Benchmarking Datasets of Building Environmental Conditions and Occupancy Parameters

Lead Performer: National Renewable Energy Laboratory (NREL) — Golden, CO
Partners:
-- Pacific Northwest National Laboratory (PNNL) — Richland, WA
-- Oak Ridge National Laboratory (ORNL) — Oak Ridge, TN
-- Lawrence Berkeley National Laboratory (LBNL) — Berkeley, CA
FY19 DOE Funding: $1,000,000
Project Term: October 1, 2018 - September 30, 2022
Funding Type: Direct Funded

Project Objective

Optimizing the operations of buildings requires access to data from a variety of relevant sensor types including temperature, humidity, occupancy, air quality, and irradiance. This information when taken at a high frequency can establish comprehensive datasets that portray building operation and performance. Correlations in the data can suggest least-cost pathways to accomplish tasks like nonintrusive load monitoring, virtual sensing, building energy model calibration, forecasting, benchmarking, control optimization, fault detection, and many others. However, there is a lack of publicly available, high-resolution sensor datasets with broad applicability to a variety of high-impact use cases.

This project aims to develop the experimental plan for a subsequent multi-year effort in the collection and curation of high-quality, well-calibrated datasets of building operations through monitoring of equipment health, environmental variables, and occupancy parameters. Through a “more data than needed” approach, the project team will identify the parameters that are most valuable for a variety of purposes including load forecasting and baselining, virtual sensing, building energy modeling, building performance benchmarking (at the whole building and system levels), and non-intrusive load monitoring.

Project Impact

This project will help determine the level of resolution required for most effectively optimizing building operations through advances in data analytics and control technologies, as well as spur innovation in the research community through a common benchmark to develop from and evaluate against. These newly collected datasets can assist in the identification of low cost pathways to improve building operation and performance and lead to significant cost and energy savings.

Contacts

DOE Technology Manager: Michael Specian
Lead Performer: Tianzhen Hong, Lawrence Berkeley National Laboratory (LBNL)

Related Publications