Stochastic and multi-objective optimal sizing of energy resources within a residential home

Abstract

With the growing need for increased adoption of greener sources of energy, there has been a continued increase in the deployment of solar photovoltaic generating systems. When solar PV systems are deployed without any storage system, due to the mismatch between generation and demand coupled with the intermittency of the solar sources, PV generation would need to be curtailed or wasted. Furthermore, a battery energy storage system that is not optimally sized would lead to increased cost of the system. Also, building energy consumption continues to represent a large percentage (about 36%) of total national energy consumption. Increased electrification of heating and energy efficiency improvements have been identified as solutions to reduce the greenhouse gas emissions related to energy consumption in buildings. An optimally sized and operating heating system can help improve energy efficiency and reduce energy waste. Optimization can help unlock benefits to a system and ensure systems are optimally sized, operational, and economically viable. The thesis begins by proposing a stochastic optimization problem to determine the optimal size of the energy storage system in a PV-integrated system. The residential load demand varies over a given year. Hence the proposed mathematical optimization problem considers the uncertainty in the residential load demand. The benefits of considering uncertainty in the optimization problem are then investigated. The proposed problem was formulated with different objective costs and the solutions obtained showed the importance of considering uncertainty in the sizing problem. Then a multi-objective approach is proposed for the sizing problem. A multiobjective problem is formulated consisting of two objectives: to minimize the cost of purchasing the battery energy storage system and to minimize the amount of energy imported from the grid within the optimization period. The stochastic problem is formulated as a two-stage scenario stochastic problem. The Monte Carlo approach is used to deal with uncertainty in load demand forecasting. Results show that the proposed approach can estimate an optimal battery energy storage system at the current cost of the battery energy storage systems and clearly indicate the benefit of a stochastic approach. Lastly, dynamic thermal models of residential homes typically in the UK are obtained using the resistance-capacitance network approach. The thermal models developed showed good performance when compared to the estimation data with about 72-82% fitness attained. The obtained model was used in the sizing operation of the heating system. A stochastic multi-objective approach was also adopted here. Analysis was conducted on a linear and quadratic formulation of the regulation objective function. It was shown that the solution of the stochastic optimization approach performs better with regard to the operation of the system and the minimization of the objective cost.

Date of Award	1 Aug 2024
Original language	English
Awarding Institution	The University of Manchester
Supervisor	Haiyu Li (Supervisor) & Zhirun Hu (Supervisor)