Are you comparing apples to apples with your energy efficiency options?

Say you are a building owner, seeking to optimize your building’s performance and maximize your energy savings by implementing a few energy efficiency measures. You receive some bids from different building product salespeople—but they vary wildly in predicted energy and cost savings. How do you know which estimate is the closest to being accurate?

The answer is in the inputs. Pay attention to what assumptions are being made to calculate savings. One of the easiest ways to manipulate an energy performance model is to adjust the usage schedule of a building. Care must be given in setting up schedules in an energy model to avoid over or under predicting the performance of energy efficiency improvements.

In the process of developing Condense’s revolutionary energy modeling tool, we built hundreds of thousands of parametric models. We discovered that one of the most important factors in predicting energy usage is usage schedules. In multifamily buildings, a wrong guess about how people take showers can throw off your energy savings estimates.  In a typical 3-story apartment building in Austin, Texas, a difference of just 5 minutes in daily hot showers can throw total predicted energy use off by almost 1%.   (See the table below for an example.)

In office buildings, lighting schedules and heating and cooling setpoint schedules can have a significant impact.  For example on a 32,000 f2 office building in San Antonio, a simple oversight like leaving the lights on an extra hour each day can swing the total energy use by 1.7%.  (See table below for details.)

In schedules, there are lots of levers to pull to swing the predicted energy use.  Schedules in Energy Plus include not just mapping over every hour and minute of the year, but ability to adjust fractions of use (such as only 50% of lights are on in mid-day). Because of these layers of complexity, it’s easy to lose track of schedule adjustments, and it’s also easy to create compounding effects.

In a 2006 Stanford study, researchers looked at elementary schools in two different climate zones to study the impact of occupant behavior on predicted energy usage. The results concluded that predicted energy consumption varied by more than 150% using all high or all low values for what experts believe reasonably represents occupant behavior. Adjusting a single parameter in a model could impact model results by up to 40% over a run using “all medium” values.

Similarly, a study performed by researchers at LBNL in 2015 also focused on predicting occupant behavior in building energy models, using various occupancy schedules and environmental preferences. The study found that the energy consumption differed over 150% when usage schedule inputs were maximized and minimized for typical occupant behavior patterns.

So what does this tell us?

When it comes to evaluating building product alternatives, to get an honest performance comparison, you must use consistent schedules in modeling. A third-party validation tool for modeling alternatives, such as Condense, is critical in the pursuit to objectively compare different building performance products, apples to apples. Condense limits the ability of users to manipulate occupancy schedules, thereby avoiding the possibility of over or under predicting energy usage for design bias. Condense uses Energy Star recommended schedules for every energy model it produces. Condense also provides alternative schedules for special uses such as extended hour office occupancy, or for energy saving measures. But these are standardized and are not able to be edited in detail.

While the relevance of energy models lay in their ability to evaluate design alternatives rather than reliably predict energy performance, using consistent schedules to achieve those comparisons is key.