One of the most consistent, most frustrating problems in the building operations world is just how infrequently buildings in the real world live up to the operational efficiency promised on the drawing board. In real estate, this is so common that we’ve given the discrepancy a name: We call it the “efficiency gap”. Nailing down just how much buildings underperform their expectations is difficult because there is so much variance, but one study found an average figure of 34 percent. Where is this discrepancy coming from and why is it so rarely taken into account?
Part of what makes these questions so difficult to answer is that the performance gap in most buildings is the result of several interplaying factors, all of which may have a small or large effect, depending on the specifics of the building. Researcher Ramallo-González classified the main causes of the performance gap into three main categories: environment, workmanship, and behavior.
The usual suspects
In terms of the gap we can already see, environmental uncertainty relates to our use of imperfect weather models to design or evaluate a building’s expected energy use. At the evaluation stage, it is common to use “synthetic weather years” that are built using data that does not represent a real year or that has been generated using data from the closest weather station and not the exact location of the building. Either of these approaches introduces uncertainty that can negatively impact the design process or the accuracy of an energy efficiency prediction.
The good news is that the effect of imperfect weather models is generally small – in the range of +/-5% for Washington, D.C., according to researchers Kershaw and Coley, with only a relatively small skew toward overestimating efficiency. Other factors play a significantly larger role.
Sticking with environmental factors for a moment, though, Ramallo-González also classifies uncertainty due to climate change as a factor that contributes to the efficiency gap. Most buildings are designed to last a significant length of time. In the U.S., almost 40 percent of commercial buildings are over 50 years old, and over a quarter were constructed before 1959. Over time frames like this, especially going forward, the climate may shift so that a building that is efficient now finds itself in an environment that it is no longer suited for.
The next category, workmanship, takes into account both the ability of builders to accurately construct the building described in the design phase and the materials used. Of particular importance is the amount of thermal bridging that takes place in a space and the actual r-values of materials compared to the expected r-values. In a research paper with the lengthy title Evaluating the impact of an enhanced energy performance standard on load-bearing masonry domestic construction, researchers consistently found thermal bridges that were not anticipated in the design phase.
Thermal bridges are materials or spaces with low insulation ratings that are found among materials with high insulation ratings. The classic example is a wooden stud in an exterior wall. The rest of the wall is typically filled with insulation, but heat can pass more easily through the stud. This allows the heat to transfer out of the home as if the stud were a bridge.
Finally, occupant usage is a major variable that is difficult to account for in the design phase. This last point will surprise precisely no one in the real estate world. Occupants leave windows open when they shouldn’t, leave lights on overnight, and generally make accurate predictions difficult. The engineer’s refrain that buildings would run perfectly without the people in them comes to mind.
Kershaw and Coley found that the variability in performance due to operation parameters linked with occupants’ behavior could be as high as 79 percent in San Francisco and 57 percent in Washington, D.C.
Another, somewhat more disturbing source of the performance gap is a failure on the part of the architects, engineers, and energy consultants who make or use energy performance models. The same David Coley from the study cited above, working with Salah Imam and Ian Walker, wrote a controversial paper titled “Are Modelers Literate?” that looked at 108 of these professionals and their ability to rank improvements to a home according to how they would impact efficiency. As you might imagine from the title, the results were not encouraging.
The paper found only a slight correlation between the variables chosen by modelers and the variables that are objectively important. Even more concerning, when modelers were grouped into clusters, a group of about 25 percent was found to have performed “worse than a person responding at random.”
The study attracted some criticism from industry insiders, one of whom claimed that the sample size that Coley and his co-authors used included too many junior professionals. Of the 108 members of the study, about two-thirds had five or fewer years experience in the industry. Coley, Imam, and Walker contended that their more qualified test subjects failed to perform better.
If modelers aren’t able to create effective models for a building’s energy use, it wouldn’t be at all surprising to find significant performance gaps.
So, what is to be done?
The causes of the performance gap are relatively easy to point to. Solutions are a little bit more elusive.
To fix the long-term problems involved with the energy efficiency modeling industry, we need better data and better training. Weather data gets better all the time, but there are still large areas where the resolution of weather data is too low for effective models. There are also other factors that affect how a building will perform in a microclimate. Coley told Yale Environment 360 that one building he visited to investigate a particularly serious performance gap, had made use of tinted windows and shading to reduce solar gain. The designers had created their model based on a classroom on a flat, sunny landscape, despite the fact that their building was in a shady valley. As a result, lights had to be on virtually all day.
In terms of workmanship, more work needs to be done to understand how designs will perform in practice, particularly with regard to thermal bridging. The name of the category seems to imply that the builders are the cause of the efficiency gap in this area, but in reality, a failure to account for how a building will work in real life is evidence of deficient models. The same is true for occupant behavior. Basing models on how the building might perform with ideal occupants is like basing vehicle safety standards on the notion that drivers won’t crash. It’s unrealistic.
Finally, stiffer education and accreditation requirements for energy modelers may help to improve modeler literacy, but Coley thinks that the real problem is feedback. He acknowledges that this area of the industry is very new and that large accreditation requirements up front might put a strain on the number of people qualified to make and use energy models at this point. Instead, he thinks that the first change should be a shift toward requiring modelers to follow up once buildings are constructed and see how the models compare to actual performance. Over time, this would allow the models to be tweaked until they can better predict final performance.
Buildings are a massive source of energy use and greenhouse gas emissions – they account for about 40 percent in each category. The energy modeling industry isn’t going away, and in fact, it’s essential. To fairly serve the real estate industry, though, it will need to improve on its performance gap problem.