UC Davis

Oversizing, the common engineering practice of specifying devices with capacity exceeding the actual load requirement, is a widespread practice across virtually all building technologies end-use categories, including HVAC, electrical systems, lighting, appliances, and plug loads. This practice, driven by factors like design uncertainty, institutional pressures, and risk aversion, results in wasted capital investment, control difficulties, and excessive energy consumption due to inefficient part-load operation. Part-load operation, where devices run below maximum capacity, is the dominant operating mode in most energy systems and presents a complex design challenge. Solutions to match output to load fall into three broad categories: constraining the output, adjusting the device’s internal behavior, and linking output to energy storage or other waste-heat reuse applications. The energy implications of part-load are critical, as efficiency often drops sharply as load decreases across a wide diversity of devices. To quantify the extent of this problem, we derive a dimensionless Part-Load Metric (PLM) based on device efficiency and its frequency distribution of operating hours at various output levels. The PLM quantifies the deviation of a device's actual efficiency from its maximum design efficiency. This metric also serves as a measure of "capital inefficiency," enabling engineers to compare the impact of different part-load solutions and providing a unified framework for evaluating performance across various devices and systems.