Europe's ambitious climate target of achieving net-zero carbon emissions by 2050 requires a fundamental shift in grid management strategies, focusing on adjusting volume and time of demand to reliably integrate an increasing share of non-programmable renewable generation. To effectively inform policymakers and decision-makers on this front, household energy behaviour should be accounted for, given consumers are expected to actively engage with Demand-Side Management (DSM) programmes, whether through manual adjustments or preference settings for automated control. Thus, delving into household energy behaviour can provide greater clarity on how this can effectively and practically take place. Yet, conventional energy models often overlook the dynamic complexities of household energy end-uses, assuming overly simplistic and uniform consumption patterns across the population.

To contribute to the behavioural alignment of DSM program design, this thesis aims to investigate appliance use patterns and their heterogeneity across the population.

Using dishwashing and laundry habits in Germany and the German-speaking Cantons of Switzerland as illustrative case studies, this research follows three steps. First, statistical methods including hierarchical clustering, multinomial logistic regression, and analysis of variance were applied to survey data (N=1,188) to identify distinct appliance usage patterns, explore their determinants, and assess variations in load-shifting potential and perceived inconvenience. Second, these patterns were operationalized into an activity-based energy demand model, establishing a pattern-dependent link between occupant's activities and appliance usage schedules. Third, to investigate consumer preferences for load-shifting with respect to different DSM scenarios, a model-driven adaptive survey was conducted (N=1,594). Such a tool established a dynamic feedback loop between user responses and an energy demand model, supporting the decision-making process with tailored feedback.

The results demonstrated that meaningful patterns could be identified among the large diversity of dishwashing and laundry habits. Pattern membership was associated to multiple factors that resist a narrow categorisation, and households with more energy-intensive patterns tended to perceive load-shifting as more inconvenient. By operationalizing the identified patterns into an energy demand model, a deeper understanding of energy demand heterogeneity was gained compared to conventional approaches reliant solely on socio-demographic attributes. This supported a better identification of target segments and actionable DSM interventions. Moreover, a model-driven adaptive survey showed potential in exploring household preferences for load-shifting, supporting them in navigating the complexity associated with dynamic energy scenarios.

This interdisciplinary work advocates for a comprehensive reorientation of DSM programme analysis: shifting focus from household-centric to individual appliance usage-oriented perspectives; considering temporal and intensity aspects of appliance usage rather than relying solely on aggregated load profile data; and accounting for variations in the multiple factors shaping household energy practices beyond just socio-demographic attributes. By doing so, current analyses can better inform policymakers and decision-makers on the design of targeted and actionable DSM programmes.