Insight Report: Domestic Time of Use Tariff: A comparison of the time of use tariff trial to the baseline domestic profiles

Abstract

Peak electricity demand poses a particular challenge both to network operators and to energy suppliers. A reduction in peak demand would allow existing networks to accommodate load growth with lower investment, and also reduce the cost of electricity generation during peak periods. To explore the potential for peak reduction, the Customer-Led Network Revolution project has trialled a Time of Use (ToU) tariff scheme. By increasing electricity prices during the weekday peak period (4pm-8pm) for throughout the year and reducing prices in off-peak periods, the tariff incentivises a shift in consumption out of the peak period. A static ToU tariff was used – that is, the tariff remained constant and did not change dynamically depending on expected network loading. Based on smart meter data and survey responses, test cell (TC) 9a investigated the electricity use patterns of 574 domestic users on a ToU tariff between October 2012 and September 2013 and compared them to those of the control group (test cell 1a). The impact of demographic profile on customers’ responses to the ToU tariff was also considered. On average, when compared to consumers in the control cell TC1a, customers on the ToU tariff had lower consumption during the peak period on weekdays, and higher consumption at other times – indicating that the tariff achieved the intended behaviour change. There was a small net reduction (0.8%) in annual consumption, although this was not enough to be statistically significant. In particular, the trial showed; Lower electricity consumption during the peak periods (between 1.5% and 11.3% less than TC1a). This is in line with our qualitative research where customers claim changing time of use of certain appliances. Lower average peak power demands1 during the peak period (between 3.2% and 12.5% lower than TC1a when averaged throughout the year and across all customers). On average, customers showed a lower maximum half-hourly peak demand (between 2.1% – 10.3% lower than TC1a) during the peak period. However at the time of greatest system peak demand – specifically a single half-hour in the year there was no (statistically significant) difference in the mean peak demand observed between TC9a and TC1a