Please use this identifier to cite or link to this item: http://scholarbank.nus.edu.sg/handle/10635/136280
Title: DEVELOPMENT OF AN AUTOMATED ENERGY AUDIT PROTOCOL FOR OFFICE BUILDINGS
Authors: CHIRAG DEB
Keywords: Energy auditing, office buildings, energy consumption, retrofitting, machine learning, prediction models
Issue Date: 20-Jan-2017
Source: CHIRAG DEB (2017-01-20). DEVELOPMENT OF AN AUTOMATED ENERGY AUDIT PROTOCOL FOR OFFICE BUILDINGS. ScholarBank@NUS Repository.
Abstract: This study aims to enhance the building energy audit process, and bring about reduction in time and cost requirements in the conduction of a full physical audit. For this, a total of 5 Energy Service Companies in Singapore have collaborated and provided energy audit reports for 62 office buildings. Several statistical techniques are adopted to analyse these reports. These techniques comprise cluster analysis and development of prediction models to predict energy savings for buildings. The cluster analysis shows that there are 3 clusters of buildings experiencing different levels of energy savings. To understand the effect of building variables on the change in EUI, a robust iterative process for selecting the appropriate variables is developed. The results show that the 4 variables of GFA, non-air-conditioning energy consumption, average chiller plant efficiency and installed capacity of chillers should be taken for clustering. This analysis is extended to the development of prediction models using linear regression and artificial neural networks (ANN). An exhaustive variable selection algorithm is developed to select the input variables for the two energy saving prediction models. The results show that the ANN prediction model can predict the energy saving potential of a given building with an accuracy of ±14.8%.
URI: http://scholarbank.nus.edu.sg/handle/10635/136280
Appears in Collections:Ph.D Theses (Open)

Show full item record
Files in This Item:
File Description SizeFormatAccess SettingsVersion 
Thesis_Final_Electronic.pdf9.43 MBAdobe PDF

OPEN

NoneView/Download

Page view(s)

67
checked on Jan 14, 2018

Download(s)

520
checked on Jan 14, 2018

Google ScholarTM

Check


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.