What architects should know about building modeling in early design
The 2017 chair of AIA's Building Performance Knowledge Community provides an energy modeling overview
What do architects need to know in regards to ensuring that our buildings perform as designed? This article will provide an introductory overview, touching on required benchmarking, gathering actual operating data for green building certification, relying on analytical testing rather than intuition, showing building owners the incredibly fast return on modeling services, and extending the value of the services we provide.
Performance-based design and early modeling
Energy modeling informs the design process, leading to a higher quality and more efficient building. Early design performance modeling allows the design team—including the owner and operator—to factor data into fundamental design decisions, which affects a number of interrelated building systems through comparative analysis. This information is not meant to predict energy use but rather to see which design option would lead to a more efficient building by reducing overall demand as well as peak heating and cooling load. Orientation, massing, window-to-wall ratio, and permanent shading devices can all be studied in the early design analysis process.
This analysis should be done prior to modeling whole-building systems energy use, as architects can better set and meet performance goals by employing computer models early. Modeling can help the design team to understand climatic data, test design options to make data-based decisions, discover integrated solutions, set project goals, and find first-cost savings. As an example, reducing the window to wall ratio, reorienting glazing, or using improved glazing might reduce cooling loads and save the project the cost of a chiller, immediately paying for the envelope improvements.
Market forces are already encouraging architects to deliver buildings that perform as designed. According to the Institute for Market Transformation, energy disclosure laws—which require regular reporting to the government or make energy performance data available when a property is leased or sold—are becoming more widespread. At the same time, reports about sustainably designed buildings missing their energy goals by a wide margin after occupancy are surfacing with regularity.
The industry is headed for more performance-based and outcome-based compliance, and the architect has a significant role to play in ensuring that the energy performance of the project is integrated into the overall building design.
The variance of actual building performance from energy model predictions can be attributed to many causes: occupant behavior and plug loads are normalized in the simulation process and may vary significantly from expectations, energy modeling may oversimplify system interdependence, construction quality may vary from plan, or operations may vary from anticipated use. But as our profession looks to a carbon-neutral future, it becomes increasingly important to connect building design expectations with actual performance.
Voluntary rating systems like the Living Building Challenge require actual performance data for certification, and LEEDv4 is moving away from projected energy use and toward actual performance by requiring energy metering as an EA prerequisite. Many jurisdictions are moving in the direction of true performance-based energy code compliance—meaning 12 consecutive months of code-compliant performance within three years of delivery—opting for voluntary compliance in the immediate future and mandatory compliance over the next few years. The growth of the whole-building commissioning process is evidence of a market where building owners are interested in making certain that buildings are constructed and performing as designed. The industry is headed for more performance-based and outcome-based compliance, and the architect has a significant role to play in ensuring that the energy performance of the project is integrated into the overall building design.
Use scientific method instead of assumptions
Architects should not rely on their intuition to make decisions that may not be correct or do not have the ability to be measured. A lack of real data makes their choices vulnerable later in the design process, which can be costly and inefficient because so much design development has already been predicated on these early directions.
For example, while architects assume orientation has a significant impact on building loads, it may have less impact than anticipated in a location with pervasive cloud cover. It is better to use the scientific method: use systematic observation to craft a hypothesis, make predictions based on that hypothesis, test the predictions, modify the hypothesis based on the results, and test again until the discrepancies between theory and results are resolved.
How often have we assumed that a certain orientation will be better but lacked the ability to quantify how much better? If, instead of using assumptions, we are able to calculate the improved performance of a particular orientation or window-to-wall ratio on a specific site, we would understand the benefits of these decisions and be able to communicate them effectively to our clients and the rest of the design team.
Design teams that model their projects during concept or schematic design—using software such as Sefaira, Revit 360, or AECOsim to test different options for the orientation, massing, and ratio of glazing to solid wall—know which of the options will perform better and by how much. These teams say that they are surprised by how often their initial assumptions are incorrect.
The location of a project is not just the physical characteristics of the site but its climatic conditions as well. Software such as UCLA's Climate Consultant allows us to analyze different aspects of particular climates through data, which is represented in easy-to-understand graphic charts. The program allows us to compare climates we know to the climate of the project's location, and to dig deep into specific aspects of a region. The sun shading chart allows us to see which shading solutions will be most effective for a building with a particular orientation in a specific climate, and the psychometric chart shows how human thermal comfort will be affected by that climate's temperature and humidity. It even quantifies 16 specific strategies to see which will be most effective.
On most building types, the building enclosure decisions have the greatest impact on the completed project's energy efficiency. Design elements like the orientation, massing, and ratio of glazing to solid wall are impactful and require significant effort to change as the design progresses further. The amount of insulation, glazing selection, exterior shading, and exterior cladding systems—even if very detailed and specified—will not make up for opportunities lost if these first steps are not taken with consideration and intent.
Will clients want to pay?
Early building simulation can lead to savings in the estimated construction costs by identifying efficiencies early in design. It can also save design time, as more cost-impactful decisions are made earlier and modeling will allow the design team to make decisions based on performance impact.
Amir Roth, the building energy modeling technical manager for the Department of Energy, cites a study by HOK that reported of energy modeling costs as being paid back within the first several months of building operations. According to Anica Landreneau, the director of sustainable design and consulting at HOK, large building modeling costs run from $20,000 to $200,000 for their projects, depending on the number of iterations of modeling, and the complexity and size of the project.
Architects should not rely on their intuition to make decisions that may not be correct or do not have the ability to be measured.
Often, modeling identifies first-cost savings by initiating early cross-disciplinary conversations regarding energy conservation measures (ECMs) such as using a radiant HVAC system instead of an air distributed system or the application of improved glazing to obviate perimeter heat or reduce the number of chillers required. This approach has to be coordinated with the thermal performance of the enclosure, but if its benefits are identified as a project goal then the team can move forward with an understanding of the importance of incorporating this particular aspect.
First-cost savings commonly attributed to energy modeling is the identification of unnecessary costs in the form of oversized and expensive HVAC systems. The identification of first-cost savings relies on an integrative design process wherein the architect reduces building loads and the mechanical engineer responds to the load reduction in the HVAC design.
How do we model properly?
After climate and site analysis, the team should set a benchmark or baseline model to understand the performance of a similar building. In predesign, the team can establish a baseline model using one of several methods against which the developing design can be compared. In a Building Green webcast on energy modeling for early design decisions, Prasad Vaidya explains that one can create a benchmark or baseline model for model calibration based on existing energy use, such as on a campus where this information is available.
A benchmark can also be created using one of four available databases: EPA's Target Finder, EIA's Commercial Buildings Energy Consumption Survey, the Labs21 Benchmarking Tool, the California Commercial End-Use Survey, and DOE's Building Performance Database. Vaidya also explains that, in his experience, early models are 70 percent accurate in predicting the actual performance of a building. The model’s accuracy grows to 90 percent at the end of construction documents, and 95 percent during operation.
Benchmarking and climate studies will help the team to identify the ECMs and set target goals. This effort can be initiated during an eco-charrette, where the design team—including the energy modeler, client, and users—are represented. ASHRAE 90.1 is a good starting place for getting an understanding of baseline goals, as it is the standard for all LEED projects.
According to Amarpreet Sethi, an energy modeler for DLR Group who presented in the Building Green webcast on full-blown energy modeling, the next step is to find strategies for reducing peak loads by using parametric modeling to study massing, orientation, window-to-wall ratio, shading, daylighting ,and the effect of these elements on each other. This process should result in a few options that can then be evaluated for comparative effectiveness. During the schematic design process, the team will want to do a preliminary life-cycle cost analysis to understand the financial impact of each option. By the time of design development, the systems have been selected and are being developed. The majority of the big architectural decisions will have been made, and the balance of the modeling during the design process will finalize the ECMs and evaluate the updated designs to ensure meeting energy use goals.
As our profession looks to a carbon-neutral future, it becomes increasingly important to connect building design expectations with actual performance.
Jillian Burgess, a building enclosure consultant with the Façade Group in Philadelphia, illustrates the distinction between comparative modeling and predictive modeling: Comparative modeling looks at the relative impact of various design components, while predictive modeling is a very specific modeling technique that seeks to identify the exact energy use of a building by identifying the exact internal loads and external loads at a highly detailed level. It is a very in-depth process that is rarely necessary. Most projects employ a type of comparative modeling that does not target the exact energy use but rather quantifies how much better one design option is over the other or how the design is an improvement over minimum code.
Parametric modeling, on the other hand, uses a range of outcomes on a set of outputs. The input geometry or attribute ranges, and the output goal metric ranges are usually scripted in coding software to harness computing power and run several models based on a genetic algorithm. It is not related to level of design and can be done on simple massing models or highly detailed models. It is a powerful tool that can help teams set energy goals and understand which variables have a bigger or smaller impact on the project’s goal metric.
"Many architects don’t get specific enough when asking questions in early schematics," Burgess cautions. "I think the variables are clear—sunshades, orientation, glazing—but the goal metrics are often not clearly stated. Architects often ask for energy use in early design, but often what they mean is loads, and probably more specifically external loads because that’s the only thing they are impacting with the design of the envelope. So, then, the first question should be what is the dominant external load, and how can I affect it or how does this design affect it? And in parallel, how do external loads stack up to internal loads and how do these relate to the building energy use? These questions will help you target design strategies with very little modeling effort."
Extending value of the architectural service
We wouldn’t buy a car if we didn’t believe that the miles-per-gallon performance data on the window sticker was aligned with the actual performance of the vehicle. It is a simple concept: demystify the process, set measurable goals, and work to achieve these goals. The tools for this are widely available and becoming easier to use, thanks to graphic interfaces and elimination of the need to switch between platforms to capture all the information being assessed. The industry is headed in a positive direction, and there is a real opportunity for architects to deepen their expertise, increase the value of our services, and become leaders in this endeavor.
Pamela Sams, AIA, is a design realization leader for Gensler’s Southeast Region and the 2017 chair of AIA's Building Performance Knowledge Community.