Addressing the building energy performance gap with measurements

An extensive amount of research, development, regulation, and investment has gone toward reducing energy consumption of building HVAC systems. Nevertheless, actual building energy consumption is often significantly higher than that planned at the design stage. This difference is sometimes referred to as the building energy “performance gap” (de Wilde 2014). Spitler and Gehlin (2019) reviewed some of the work exploring the performance gap based on energy-use intensity (EUI). The results are eye-opening, to say the least. Scofield (2013, 2009) showed no energy savings on average for LEED-certified buildings compared to comparable buildings in the U.S. commercial building stock. Geng et al. (2019) found similar results for buildings in China and the United States. Kurkinen et al. (2014) also found significant performance gaps in Swedish buildings.

These studies have used energy use intensity as a metric because it is relatively easy to measure. While EUI can be very useful in showing that problems exist, it has inherent limitations with regard to identifying the cause of the performance gap. For example, it is difficult to differentiate the effects of a poor building envelope from those of a poorly operating heating and cooling system.

One path forward is measurement of the actual system performance. Performance measurement of commercial-building-scale ground source heat pump systems is the subject of an ongoing international collaborative research project, IEA HPT Annex 52 (IEA HPT 2019). The activities of the annex include defining a more comprehensive boundary schema for system performance factors, developing instrumentation and analysis guidelines, and preparing a set of about 50 case studies of real-world performance measurements. There have been a number of interesting and sometimes counterintuitive findings in the work to date. From my own perspective, the degree to which part-load effects dominate the performance in some systems has been surprising.

But whither goes the rest of the heating, ventilating, and air-conditioning industry? While the developments in performance measurement of ground source heat pump systems are interesting, it is not clear that there is any activity in measuring performance of other system types. Several years ago, I looked at length, but to no avail, for published measured performance of conventional variable air volume systems with a chiller and boiler. I would be quite pleased to find out that I did a poor literature search, and that there are publications that I’ve missed that could enlighten me on how system performance varies with load and temperature, and how much of the system energy over time is consumed by the distribution subsystem versus the chiller and boiler. I would be happy to hear from readers regarding publications I’ve overlooked.

Arguably, while this information could be quite valuable for a single building, widespread knowledge of system performance will have much more significant benefits. With continuing developments in the Internet of things, machine learning, and data analytics, one can imagine many ways that this information might be used. Imagine a building owner knowing that their system for the last year ranked 50th of 55 buildings with similar systems in the same area. By comparing system performance factors and other metrics, the capability to detect faults should be greatly increased. There are, of course, hurdles to be overcome, such as sharing of information.

But before we get to that hurdle, there are other hurdles to overcome—such as how to make measurements that are robust and not overly expensive on a wide variety of systems. That leads, in turn, to another hurdle—how to encourage sharing of knowledge in this area. I believe Science and Technology for the Built Environment (STBE) could be one part of the answer.

From the perspective of a scientific journal such as STBE, research on performance measurement often falls into the category of “case studies,” which the journal doesn’t generally publish unless these lead to generalizations that can be applied with confidence to new situations (Spitler 2018). However, articles that make new contributions in methodology or analysis are certainly welcome, as are articles that lead to new insights. I’m considering organizing a special issue on the subject. If you have a potential contribution, please let me know of your interest.

Jeffrey D. Spitler, PhD, PE, Fellow ASHRAE, Fellow IBPSA Editor-in-Chief, Science and Technology for the Built Environment Regents Professor and OG&E Energy Technology Chair School of Mechanical and Aerospace Engineering Oklahoma State University Stillwater, Oklahoma, USA