PV impacts may cause certain utilities to refrain from integrating PV systems into their networks. However, each electric network has a certain limit beyond which the impacts become substantial and may adversely interfere with the system operation and the equipment along the feeder; this limit is referred to as the hosting limit (or hosting capacity). Therefore, it is important for utilities to identify the PV hosting limit on a specific electric network in order to safely and confidently integrate the maximum possible PV systems. In the following dissertation, two approaches have been proposed for identifying the hosing limit: 1. Analytical approach: this is a theoretical mathematical approach that demonstrated the understanding of the fundamentals of electric power system operation. It provides an easy way to estimate the maximum amount of PV power that can be injected at each node in the network. This approach has been tested and validated. 2. Stochastic simulation software approach: this approach provides a comprehensive simulation software that can be used in order to identify the PV hosting limit. The software performs a large number of stochastic simulation while varying the PV system size and location. The collected data is then analyzed for violations in the voltage levels, voltage fluctuations and reverse power flow. It is important to note that there are multiple factors that affect the hosting limits in a distribution network. Moreover, the limit can be assessed based on different parameters; however, it will be shown in this dissertation that in most cases the voltage level is the first parameter to be violated under high PV penetration conditions. Therefore, in both approaches, the voltage is considered the main factor to be monitored for violations for PV hosting limit identification. The work presented hereinafter focuses on providing novel, innovative and practical solutions for fulfilling certain gaps in power system simulation. A novel hybrid simulation tool is presented in this dissertation as a solution for some of the issues facing the simulation of distribution networks with high PV penetration. Hybrid simulation is a relatively new concept in power system simulation and has not yet been applied for studying PV impacts in distribution networks. The presented hybrid tool offers accurate results and fast simulations. It can be used for various applications regarding the study of PV impacts as will be shown in this dissertation. It interfaces an EMT model of a grid-tied PV system with a phasor-domain model of a distribution network. The presented hybrid simulation tool incorporates a phasor-domain QSTS simulation with a time-domain EMT simulation which allows for a wide range of frequency visibility. The tool offers full EMT-level visibility at the point of common coupling (PCC), as well as slow dynamic visibility through the QSTS simulation. The tool is validated and tested by comparing the results with a full EMT simulation. It is used for studying the impacts of PV systems on the distribution network during fault conditions, islanding situation, solar irrandiance variation, among many other applications. The developed tool is made completely open-source in order to promote the hybrid simulation concept in power systems simulations as a viable solution for many of the conventional simulation tool limitations. Moreover, the work presented hereinafter proposes a novel co-simulation architecture with power hardware-in-the-loop (PHIL) simulation. The proposed architecture is the first of its kind developed at the National Renewable Energy Laboratory (NREL). The co-simulation testbed is developed in order to allow for wider range of hardware testing before deploying new technologies into the field. The testbed is composed of: 1. A full phasor model of the distribution network developed using a commercial distribution management software (DMS) environment. 2. A reduced form of the full network model developed in a real-time EMT environment using Opal-RT real-time simulator. 3. A hardware setup tested in a PHIL simulation environment. The hardware setup represents a grid-tied PV system at the PCC composed of a grid simulator, PV simulator, and a PV inverter. The co-simulation allows a slow QSTS simulation to be performed using the DMS model where slow variations are simulated, such as voltage regulator operations and slow variations in loads and solar irradiance. The QSTS simulation updates the EMT model components (loads, generators, voltage sources, and voltage regulators). The reduced EMT model is solved in real-time which allows for detailed and accurate visibility of the transient phenomena occurring in the network. Finally, the EMT model communicates with the physical hardware device at the PCC in order to close the PHIL loop. This architecture expands the capabilities of conventional PHIL testing and allows for more tests and scenarios to be implemented. The co-simulation testbed is tested by solving a real feeder network model in the DMS using historic load and solar irradiance data. The phasor model updates the EMT model's loads, voltage sources, and voltage regulator status at every QSTS time-step. The EMT model communicates through the hardware interface of the real-time simulator with the hardware setup by sending command control signals to the grid simulator and the PV simulator in order to replicate the simulated conditions in the real physical hardware. The inverter is tested under different operation modes, and its capabilities to use advanced algorithms for voltage regulation are put to test. The co-simulation architecture also addresses the stability and accuracy concerns of the PHIL experiments. A detailed stability and accuracy analysis is discussed in Chapter 6.