A method to estimate absence in apartments based on domestic water use

Abstract

Today, every new building is constructed with the ambition of it having a very low energy demand or even being completely neutral with regard to supply and demand over a year. In modern low energy or nearly zero energy residential buildings like these, the behaviour of the occupants becomes increasingly more significant in terms of the energy performance of the building. Occupants who are present and using appliances reduce the amount of heat needed to be supplied by the heating system. Knowledge about occupant behaviour is, consequently, of great importance. As occupants’ presence, absence and behaviour are hard to measure without being intrusive, the level of occupancy is often derived from statistics gathered from questionnaire surveys or from measurements of other variables connected to occupancy, such as household electricity use. This study investigates such measurement data with regard to the use of household electricity, domestic hot water and domestic cold water gathered at hourly intervals over a whole year in over 1,000 Swedish apartments. A method is proposed whereby an apartment is regarded as being empty if no water (hot or cold) is used during an entire day. Using this method, 50% of the apartments were found to have 10 days of absence or more, and for apartments in the interquartile range, i.e. the 50% closest to the median, the periods of absence varied between 1 and 31 days. The validation showed an increase of absence during national holidays by about 300%, and by about 100% during weekends. The outcome of estimating absence using only one of the variables was evaluated and it was shown that domestic cold water use would be a good substitute method with 98% precision.