User profiling and saving potentials

Posted by Tassilo Pellegrini on April 20, 2010 under case study, efficiency gains, user profiling | Be the First to Comment

To gain an insight into behavioral patterns four in-depth interviews were conducted with two men and two women living in individual households. The sample was not meant to be in any way representative for a certain kind of cohort, social class or target group but was intended to identify occupancy patterns, device types and their corresponding usage. Therefore, two time frames have been analyzed. The interviewees were asked to describe in detail their appliance usage during morning time from getting up to leaving for work – a time frame which is in so far crucial as morning hours generate peak loads at the energy providers’ side. Additionally, participants were asked to describe a typical week in their lives which was necessary to capture occupancy and sleeping patterns during workdays and weekends.

From this information we modeled a normalized seven day period in a single resident household representing the devices in use, occupancy and sleeping patterns. The average energy consumption per device-type was calculated and differentiated by active and passive use. By applying tariff schemes from the Austrian Energy Exchange  we also calculated the average energy costs for a single household per day and per week. These calculations were crucial as simple scenarios showed that a change in behavior might lead to a reduction of energy consumption but not necessarily to a cost reduction and vice versa if the prices of the energy market would be used by the end customer. These effects have to be observed closely as cost-efficiency and energy-efficiency do not necessarily correlate positively.

The empirical modeling of usage patterns performed within this diary experiment lead to the conclusion that policy based energy control could affect energy savings of up to 24% by simply applying automated turn-off rules to stand-by devices and ad-hoc devices alone when streamlined with the behavioral patterns of the user.

We will continue with this empirical work and search for more granular and non-bivious saving potentails in the following months.