13 Jul Lies, Damned EUIs and Statistics – The Pitfalls of Relying on EUIs for Benchmarking
There are three kinds of lies: lies, damned lies, and statistics. – Mark Twain.
One of the big problems with energy efficiency is that it’s complicated. I feel like there’s a big need to simplify it for clients so they can make the most informative and cost-effective decisions about their buildings. But sometimes simplicity results in oversimplification, causing us to make the wrong decisions. In that vein, let’s examine one of our industries sacred cows, the Energy Use Intensity (EUI), and how its usefulness is more limited than we might think.
What is an EUI and why do we use it?
An EUI is simply an expression of the energy use of a building, divided by the floor area as a means of normalization. The idea is that you can compare buildings of varying sizes with industry benchmarks and get an idea of how efficient your building is. This is the basic idea behind the EPA’s ENERGY STAR® certifications, and a plethora of benchmarking sites like LBNL’s building performance database (BDP).
The great thing about BPD is that it also lets you filter your sample group by location, weather, building type, etc. You can even compare the EUIs of similar types of buildings in different states. For example, in the graph below it appears the BDP sample of data shows grocery stores in California (in purple, N=254) use more energy per square foot than those in Illinois (in yellow, N=50).
BDP Plot of Grocery Stores in California and Illinois
The idea of benchmarking is a cornerstone of energy efficiency for good reason. In particular, if you have a portfolio of buildings that have similar purpose, like a chain of drug stores, comparing EUIs can tell you a lot and the “energy hogs” certainly stick out of the crowd. One of my favorite plots demonstrating this is this graph shown below. It comes from kW Tracker™, our in-house portfolio management platform.
Energy Spend vs EUI from kW Tracker™
The graph above shows not just the EUI (y axis in kBtu/sf), but also plots the total energy spend. That let’s us focus not just on high EUIs, but on the ones that really matter (i.e. the top right corner) where the EUI is high, and the site is large.
To the EPA’s credit, additional factors are taken into account to make ENERGY STAR® ratings more meaningful – not just square footage. Those factors depend on the building type and the statistical validity of influencing factors. For example, for office buildings the score also includes an adjustment that depends on the operating hours, number of workers, number of computers, etc. ENERGY STAR® and Portfolio Manager® use the 2003 data from the U.S. Department of Energy’s periodic Commercial Building Energy Consumption Survey (CBECS or “sea becks”).
The value of EUIs is clear, and I’m happy that we are seeing the propagation of building disclosure laws that require reporting building energy use regularly, or at time of sale. That information helps drive awareness of energy efficiency and helps building owners understand the impact of energy efficiency on value in real estate purchasing decisions.
The Problem with EUIs
There lies a fundamental problem in the calculation of an EUI which makes them an incomplete way of measuring energy efficiency in buildings. The idea is just as flawed as it is simple. Here’s why.
The EUI was intended to give us a quick read on the energy efficiency of a building by comparing it to peers. The problem is, there are so many differences in building performance, and many factors contribute to those differences. So finding a statistical model for energy use based on sampling and identification of independent, valid variables is illusive.
Small Data Set
The best comprehensive sample of buildings to figure out those relationships is CBECS, and that data set isn’t nearly big enough. It is the biggest collection of data we have that isn’t self-reported (i.e. probably incorrect). Yet even so, the field surveys funded by CBECS still number in the thousands, and given the differences in building types, geography, ages, etc. that sample size just isn’t large enough to extract regression coefficients for things like whether the building has setback thermostats, or chilled water vs. VRF cooling (fancy engineering speak for building equipment factors that impact its energy use). Bottom line is that our industry isn’t nearly as well-researched as people might assume it is.
The Impact of Occupancy on Energy Use
Measuring efficiency is really, tricky business and trying to boil it down into one number, while desirable, is going to be kind of wrong. One of the biggest failings is the relationship with building occupancy. Credit to Don Winston and Dennis Landsberg for pointing this out to me, based on their experiences in New York City (NYC).
The NYC real estate market is one of the most expensive on the planet, which causes people to rethink the way they use buildings. At thousands of dollars per square foot, you’d be foolish not to put as many desks in an office as you could. As a result, occupancies in NYC tend to be very high with more people in a space. More people means more plug loads, more cooling loads, more ventilation loads – a lot of very legitimate factors that don’t typically get figured into an EUI.
ENERGY STAR® takes occupancy into account but what I hear from my colleagues a full building doesn’t look much better than an empty one. We like to encourage folks to fully utilize a space, but a simple EUI just makes an empty building look better.
Another complication arose when we were designing the reporting forms for Standard 211, the latest ASHRAE standard defining energy audit levels. In the Standard, we required energy auditors to report an EUI for the site but soon realized we needed to be clear about what that EUI represented, especially when it comes to how we handle on-site generation.
Is a building really energy efficient if its offsetting excessive energy use with on-site generation?
This building has solar on the roof which is probably making the EUI for this building look better. But the building also has “window shaker” air conditioners and single-pane, louvered windows. Not an energy efficient combo at all. Yet this building might have a great EUI, if the energy produced by the solar panels offsets the consumption.
In theory, you’d like to subtract that solar generation out, so that you have a measurement for the building itself, not just the grid impact. However, you can’t always do that – most solar installs in California for example are net-metered. If you can’t meter the solar separately, you can’t exclude the generation from your calculation. In Standard 211 we borrowed from ASHRAE Standard 105 which defines separate EUIs for “building” or “site” depending on whether the on-site generation was added back into the EUI (as in the “building” EUI) or simply displaced apparent consumption (as in the “site” EUI). For a solar system, for example, the EUISITE will be low, and the EUIBLDG will be the same as if the solar wasn’t there.
Will New CBECS data help or hurt?
Word on the street is ENERGY STAR® will be updated with new CBECS data sometime in August 2018. At that point, the more recent 2013 survey data is already 5 years old (ahem why does it take so long to compile anyway?). So, keep your eye out for that update and consider how your building’s score will be impacted.
When the new CBECS data update hits Portfolio Manager®, your score is more likely than not going to go down. The EUIs of building stock have overall trended downward over the last 15 years. So when those CBECS averages come down and your building still is what it is, its score is going to look worse. The good news is that the EPA doesn’t have the funds to uninstall your ENERGY STAR® plaque – so you can continue to sport that with pride!
Use EUIs to Your Advantage
Simplicity in energy efficiency has its pros and cons. It can encourage people without a background in the nuances of buildings and energy efficiency to easily identify potential energy offenders and consider projects that lower energy use, which is very beneficial to the planet and their bottom line. As with any scoring system, it’s important to know where the potential pitfalls lie. Doing so allows you use the value in the manner it was intended and make informed decisions about operating your building efficiently and economically. If you need any help calculating your EUI, understanding it or managing your portfolio be sure to contact us.