University of Oklahoma researchers have devised a mathematical formula to monitor energy use in HVAC units that can be used to detect unreliable systems and faulty equipment that affect energy consumption.
The formula, devised by a team led by Li Song, University of Oklahoma assistant professor in the College of Engineering’s School of Aerospace and Mechanical Engineering, uses existing output data, such as pump speed and power to monitor the units to create virtual sensors to identify energy waste at the air-handling unit as well as at a whole-building level.
Nearly 95 percent of all US companies don’t monitor their building energy efficiency due to a lack of awareness, existing infrastructure restrictions or the prohibitive cost of commercial monitors, according to the University.
Besides saving companies money on utility bills, Song’s formula is a low-cost option to commercial monitors. Song estimates one ultrasonic flow meter, which monitors water pump performance, could cost as much as $5,000, and an organization would need to buy several monitors to get an accurate picture. Song’s virtual process uses little to no hardware and is within a plus/minus 2 percent degree of accuracy compared with commercial meters.
Prior to developing virtual sensing, Song implemented the energy monitoring and operation fault detection and diagnosis manually in more than 100 buildings with cumulative savings exceeding $70 million. In one building alone, her method reduced annual electricity consumption by 53 percent, electricity demand by 21 percent and gas consumption by 49 percent in one year. In another building the structure qualified as an Energy Star building five months after the evaluation.
Song’s research, which has focused on corporate efficiencies, is now entering the government sector. Her research team was recently awarded a three-year, $1 million contract from the US Department of Defense to increase building efficiencies at military installations. Her first analysis focuses on Tinker Air Force Base in Oklahoma City, Okla.
