2021/11/02 by Jianli Chen, Rajendra Adhikari, R.S. Adhikari +12
Computer Science · Energy · Engineering · Physics and Astronomy · #Building Energy and Comfort Optimization #Computer science #Data Analysis #Economics #Electricity #Energy Efficiency and Management #Energy consumption #Engineering #FOS: Electrical engineering #FOS: Physical sciences #Markov chain #Probabilistic logic #Renewable energy #Simulation #Smart Grid Energy Management #Statistics and Probability (physics.data-an) #Stochastic modelling #Stock (firearms) #Systems and Control (eess.SY) #cs.SY #eess.SY #electronic engineering #information engineering #physics.data-an
paper · pdf · doi:10.48550/arxiv.2111.01881
openalex publication_date 2021/11/02 · arxiv created 2021/11/04 · arxiv updated 2021/11/05 · openalex created_date 2021/11/08 · openalex updated_date 2026/07/28
The residential buildings sector is one of the largest electricity consumers worldwide and contributes disproportionally to peak electricity demand in many regions. Strongly driven by occupant activities at home, household energy consumption is stochastic and heterogeneous in nature. However, most residential building energy models applied by industry use homogeneous, deterministic occupant activity schedules, which work well for predictions of annual energy consumption, but can result in unrealistic hourly or sub-hourly electric load profiles, with exaggerated or muted peaks. This mattered less in the past, but the increasing proportion of variable renewable energy generators in power systems means that representing the heterogeneity and stochasticity of occupant behavior is crucial for reliable energy planning. This is particularly true for systems that include distributed energy resources, such as grid-interactive efficient buildings, solar photovoltaics, and battery storage. This work presents a stochastic occupant behavior simulator that models the energy use behavior of individual household members. It also presents an integration with a building stock model to simulate residential building loads more accurately at community, city, state, and national scales. More specifically, we first employ clustering techniques to identify distinct patterns of occupant behavior. Then, we combine time-inhomogeneous Markov chain simulations with probabilistic sampling of event durations to realistically simulate occupant behaviors. This stochastic simulator is integrated with ResStock, a large-scale residential building stock simulation tool, to demonstrate the capability of stochastic residential building load modeling at scale. The simulation results were validated against both American Time Use Survey data and measured end-use electricity data for accuracy and reliability.