Behind the energy model
Behind the energy model
June 20, 2016
|Marge Anderson, Seventhwave’s Executive Vice President, explores the energy modeling landscape during a Q&A session with Keith Swartz, PE, BEMP, LEED AP, Senior Energy Engineer.|
Keith Swartz shares the art of energy modeling with an engineering intern.
Accurate energy modeling relies on powerful software tools. What is your take on how the industry’s most popular programs are meeting the challenge?
There is a trend toward performing energy modeling early in the design process rather than as an accounting exercise afterwards. But some early design modeling tools are not suitable for more detailed energy models later in design. All tools have strengths and weaknesses. I’ve also found that powerful programs generally are not very user-friendly, and user-friendly programs are not very powerful.
Another problem we face is software not keeping up with today’s energy technologies. For example, many common programs do not model VRF (variable refrigerant flow) systems, an HVAC technology that is becoming more common in building design. Programs that do incorporate VRF typically model only those systems that do not have heat recovery.
Also, many programs do not allow you to easily modify the building’s architecture. And to further refine energy models, we need to incorporate better information about plug loads and infiltration. A software feature that could be helpful is feedback on sensitivity of inputs and uncertainty of outputs.
But software is getting better. For example, there’s progress toward an “easy button” approach to automatically generate a model for a baseline minimally code-compliant building. And some programs provide the capability to auto-calibrate energy models for existing buildings to past energy bills.
What’s cutting edge right now, and what do you expect to see in ten years?
Right now I’m seeing more use of EnergyPlus as the fundamental program and less dependence on DOE2. Trane TRACE is working on a new version that is based on EnergyPlus that is expected to be released this fall.
User interfaces for EnergyPlus are getting better, and integration with Building Information Modeling (BIM) is improving. I haven’t been doing this directly, but I have heard from others at the ASHRAE Energy Modeling Conferences that it seems to work well with simple buildings and less so for more complicated buildings. Inputs to BIM need to be suitable for export, but rarely are. It is still often easier to start from scratch than to clean up a model started by BIM.
In the future I expect there to be more automation of inputs, like auto-generation of baselines, and better translations from BIM.
There’s a lot of emphasis in the codes community about outcome-based energy codes. How will the role of energy modeling change if outcome-based codes take off?
Outcome-based codes could have a huge impact on the role of energy modeling. The most significant impact could be that energy modeling will need to focus more on predicting energy use instead of some other goal. Currently most energy models are used for LEED, code compliance, or comparing design options.
Some modeling rules take the model away from reality. Energy models to evaluate systems inside the building might not include, say, parking lot lighting because it doesn’t interact with the building’s internal systems. Outcome-based energy models will need to be more realistic and include everything that will spin the meter, and this will take more effort. The expected building operation needs to be better known at the energy modeling stage.
Although different from code compliance, some owners are using performance-based requirements in their design contracts, and this has a similar impact on how energy models are done.
In what other ways have you seen energy modeling improve buildings—not just in terms of energy?
Energy modeling fosters collaboration, informing better design decisions by bringing the design team together to discuss system interactions.
Energy modeling also contributes to the process of fine-tuning a building’s performance. Calibrating an energy model after a year of operation raises questions about why the model differs from reality. Finding answers to these questions can reveal problems with systems that have not been operating properly or reveal unrealistic assumptions in the model. Post-occupancy, learning about how the building is actually used helps energy modelers enter better inputs into their future models.
We’re seeing a lot more automated modeling tools popping up. What’s the best application of these automated tools vs. the “hand-crafted” energy model? And what role does engineering experience still play in the process?
Simplified, automated programs have fewer inputs, so a lot of assumptions are behind the calculations. To the degree that the actual building deviates from those assumptions, the results will be less accurate. For example, if the program assumes that windows are evenly distributed around a building, but in reality most of the windows are on the south side, then the program will understate energy savings from window shading.
The best use of automated modeling tools would be for high-level, early design decisions like building massing and orientation. They are also better suited to homogeneous buildings, not those with a mix of uses.
Engineering experience makes or breaks an energy model—much more so than the software used. A powerful program with lots of inputs loses its benefits when an inexperienced user doesn’t know what to enter. An experienced energy modeler knows what factors are important for a particular project. An inexperienced user would more likely spend time working to refine factors that are insignificant.
A lot of training and practical experience is needed to become a good energy modeler. What could college engineering programs do to prepare students for what they will face on the job?
This reminds me of a story I heard about the U.S. Army bringing soldiers from Fort Benning, Georgia to Alta, Utah in 1942 to train them to ski for winter warfare. Some were injured in the process and success was questionable, according to the ski instructor: “the end of that winter we came to the conclusion that maybe a third of them would become pretty good skiers, the middle third could get by, and the remaining third had better stay in the paratroopers and forget it.” It was probably more effective to train skiers to be soldiers.
This is how it is with energy modeling—it is better to teach software to someone who understands buildings than it is to transform a computer programmer into a buildings expert. Students need to learn how buildings use energy and understand the interactions that occur among building systems.
What advice would you give new engineers and architects who want to become effective energy leaders?
Know the various building energy systems and how they interact. Also learn the non-energy aspects of buildings like comfort, health, and costs of construction and operation. And finally, consider the financial constraints of a building project. Instead of seeking energy efficiency or renewable energy at any cost, view the challenge as determining the best long-term investment of the project’s limited budget.
Check out our collection of energy modeling resources—articles, tools, webinars and more.