This work presents a two-stage model for the data analysis of electricity consumption. The first stage divides the consumption in two parts: weather- and illumination-related, and residual consumption, where weather-related consumption refers to heating, ventilation, and air conditioning (HVAC). Given the hourly total consumption, we obtain the hourly weather-related and illumination-related electricity consumption, and subtract this out to get residual consumption. The second stage of the model is a flexible, agent-based analytical tool that allows disaggregation of residual consumption into a sum of consumptions by different groups of appliances. This tool can be used for a variety of applications including an optimization of demand-side management and/or a development of a set of desirable patterns for schedules of electricity-related activities in households (within an acceptable range of course) in order to minimize negative effects of high peak demand.