Prevention of compressor hunting in DX rooftop unit through supply air temperature optimization

A rooftop unit (RTU) is typically referred to as a unitary system, which employs direct-expansion (DX) air-condensing equipment for cooling and then provides the conditioned air to the space through dedicated air distribution system. Over the years, the DX RTU has wide applications due to its greater flexibility, lower installation cost, and more useful space. System challenges may include reliability, noise, vibration, and humidity control; however, most of these issues have been resolved by technology advancements. One challenge, the compressor short-cycling, has not been successfully solved with satisfactory partial load performance and minimum initial cost. This excessive on/off cycling drastically reduces the compressor reliability and increases space humidity. The problem is attributed to the system load modulation not matching the building load variation. Although significant progress has been made to resolve short-cycling, current methods cannot provide a satisfactory solution. Dr. Mingsheng Liu (2004) developed and patented (US60/701,608) an advanced control algorithm to resolve the DX RTU\u27s short cycling. This algorithm is based on the fundamental vapor compression theory that different evaporating temperatures will produce varied refrigeration effects. Elimination of the system cycling may be achieved by optimizing supply air temperature reset to match building load condition. This research develops and justifies this advanced control algorithm. A mathematic model has been built to simulate the system performance loads for different weather conditions. Results show short-cycling to be completely eliminated under various operating conditions with no extra energy cost under higher load or humid weather condition and significant savings under the lower load or dry weather condition. In addition, field investigation has been conducted in a functioning office building, which further proves the effectiveness of this advanced control algorithm