Residential HVAC Fault Data Collection Plan – Refrigerant Undercharge and Overcharge Faults

Technical Report · Fri Mar 01 04:00:00 EST 2024

DOI:https://doi.org/10.2172/2338269· OSTI ID:2338269

Yoon, Yeobeom ^[1]; Jung, Sungkyun ^[1]; Im, Piljae ^[1]; Luo, Junjie ^[1]; Sharma, Vishaldeep ^[1]; Safir, Islam ^[1]

---

1. Oak Ridge National Laboratory (ORNL), Oak Ridge, TN (United States)

Heating, ventilation, and air-conditioning (HVAC) systems can develop faults due to poor installation practices or gradual wear and tear, leading to decreased HVAC system’s efficiency, compromised thermal comfort, and shortened equipment lifespan (EERE, 2018). Automated fault detection and diagnosis (AFDD) technologies offer a solution by identifying energy-wasting HVAC faults, such as inadequate indoor airflow and incorrect refrigerant charge, and guiding technicians to enhance system efficiency. In the realm of residential HVAC, AFDD can be implemented through various fault detection and diagnosis capabilities, sensor configurations, and target applications. These technologies typically fall into three categories: smart diagnostic tools, original equipment manufacturer (OEM)-embedded tools, and add-on tools. Smart diagnostic tools employ temporarily installed sensors to directly measure HVAC system characteristics, while OEM-embedded tools utilize factory-installed sensors to identify faults or assess system performance. However, both these types of AFDD technologies are often only accessible for high-end HVAC equipment or require additional sensor installation by qualified technicians, resulting in high investment costs and limited applicability for low-income residential buildings. On the other hand, add-on tools rely solely on data from smart thermostats and meters to detect faults by continuously analyzing equipment runtime or energy usage. As smart thermostat and meter costs decrease and their prevalence increases, these tools can be readily deployed in low-income residential buildings. However, they possess limited capabilities as they rely solely on basic trend analysis. Enhancing such tools with advanced machine learning algorithms can significantly improve their effectiveness.

• Research Organization: Oak Ridge National Laboratory (ORNL), Oak Ridge, TN (United States)
• Sponsoring Organization: USDOE Office of Energy Efficiency and Renewable Energy (EERE)
• DOE Contract Number: AC05-00OR22725
• OSTI ID: 2338269
• Report Number(s): ORNL/TM--2024/3347
• Country of Publication: United States
• Language: English

Related Subjects

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATION