Robust Allocation of Decentralized Photovoltaic and Energy Storage Systems in a Power Grid under Uncertainty


Cangul, Ozcel (2021) Robust Allocation of Decentralized Photovoltaic and Energy Storage Systems in a Power Grid under Uncertainty. PhD thesis, University of Liverpool.

[img] Text
200977076_Jul2021.pdf - Unspecified
Access to this file is embargoed until 1 August 2024.
Download (5MB)

Abstract

Power grids are already complex systems with many parameters such as generation, transmission, distribution and demand of electrical power. In today’s world, the amount of renewable energy sources (RESs) integrated in a power grid are continuously increasing. This gradual transition of increased harnessing of RESs introduced a new perspective to the electrical energy generation, i.e. decentralized generators. Correspondingly, the common way of generating power has been shifting from central power plants to a combination of central and distributed generation. Consequently, energy engineers have been searching for methods for establishing an effective and robust balance between centralized and decentralized generators. Due to their stochastic nature of availability, the RESs have recently started to be supported with energy storage systems (ESSs) which enable controlling of their exploitation over a wider time horizon rather than being limited by their availability. As a result, a new problem arises with the integration of growing amounts of RESs and ESSs: Optimal allocation of RESs and ESSs in a power system. This work introduces a new probabilistic framework for the optimal positioning and sizing of utility-scale solar photovoltaic (PV) systems in a transmission network. Two different perspectives of analysis are conducted for this purpose: economic and technical. For the economic analysis, a novel financial metric, the Unit Financial Impact Indicator (UFII), is proposed to minimize the pay-back period for the solar PV investment and offers a direct quantification of the financial efficiency of a RES allocation strategy. The siting and sizing problem of solar PV units is optimized using this indicator. For the technical analysis, the utility-scale PV units and ESSs are optimally allocated by minimizing the bus voltage deviations from unity in order to acquire a better voltage profile. Inaccurate quantification of environmental and operational uncertainties, such as variable power demands, component failures, and weather conditions, can threaten the reliability of the investment. To overcome this difficulty, a new cloud intensity simulator affecting PV power production is defined based on historical data and is embedded within a Monte Carlo simulator specifically designed to quantify the uncertainty in the output of interest. Sensitivity analyses are conducted to investigate the relationship between the uncertain inputs and the output of interest. Genetic Algorithm and Nonlinear Programming optimization strategies are employed to identify optimal designs. The efficiency and the usefulness of the proposed approaches are tested on both 14-bus IEEE power grid and 35-bus North Cyprus transmission network as case studies. Findings of the optimizations using UFII are also compared with those obtained by other economic metrics used in the literature. The results prove the efficacy of the new metric and methods towards quantifying the financial and technical effectiveness of solar PV and ESS investments on different nodes and geographical regions in a power grid.

Item Type: Thesis (PhD)
Uncontrolled Keywords: decentralized generation, distributed generation, energy storage, genetic algorithm, monte carlo simulation, optimal allocation, photovoltaic, power grid, risk analysis, uncertainty quantification, unit financial impact indicator, voltage deviation
Divisions: Faculty of Science and Engineering > School of Engineering
Depositing User: Symplectic Admin
Date Deposited: 06 Sep 2022 10:07
Last Modified: 18 Jan 2023 21:04
DOI: 10.17638/03154378
Supervisors:
URI: https://livrepository.liverpool.ac.uk/id/eprint/3154378
Dimensions
Altmetric
CORE (COnnecting REpositories)