Modeling, Simulation and Forecasting of PV Solar Power Plants in Pakistan using Agent Based Approach

Abstract

Direct relation of economic growth in a country and its energy production has always been selfevident. Currently, Pakistan is exploring its renewable resources vastly including solar energy potential, in order to achieve clean and reliable energy goals. This births the need for effective simulation and modeling of solar power generation across the country. A reliable framework that generates accurate energy profiles of renewable power generation projects is lacking. Such framework would further aid in cost effective integration of renewable projects in energy mix for Pakistan. This research proposes modeling and simulation of PV solar power plants in Pakistan for effective forecasting of solar power generation. This framework is aimed to be scalable and robust enough that it proves efficient against conventional approaches towards power generation profiling. The framework presented in this thesis starts with initially developing a dynamic model for a PV solar power plant and developing its power generation profile using agent based modelling (ABM) approach. The ABM methodology adopted in this research, unlike discrete event modeling uses individual components as active components of the system by identifying them and defining their behavior. These active components are termed as Agents and are solar cells, solar modules, solar arrays, solar strings, dc combiner terminals, DC-AC inverters and transformers in this framework. This research framework is successfully validated for generation data acquired from 100MW Quaid-e-Azam Solar Park phase-1 Bahawalpur, Punjab, Pakistan. This implemented and validated PV solar power plant model will prove to be of vital resource for integrating renewable power generation in existing power grid. This scalable dynamic model of 100MW solar plant will further aid in realistic analysis of the future dynamics associated with renewable power generation projects.