Towards Measured Data
Sustainability in architecture has always been based on metrics. The idea is that buildings should achieve a certain numeric threshold to be considered “sustainable” or that metrics can be used to compare project performance. While metrics have been invaluable in applying rigor and objectivity to our understanding of the quality of architecture, these numbers serve only as a proxy for the outcomes we aspire to achieve. Health, climate, and community are now the goals of architecture. The challenge is in determining which metrics help us get there.
Energy Use Intensity (EUI) is the classic sustainability metric and has long been used to represent project performance. Though only touching on one measure of sustainable design, energy, EUI’s use has been widespread because it’s easy to calculate, easy to benchmark, and easy to understand given its typical whole number, 1 to 3 digit range. The problem with using EUI to determine performance for awards or certification is that it’s generally predicted or modeled (pEUI), meaning it’s not real. For architects to become serious about building performance and to realize the impact that we claim to be aiming for, the profession will need to shift its focus from predicted metrics to measured metrics, beginning with EUI.
Ultimately, it's not predicted energy use that will cool our planet, but the actual reduction in carbon that our buildings are able to achieve. This year, the COTE Top Ten awards is leading this charge and for the first time, requiring that submitting firms report measured data for their projects. This change will ensure that project performance is judged accurately and that projects are awarded for their real-world impact rather than for the outputs of their energy models. Reporting pEUI is welcome, but it’s now optional and subsidiary.
For too long, the profession has been awarding and lauding buildings based on self-reported estimates that may or may not have any bearing on reality. Energy modeling is a valuable tool for informing design decisions, but not effective at predicting a building’s actual future energy use. Modeling is based on numerous assumptions and outputs will vary greatly depending on the knowledge, experience, and incentives of the modeler. Furthermore, the self-generated and self-reported nature of pEUI invites gaming of the system, likely leading to design juries unintentionally elevating energy hogs as examples of projects to emulate.
There have been frequent arguments for maintaining the current focus on predicted metrics, but none of these hold up to scrutiny. One argument often posed is that predicted data is a more appropriate judge of performance since it captures the intent of the design team and that judging measured data would be unfair as it includes factors outside of the architect’s control. To apply a sports analogy, this would be the equivalent of awarding the team that trained harder rather than the team that actually won. Since a narrative, not a metric, is the best judge of design intent and because pEUI has as little correlation to intent as it does to actual performance, judging a project by its pEUI is actually even more bizarre than the above example. It would be the equivalent of awarding the team who says they trained harder, regardless of what they actually did or who actually won.
Another popular argument against measured data is that it’s too hard to acquire and that requiring real numbers would reduce awards submissions or participation in programs such as the 2030 commitment. Many AIA chapters depend financially on award submission fees, so adversity to change is understandable, but I've yet to see any data to support this fear and I very much doubt that it's true. After all, it's the energy model that's onerous for many firms, particularly small firms.
It's much easier to add up 12 months from utility bills and divide by the GSF than it is to run an energy model. It’s also faster, cheaper, more accurate, and more meaningful. All built projects have actual energy, but only a small fraction had modeled energy beforehand, and only a smaller fraction had modeled it well. It's likely that swapping out predicted metrics for measured metrics will increase accessibility and thus increase the number of submissions. In a world where measured data is the most valuable currency, you don’t need unique skills or higher fees to play the game, all that submitting firms would need to do is pick up the phone. As the profession shifts from predicted to measured data, both quality of projects awarded and the accessibility of recognition should improve due to the realignment between incentives and the outcomes we all want to achieve. The rest of the profession would be wise to follow COTE’s lead and begin to recognize projects based on their real-world impact.
Notes:
EUI, or Energy Use Intensity, is defined as kBtu consumed per gross square foot of a building area over one year or kBtu/sf/yr. To calculate measured EUI, simply sum the monthly energy use reported on 12 consecutive months of utility bills, convert kWh to kBtu (kWh * 3 41 = kBtu), and then divide this number by the building’s size in GSF.