Simulated Building HVAC Fault Detection Data from DOE Labs

LBNL, PNNL, NREL, ORNL, and ASHRAE RP-1312 created a dataset for testing automated fault detection and diagnostics algorithms in buildings. It contains simulated data for air-handling units and rooftop units generated using PNNL's large office building model. The dataset was last updated in April 2022.

Use Cases

• Test fault detection algorithms on simulated data for air-handling units and rooftop units.
• Benchmark diagnostic performance for HVAC systems using EnergyPlus simulation outputs.
• Analyze building performance under fault conditions for commercial building models.

Strengths

• Created by a consortium of leading national laboratories (LBNL, PNNL, NREL, ORNL) and ASHRAE.
• Specifically designed for evaluating fault detection and diagnostics algorithms in a controlled simulation environment.
• Includes data for two major HVAC system types: air-handling units and rooftop units.

Limitations

• Data is simulated, not measured from real buildings, which may limit real-world applicability.
• The dataset size, row count, and specific column features are not documented.
• Temporal coverage and geographic scope of the underlying simulation model are unspecified.

Provenance

Source

LBNL, PNNL, NREL, ORNL, ASHRAE RP-1312 (Drexel University), Department of Energy

Collection Method

Simulated using PNNL's large office building model with EnergyPlus.

Freshness

2022-04-26