Title: Modelling residential house electricity demand profile and analysis of peaksaver program using ANN: case study for Toronto, Canada

Authors: M. Ebrahim Poulad; Alan S. Fung; Lei He; Can Ozgur Colpan

Addresses: Mechanical and Industrial Engineering Department, Ryerson University, Toronto, Ontario, Canada ' Mechanical and Industrial Engineering Department, Ryerson University, Toronto, Ontario, Canada ' Mechanical and Industrial Engineering Department, Ryerson University, Toronto, Ontario, Canada ' Department of Mechanical Engineering, Dokuz Eylul University, Buca, Izmir, Turkey

Abstract: A technique is proposed and developed to predict the household hourly electricity demand. The developed artificial neural network (ANN) model of residential hourly demand is employed to estimate the potential impacts of load curtailment activation (LCA) on electricity demand on the activation days. Results are separately discussed in two seasons: summer and winter. LCA occurs once per day for no more than four consecutive hours. Electricity demand increases dramatically after peaksaver/LCA is completed on July 6 and August 30 of 2010. Both days show saving if the data are not normalised. Unnormalised load reductions for individual event hours ranged between 0.35 and 0.64 kWh/h or 14% and 24%, respectively.

Keywords: demand management; artificial neural networks; ANNs; greenhouse gases; GHG emissions; peak saving; load curtailment activation; LCA; modelling; residential housing; electricity demand; case study; Toronto; Canada; hourly demand.

DOI: 10.1504/IJGW.2016.077911

International Journal of Global Warming, 2016 Vol.10 No.1/2/3, pp.158 - 177

Received: 23 Oct 2014
Accepted: 16 Mar 2015
Published online: 22 Jul 2016 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article