Zoning Effects and Stochastic User Profiles: A Multi-Method Sensitivity Analysis for Residential Building Simulation
Year: 2026
Type: conference-paper
Abstract: Building energy simulation can be strongly influenced by thermal zoning and coupling assumptions. However, zoning decisions are rarely assessed together with uncertainties from building data and user behavior. This work quantifies the influence of zoning granularity, layout, and stairwell representation for a multifamily building relative to physical input uncertainty and stochastic occupancy. Eight zoning variants of a single building are generated in a workflow for reduced-order building models. The uncertainty propagation samples construction state, age class, window-to-wall ratio, setpoints, internal gains, weather, and stochastic household profiles. In addition, Morris screening and Sobol analysis are used to rank influential input parameters and quantify first- and total-order effects. The results show that zoning acts as a structural uncertainty: it affects absolute KPI values, uncertainty ranges, and parameter sensitivities, but not equally for all outputs. For annual heating demand and peak heating load, the dominant structural effect is the transition from a single-zone model to explicit multi-zone representations, whereas overheating indicators are more sensitive to zoning resolution, layout, and retrofit state. Sensitivity rankings are target dependent: air change and construction assumptions dominate heating demand, whereas overheating is more strongly influenced by window, shading, solar, and weather effects.
Source: Proceedings of BauSIM
Language: English
Subfield: Building and Construction
Field: Engineering
Domain: Physical Sciences
Open Access status: bronze
OpenAlex ID: W7224934986