Impact of Load Modeling on DG Integrated Distribution Systems

Abstract

Renewable energy based Distributed Generation (DG) resources are crucial for sustainable energy supply infrastructure as they are non-polluting and inexhaustible. The uncertainties associated with DG resources may cause distinct economic and technical challenges which require comprehensive investigation to facilitate their integration in Distribution System (DS). Generally, DG planning studies are conducted while assuming constant generation and load models. However, such assumptions may result in misleading and inconsistent values for voltage profile, loss reduction, payback period, deferral values, and other relevant calculations. Therefore, to achieve accurate and realistic results, it is necessary to consider variations associated with generation and load. This thesis presents time varying load modeling and probabilistic solar irradiance modeling techniques and investigates their impact on Photovoltaic (PV) based DG planning. A novel Beta distribution based probabilistic generation model is proposed for solar irradiance uncertainty modeling to compute the hourly output power values produced by PV based DG. The beta distribution parameters are found by assuming the variations of irradiance patterns at consecutive time steps. Subsequently, the proposed model is employed to generate various solar irradiance generation scenarios. Then, a time varying load modeling approach is presented for PV based DG planning. Five different types of time varying load models (i.e. residential, commercial, industrial, mixed and constant) are considered. These loads are modeled by combining the time varying characteristics of residential, commercial and industrial loads with the voltage-dependent load model while assuming suitable voltage exponents. The application of these load models make the PV based DG integration more realistic as compared to the conventional model. Furthermore, a methodology has been developed to determine intermittent DG allocation for DS while considering varying load and generation. The objective is to minimize the multiobjective optimization function which involves voltage deviation, active and reactivepower loss indices. Finally, the impact of time varying load modeling approach on DG integrated DS performance has been investigated. The proposed DG planning framework has been validated on IEEE 33-bus and 69-bus standard distribution test systems in MATLAB environment. A comparative assessment of different impact indices, penetration level, active and reactive power intake, active and reactive power loss and MVA support offered by the installation of PV based DG for different time varying load models has been performed. The results demonstrate that the proposed generation model is suitable for solar irradiance modeling. Moreover, time varying load modelling approach has a significant impact on DS planning studies under uncertain scenario.