Conference Year

2014

Keywords

zero energy building, low energy building, single objective optimization, multi-objectives optimization

Abstract

Low energy building and zero energy building have attracted increasing attention in both academic and professional fields following the ambitions of many governments in reducing building energy consumption and carbon emission. This paper presents an investigation on the optimal design of renewable energy systems in two types of buildings: Low Energy Buildings and Zero Energy Buildings. The first zero energy building in Hong Kong, namely Hong Kong Zero Carbon Building (ZCB), is taken as a reference building in this study. The TRNSYS building model is used to generate the annual cooling load profile of the building. Simplified models are developed to simulate the building energy systems including the air-conditioning systems and the renewable energy systems in Matlab while the building annual cooling load profile is taken as the input. GA (Genetic Algorithm) method and NSGA-? (Non-dominated Sorting Genetic Algorithm) approach are implemented for single objective optimization and multi-objectives optimization respectively in Matlab. Three most important design parameters, i.e., sizes of PV, wind turbine and bio-diesel generator, are chosen as the variables to be optimized. Three objectives (total cost, CO2 emission and grid stress factor) are adopted in the multi-objective optimization. They also form the objective function in the case of single objective optimization. The performances of buildings with different combinations of renewable system sizes are compared and evaluated. The effects of the two types of buildings on the design decisions of renewable energy system sizes are studied and compared. Furthermore, the uses of single objective and multi-objective optimization methods and their advantages/disadvantages in system optimization applications are discussed.

3484_presentation.pdf (1853 kB)
Design Optimization of Renewable Energy Systems in Low/Zero Energy Buildings Using Single and Multi-Objective Optimization Methods

Share

COinS