U.S. synthetic residential building telemetry dataset generated with EnergyPlus.
- 1School of Architecture, University of Illinois at Urbana-Champaign, 117 Temple Hoyne Buell Hall, 608 Lorado Taft Dr, Champaign, IL, 61820, USA.
Abstract
High-resolution residential telemetry, including smart thermostat data, is increasingly used to develop data-driven models for building energy analysis and HVAC performance assessment. However, field telemetry often lacks detailed information on building characteristics and known physical parameters, limiting rigorous model development and validation. This article presents a U.S. synthetic residential building telemetry dataset generated using EnergyPlus that combines high-resolution time-series telemetry with fully known building, envelope, infiltration, and HVAC parameters. It comprises 540 unique models of single-family detached and attached homes across the 48 contiguous states and 13 ASHRAE climate zones. The models were generated using probability distributions of building and HVAC characteristics derived from the ecobee Donate Your Data (DYD) program and the U.S. residential building stock represented by ResStock. Statistical case weights are provided for the 540 unique models, defining a probability-weighted analytical population of 1200 residential cases. For each model, the dataset provides an EnergyPlus input file and a full year of synthetic telemetry at 5-minute intervals in Parquet format. The telemetry includes indoor and outdoor environmental conditions, thermostat setpoints, HVAC operation and runtime, infiltration and ventilation, heating and cooling rates, and energy-consumption variables. The dataset also includes building characteristics, simulation metadata, and weather-station location. Comparison of HVAC-runtime patterns with ecobee field telemetry indicates that the synthetic data reproduce broad climate-dependent patterns in residential HVAC operation while retaining known model inputs and physical characteristics. The dataset can support the development and benchmarking of surrogate and telemetry-based models, evaluation of retrofit and HVAC-sizing scenarios, and analysis of residential energy performance at national, state, and climate-zone scales.
Related
- Dynamic modeling, simulation, and analysis of solar/wind-based hybrid renewable energy system for rural residential building: a case study
- Building Seasonal Highways for Residential Energy Hubs: Sizing, planning and operating thermal energy storage
- Surfactant-Enhanced In Situ Remediation Beneath an Occupied Residential Building: Reduction of the BTEX Vapor-Intrusion Source Term and Field Validation in a Single Field Season.
- Building archetypal load profiles for Ukrainian residential consumers: economic differentiation and demand side management potential
- Development and energy savings analysis of active thermal switches with thermal storage for dynamic building envelopes in residential buildings
BibTeX
@article{kim2026synthetic,
title = {{U.S. synthetic residential building telemetry dataset generated with EnergyPlus.}},
author = {You-Jeong Kim and Yun Kyu Yi},
journal = {Data in brief},
year = {2026},
month = aug,
doi = {10.1016/j.dib.2026.113327},
url = {https://europepmc.org/article/MED/42927244},
note = {CC BY},
}
RIS
TY - JOUR
TI - U.S. synthetic residential building telemetry dataset generated with EnergyPlus.
AU - You-Jeong Kim
AU - Yun Kyu Yi
PY - 2026
DA - 2026/08/01
JO - Data in brief
DO - 10.1016/j.dib.2026.113327
UR - https://europepmc.org/article/MED/42927244
AB - High-resolution residential telemetry, including smart thermostat data, is increasingly used to develop data-driven models for building energy analysis and HVAC performance assessment. However, field telemetry often lacks detailed information on building characteristics and known physical parameters, limiting rigorous model development and validation. This article presents a U.S. synthetic residential building telemetry dataset generated using EnergyPlus that combines high-resolution time-series telemetry with fully known building, envelope, infiltration, and HVAC parameters. It comprises 540 unique models of single-family detached and attached homes across the 48 contiguous states and 13 ASHRAE climate zones. The models were generated using probability distributions of building and HVAC characteristics derived from the ecobee Donate Your Data (DYD) program and the U.S. residential building stock represented by ResStock. Statistical case weights are provided for the 540 unique models, defining a probability-weighted analytical population of 1200 residential cases. For each model, the dataset provides an EnergyPlus input file and a full year of synthetic telemetry at 5-minute intervals in Parquet format. The telemetry includes indoor and outdoor environmental conditions, thermostat setpoints, HVAC operation and runtime, infiltration and ventilation, heating and cooling rates, and energy-consumption variables. The dataset also includes building characteristics, simulation metadata, and weather-station location. Comparison of HVAC-runtime patterns with ecobee field telemetry indicates that the synthetic data reproduce broad climate-dependent patterns in residential HVAC operation while retaining known model inputs and physical characteristics. The dataset can support the development and benchmarking of surrogate and telemetry-based models, evaluation of retrofit and HVAC-sizing scenarios, and analysis of residential energy performance at national, state, and climate-zone scales.
SN - 2352-3409
ER -