Improving HVAC Performance Through Spatiotemporal Analysis of Building Thermal Behavior

Key

3619

Conference Year

2014

Authors

Michael Vincent Georgescu, University of California, Santa Barbara, United States of AmericaFollow
Igor Mezic, University of California, Santa Barbara, United States of AmericaFollow

Keywords

Building management, Data visualization, Fault detection, Monitoring, Energy Audit

Abstract

HVAC systems often make up half of the annual energy consumption of a commercial office building. Because of this, optimizing thermal factors which influence building energy efficiency is a prioritized task in the development of high performance buildings. The optimization is often achieved through a combination of effective envelope planning during building design, selection of efficient HVAC equipment during construction, and the specification of control loops and operational sequences during building commissioning. Unfortunately, these three tasks are often performed by separate practitioners with little communication between each other. Building performance decreases as a result of the original design intent being incorrectly implemented. Ideally, a building operator would like to know whether their building is comfortable, functioning normally, and running efficiently, but due to the time required in performing this assessment, often these questions may remain unanswered until a major problem arises. In building operation, an increased level of instrumentation can facilitate additional sophistication in building analysis and control, but techniques are required aggregate data into actionable information. In this work, an energy audit of a low EUI LEED silver building is performed by decomposing building data into spatiotemporal modes. These modes quickly identify spatial structures in building data and help identify influences of heat loads from the environment and internal sources such as equipment usage and occupancy. This technique has previously been used to analyze the thermal behavior of building models, but has seen only limited implementation using actual building data. Using this technique, examples of energy waste are identified including faulty equipment operation, unnecessary equipment usage, and HVAC operating conditions requiring high energy to maintain either due to improper initial commissioning or post occupancy retrofits. Through addressing the issues discovered, building energy usage is reduced by 16.5%, occupant comfort is improved. Both improvements are achieved solely through changes in building operation.