An analysis of residential water demand for Ontario and the Prairie Provinces from 1989 to 1999

Abstract

The research centred on the factors accounting for variations within residential water demand. To accomplish this task, the statistical technique of regression analysis was used to analyze water demand data compiled by Environment Canada, for five different years, within two Canadian regions: Ontario and the Prairie Provinces. Two dependent variables were analyzed: total residential water demand (TRWD) and residential water demand per capita (RWD/cap). Within the TRWD model, population served by water services was found to be the only statistically significant variable, with coefficients of determination around 0.9. It was therefore found, that for forecasting of regional water demands, multiplying a predicted population level with a per capita coefficient might be a sufficient technique at the regional level. Other variables tested included price structure, price of water services, household income and rainfall. With respect to the RWD/cap model, marginal prices were found to be statistically significant within the Prairie region, as well as for the combination of both regions. The multiple linear regression model produced R2 values that were very low (varying from 0.030 to 0.130). (Abstract shortened by UMI.)