Your goal is to reduce the energy or, perhaps, the environmental (emissions) footprint of your buildings. Your budget is limited, and you need bang for the buck. Where do you start?

A process that benchmarks building energy performance can steer you in the right direction, fast. The process is easy to accomplish and offers many advantages, including:

• achieving a quick performance read on your buildings
• identifying the high, medium, and low performers within a portfolio so you know where the best opportunities lie
• allowing you to target your energy evaluation/improvement resources
• baselining your expectations for your energy audit or upgrade up front
• providing easy performance tracking. 

Owners should no longer send audit teams to buildings just because the buildings are large or use a lot of energy. If some of these buildings happen to be high performers, your auditors would have audited them even though they were really low priority buildings. Benchmarking allows you to make smarter decisions up front and gives you the opportunity to better target your program before an auditor ever sets foot on the premises.

Historically, benchmarking consisted of simply comparing building energy use intensity (EUI) to an average EUI for a group of similar buildings. This is about as crude of a benchmarking metric as you can get. It only tells you that the building functions above, below, or at an average level of performance. You have no way of judging how well or how poorly the building performs, just that it does so above or below average. In addition, if the average itself is an unreliable gauge (which is often the case), this sort of benchmarking accomplishes even less.

A much better approach, when there is a sufficient quantity of observations from a group of similar buildings, is distributional benchmarking. The figure below shows an example EUI benchmarking distribution for 50 buildings. 

The EUI for the  building in question is 120 kBtu/sqft-yr. While the building would appear to be close to average for the entire distribution, the percentile bars show that a full 80% of the other buildings have a lower EUI. This means that the benchmarked building is in the lowest performing 20% of its peers and is most likely a great candidate for improvement. This approach allows you to immediately spot the most extreme performers, which is exactly what you are looking for if you manage a portfolio of buildings and need to strategically allocate limited resources to have the greatest overall impact.

Perhaps the most comprehensive, simple, and popular benchmarking tool in use today is EPA’s Energy Star Portfolio Manager. It relies heavily on distributional benchmarking similar to that described above. It is more advanced, however, in that it can normalize multiple drivers that influence the energy use of buildings, such as operating hours, occupant density, climate, and other variables, depending on the building type. This capability is critical to reliably benchmark many buildings.

This multi-variable normalization strategy for benchmarking buildings was pioneered at Oak Ridge National Laboratory some time ago, and while it is not necessary to achieve a reliable performance indicator for many and possibly even most buildings, there are plenty of cases in which it is necessary.  The keys to a robust and capable benchmarking system are that it be easy to use, require easy-to-acquire building information, and be a reliable indicator of building performance—each of which are features of Energy Star’s Portfolio Manager. Few available benchmarking tools have this complete set of capabilities. For example, Portfolio Manager can currently provide performance ratings for the 15 commercial building types listed below.

A number of webcasts archived on the Solution Center cover Portfolio Manager’s functionality, including those presented on January 20, 2011 and March 30, 2011. For questions or comments regarding this topic, please post your thoughts to the Blog, or you may request technical assistance from our team of experts.

Content for this Blog post courtesy of Terry Sharp, Oak Ridge National Laboratory