Analysis of Techniques for Estimating Photovoltaic Hosting Limit in Distribution Systems

The global energy demand has increased significantly alongside the worldwide population in the last decade. The global energy demand has been met for years by carbon fuel-based sources, which has increased the pace of carbon dioxide (CO2) emissions. These emissions affect the environment in a significant proportion, which has led to different efforts from governments to slow down the pace of emissions. Consequently, alternative ways of generating electricity have been investigated to have more sustainable and "greener" sources destined for energy supply. One of the technologies that utilize clean energy resources and has been widely used is solar Photovoltaic (PV). Behind the fast-growing integration of this technology into distribution networks are government incentives and regulation policies, which have resulted in reduced costs for manufacturing and installations of PV systems. The PV system integration is forecasted to keep growing, but their interconnection with traditional power systems may endanger its regular operational performance. Higher levels of PV penetration might cause an adverse impact on the different components of the grid. The protection and voltage regulation devices existent on a traditional network are designed and configured to operate under a unidirectional power flow. The PV system interconnection causes a bi-directional power flow, which will affect those components. The conventional power system is propelled to transition to a more dynamic system able to operate with different renewable sources interconnected, and even allow the participation of consumers in energy market transactions. Planning is a crucial stage for utility operators in order to achieve a successful integration without endangering the electric network's reliability and safety. Each distribution network has a limit of maximum PV penetration which within, the presence of PV systems will not adversely affect the regular operation of the system and equipment throughout the feeder; this limit is known as hosting capacity (HC) or hosting limit. Research conducted by industry companies and academia focused on solar PV technology have developed planning tools and methods capable of determining the PV hosting limit. In the following thesis work three most used techniques for PV Hosting Limit estimation are studied: 1. Iterative Approach: This method is one of the most used, and it is based on performing repetitive power flows. It represents an easy method to estimate the hosting limit but has certain limitations regarding computational time and accuracy. These limitations are demonstrated with test cases. 2. Sensitivity-Based Approach: This is a mathematical approach based on the Newton-Raphson power flow approach. It is a faster estimation approach compared with others but also has limitations caused by assumptions that are not tailored to distribution systems specifically. A detailed analysis is given by testing two different distribution systems. 3. Stochastic-analysis framework: This approach provides a comprehensive solution to the PV hosting capacity since it considers different PV sizes and locations to evaluate the impact on performance criteria. It is based on simulation software to perform numerous power flow scenarios, changing the PV penetration, and location randomly. The final results show the number of violations in the voltage levels per node, and considers reverse power flow scenarios. The three approaches are tested, and a detailed analysis is provided. The hosting limit is estimated with respect to the performance of indices or criteria since multiple factors have an impact on the final hosting capacity of a distribution system. However, the voltage is the most assessed criterion because it is the first one to be reached in the presence of High PV penetration. In all described approaches, the over-voltage is the parameter that restricts the PV hosting limit estimation. After the thorough analysis of the different approaches and the identification of limitations, the work in this thesis provided an alternative method to estimate the PV hosting limit. The proposed approach is an optimization-based method, which is based on the linearization of the traditional AC power flow equations. This allows a formulation based on linear programming when only the over-voltage criterion is considered. The objective is to maximize the PV power output that can be injected into a distribution network without causing adverse effects. The hosting capacity is assessed concerning three main factors: the voltage level, thermal rating limit of distribution lines, and reverse power flow condition. The method is validated and tested with two different distribution systems and compared with other approaches. The results demonstrate that the proposed approach offers similar solutions, while providing a balanced performance concerning the accuracy and computational time efficiency.