Duct System Energy Performance in Large Buildings

Despite the energy-saving potential indicated by studies, relatively little work has focused on the energy savings of sealing duct-system leakage in large commercial buildings— particularly leakage downstream of variable air volume (VAV) boxes. This study seeks to develop an accurate means of estimating impacts of supply duct-system leakage on energy use in commercial buildings with VAV systems, focusing on central-system cooling and outdoor air (OA) preheating. The goal is to provide a basis for the quantified justification (known as a work paper) to solicit regulatory approval for a utility program that addresses duct-system leakage in commercial buildings with VAV space conditioning systems. Current Title 24 building codes (2016) only address duct leakage in buildings less than 5,000 square feet that have constant volume, single zone, space-conditioning systems. The project team developed a simplified method for characterizing impacts of weather on cooling and preheating outdoor air. The study accounted for duct-system leakage upstream and downstream of VAV boxes, as well as the fraction of OA drawn into the supply fan. The team calculated the required cooling and preheating of outdoor air with and without duct leakage for three California locations, categorized as hot (Palm Springs) and moderate (Riverside and Sacramento). This study is based on a simplified model that uses a variety of inputs—hourly weather data, leakage level, OA fraction, and fan flow profiles—to estimate heating and cooling loads associated with the intake of OA into a typical office building. The study analyzes the energy and peak-demand implications of supply system leakage on outdoor-air conditioning, as well as leakage impacts on fan power and fan heat removal. The model assumes a fixed OA fraction, except when operating in economizer mode. Energy and savings calculations were based on the difference in outdoor airflow due to duct leakage, combined with the enthalpy difference between return air and OA for cooling, and the enthalpy difference between mixed air and supply air for heating. Results assumed an office building with 28,000 cubic foot per minute (cfm) design airflow to the conditioned zones and use of a system that provides cooled air to the supply plenum all year. Results indicate that 10% pre-VAV leakage at design flow combined with another 10% postVAV leakage flow leakage translates to an additional 9,700 kilowatt-hours (kWh) of electricity consumption for OA cooling for a roughly 33,000 square foot (ft2) building in Palm Springs. For such buildings in Riverside and Sacramento, the additional OA cooling electricity consumption is 2,800 kWh and 2,500 kWh, respectively. For a building with 70% OA (e.g., a laboratory) in Palm Springs, the OA cooling energy use due to leakage increases to 34,000 kWh, requiring an additional 2,400 therms of natural gas to preheat OA. The team also compared the impacts of duct-system leakage on different aspects of system energy consumption and costs, resulting in a breakdown of energy-cost savings for a 20% OA, 28,000-cfm building starting out with 10% upstream/10% downstream leakage, 90% of which gets sealed. Overall annual cost savings ranged between $6,000 for such a building in Riverside and Sacramento to $7,000 for such a building in Palm Springs1. A similar breakdown for the same building, but with 70% OA (i.e. laboratory), showed a much greater savings potential (particularly for OA cooling in Palm Springs) and a non-trivial contribution associated with natural gas use for preheating the mixed air entering the cooling coil. A sensitivity analysis of the combined effect of leakage and OA fraction on OA conditioning showed that climate and OA fraction strongly influence the preheat and OA cooling energy impacts of leakage and this impact increases with OA fraction. For preheating, there is no impact of leakage below a climate-dependent minimum OA fraction; this minimum OA fraction depends somewhat on system leakage rate. The team used the simplified model results to roughly estimate the energy and peak demand savings potential for the SCE service territory (Table ES-1). They assumed an average leakage of 19%, and that the sealing analyzed applies to large offices, colleges, and health-care buildings in SCE service territory. These sectors nearly (but not perfectly) match the application range of the technology.