Energy Modeling: Early and Often
For the biggest energy and cost savings, model early in design and model often, say the experts.
By Tristan Roberts
Thanks in part to green building programs and codes, whole-building energy simulation has become more common over the last two decades, but it’s not necessarily having a big impact on design decisions—let alone energy savings. “Validating the final design is really important to validate LEED points and validate code compliance,” notes Maurya McClintock, Assoc. AIA, of McClintock Façades Consulting. As McClintock and others note, though, when modeling is only used late in design—after the massing, orientation, envelope and glazing design, and mechanical systems in a building are already specified, and hundreds of hours of work have already been put into those designs—the modeling might have little value beyond keeping score.
“It blows me away that that’s where we are,” says Marcus Sheffer, an energy consultant with 7group. As critics have pointed out, a “green” building modeled to save a certain amount of energy doesn’t necessarily end up doing that. Given accurate inputs, models are accurate at forecasting energy use, says Sheffer, but “models can’t accurately predict the future”: actual operating conditions will always differ from modeled conditions. This typically happens because equipment and controls are installed differently than modeled, or because weather patterns or occupancy are different than expected. The real value of modeling is not predicting energy use but making relative comparisons among design options, says Sheffer.
Beating the 40% Savings Barrier
It’s safe to say that most whole-building projects don’t use an energy model. According to the most recent data published by the American Institute of Architects (AIA) from firms participating in its 2030 Commitment (see “,” EBN July 2012), 57% of projects used whole-building energy modeling in 2011. That figure is drawn from the 112 participating firms who reported data to AIA—a small fraction of AIA membership.
Rand Ekman, AIA, director of sustainability at Cannon Design, told EBN that his firm modeled 54% of its projects in 2011. He points out, though, “We have committed, de facto, to energy modeling 100% of our projects” due to the firm’s 2030 Commitment as well as trends in building codes. Noting that the current target for the 2030 Commitment is 60% energy savings above average performance (ramping up to 70% in 2015), Ekman explains, “Energy codes and standards are pushing commercial projects beyond what they can do without modeling.” He notes that some commercial building codes have adopted the 2012 International Energy Conservation Code (IECC), which brings with it ASHRAE 90.1-2010—estimated to be 40% more stringent than current practice. Ekman notes that California’s Title 24 energy code is also about 40% better than practice. “Right now, the best you can do just using code-equivalent design practices is 40%,” Ekman says. In other words, he argues, “You have to have the actual information you could get from doing an energy model” to know whether you’re doing any better than 40%. That’s why, in Ekman’s view, any firm that signed to the 2030 Commitment—or that wants to earn a high number of LEED points—has to be modeling of all its projects during design. And to achieve those energy savings cost-effectively, that modeling has to start early in the process.
Greg Mella, FAIA, of SmithGroupJJR agrees, noting that his firm is modeling 77% of its projects, although he guesses that only 50% of projects are being modeled early in design. “We are really actively working to get to 100%. I can’t think of any reason why we wouldn’t do it,” he says.
Stricter energy codes are making early design modeling more important than ever, says Christopher Green, FAIA, of AGO Studios. Newer codes leave less room for sloppy energy designs, and “you don’t want to find out 80% into construction documents that you’re not in compliance,” he notes. Early modeling “informs a lot of early design decisions, and especially when you’ve identified conservation goals, you want to know early whether the concepts you’ve developed are going to meet the goals or if you have to modify those before you commit to a direction.”
Green and others mention another reason to model: prescriptive paths can offer energy savings beyond code, but they limit design options. “If the owner has a specific site and is trying to take advantage of certain opportunities,” such as views from a certain orientation, “the prescriptive path can be fairly restrictive,” Green says. Architects will frequently want to increase the glazing on a given orientation, which they can do only through a performance path because it’s based on an energy budget instead of by-the-book compliance with things like whole-wall R-values and window-to-wall ratios. By taking the performance path and modeling energy use, designers can compensate for more glass with higher-performance glazing, increased insulation, or other measures.
Kirk Teske, AIA, chief sustainability officer at HKS, says that his firm models about 60% of its projects, although he adds that the number includes numerous interior fit-outs and that the whole-building rate is higher. Teske notes that modeling can be a harder sell on some of the larger, more complex healthcare projects that HKS designs. “The greater the internal loading of the project, the less impact the building-envelope component of the architecture has, so there is a stronger focus on meeting prescriptive requirements,” Teske says.
Starting with a Target
Teske and other designers agree that orienting the entire team toward an energy target is key to integrating energy modeling. At the start of a project, “we present benchmarking to them,” Teske says. “We’ll communicate to them that there is not one big silver-bullet strategy that is going to get them to their energy reduction target. It’s going to require a percent here and a percent there all along the way.” Modeling all along is the best way to provide the data that will identify those percentage gains.
“Pick a goal that has meaning for you,” says Sheffer. Typically, he says, a goal will be expressed in energy use intensity, or EUI—energy consumption per square foot—but it could be carbon-related or even cost-related, he says.
A project in Grand Rapids, Michigan, demonstrated for Amanda Bogner, P.E., of The Energy Studio how energy reduction goals can provide a unifying focus for the design team—one that pushes a team to a smarter design. On the project, a corporate headquarters visible from downtown, the team set a target of 40%–50% cost savings before renewables. At the same time, a driving force of the design was to incorporate views of downtown Grand Rapids and in turn to offer an attractive façade to the downtown. “The architects had gone through design and come up with a 70% glass south façade,” says Bogner, but “it was consuming a lot of energy,” according to an early model. Working with a daylight modeler and an energy modeler, the design team came up with several alternate façades and settled on one that was still going to make a statement but also made the energy performance a lot better. “We encouraged them to pick a couple views and make sure they were using glass wisely,” Bogner says. “Luckily, the building owner was really receptive to that; the [energy] target helped frame the context of the conversation.”
Comfort matters too
There are other kinds of targets that also matter, says Teske. “A patient room may not be that big a deal to the overall energy consumption, but the comfort of patients can be a greater driver in modeling the building and understanding the perimeter loads.” He adds, “Designing for comfort alone is a common value.” If that helps get the team on board with doing an energy model and designing smarter, more efficient systems, so much the better.
Brad Schaap, P.E., of Leo A Daly found a similar outcome on a healthcare project: his team convinced the owner to do early modeling on different building-envelope options to investigate impacts of solar orientation and shading. Because of internal loads, including ventilation requirements of 10 air changes per hour in patient rooms, the model showed minimal energy savings from a design option that was intended to reduce solar heat gain, but the modeling information helped Schaap convince the client that the design option could pay off in terms of thermal comfort: “When the glass is shaded, you’re going to feel more comfortable in that space,” he says.
Modeling and Integrative Process
The kind of buy-in that Bogner encountered in Grand Rapids is a key ingredient for successfully integrating the information coming out of an energy model, according to experts EBN spoke with. “For it to be truly effective throughout the process, there has to be some champion at a higher level that understands the value of it and really pushes for its incorporation into the decision-making process,” says Sheffer. “Absent that, quite often the modeler can say, Here’s what I think we ought to do, and if there is no champion, then you get the minimum.” Sheffer notes that on more than one occasion he, as the modeler on a project, has had to try to explain the value of modeling to a skeptical team.
Experts EBN spoke with also agree that early design modeling goes hand-in-hand with an integrative (or “integrated”) design process. “When you’re talking about energy, you’re really talking about integration of all the building systems,” says Sheffer, an advocate for integrative use of energy modeling and a co-author of The Integrative Design Guide to Green Building: Redefining the Practice of Sustainability.
Sheffer likes to see what he calls a comprehensive energy modeling scope, potentially starting in the programming phase, before design has begun. “In about half an hour, you could do a simple box model to get an idea of the distribution of energy loads in the building,” he says, which starts to inform early design concepts. Once design has begun, fairly simple models can be run to evaluate the relative performance of different conceptual designs, with a focus on massing, orientation, percentage of glazing, and other fundamental architectural design choices.
When the project has homed in on a schematic design, modeling work can concentrate on reducing major building loads, such as heating and cooling, and then move on to more refined envelope issues, glazing characteristics, internal gains, lighting loads, and plug loads. “A series of iterative models at that point can develop the design further to get the loads as low as they can be,” says Sheffer.
Up until this point, he argues, the mechanical system should be kept neutral. Only when loads are reduced does he recommend using the model to help analyze mechanical-system choices. That’s not to say that the mechanical system shouldn’t be discussed before then—and it may even make sense to include a likely system in the model—because mechanical system choices might impact how far some efficiency measures are pushed. But doing things in that order has an impact on cost, Sheffer says. After the project’s loads are reduced, the model can be used to evaluate different HVAC systems. The modeled energy use for each system is then used in a life-cycle costing analysis to determine the most cost-effective system. Once a system has been selected and the project moves through design development and into construction documents, modeling can help evaluate various mechanical-system components—energy recovery, economizers, equipment efficiencies, and more—and eventually to evaluate whether the project is meeting its goals. For a final model to validate code compliance and document LEED credits, Sheffer waits until construction is mostly done since the final model must be based on the as-built project.
Given all the earlier opportunities to affect design with modeling, it’s clear that doing only a model for the as-built structure misses many opportunities. Kim Shinn, P.E., senior sustainability consultant with TLC Engineering, emphatically agrees, telling EBN that he would ideally like to work with architects “before designers are making sketches on the backs of napkins.” He says, “Before I talk about ground-source heat pumps or variable refrigerant flow, or any of the whiz-bang stuff we do on the mechanical side, let’s look at what we can do with the design to get the energy performance.” He acknowledges that “sometimes that is not well received”—both from the architectural side and from the engineering side. Architects may not think they need that kind of early feedback or might be worried about increasing design fees, while engineers might feel more comfortable doing more conventional modeling for load calculations and equipment when design is mostly established. Shinn says, however, “I’m gradually winning over both our architecture clients and my own colleagues that this is really the most economical way.” Late-stage decisions or design changes cost more, he says, and working early to reduce loads means smaller mechanical systems as well as reduced costs both before and after occupancy.
Who does it?
Some, like Norman Strong, FAIA, of the Miller Hull Partnership, believe that integration of energy modeling will inevitably lead to more everyday use within architectural firms. “When I graduated from college, we had to hire ADA [Americans with Disabilities Act] consultants because no one could figure out how to put a ramp into a building—and now we just do projects that way.” Another common reference point is the introduction of CAD (computer-aided design) tools and the practice, common at the time, to charge an extra fee for CAD work. Now, that’s just how it’s done. Some firms are currently charging extra fees for models, particularly when hiring those services out, but some are absorbing the costs into their regular design fees.
“Energy modeling will help the practice to really understand what their design decisions are doing,” says Strong, although he notes, “We don’t want to replace the highly detailed engineer-led energy modeling.”
Mella, of SmithGroupJJR, says that there is reasonably usable free software, and “it’s the kind of thing you can learn to use in an hour.” He says that there really are no obstacles, other than habit, to architects using modeling in early design decisions: “It’s a matter of changing our design culture to routinely take on these kinds of explorations during design.”
A common refrain was voiced by Maurya McClintock: “There aren’t a lot of tools that target the early design phase,” she says, explaining, “Most of the tools for energy modeling at the moment have been developed for assessing the energy performance of an entire building design, and that includes all of the engineered systems as well as the architectural systems.” But “in early design, those systems haven’t been thought about yet—and that makes using the current tools cumbersome.” Software that has the capability to more quickly assess multiple alternatives is on McClintock’s wish list. Green echoes that opinion, noting that it’s especially true for smaller projects. See the table on page 11 for a full rundown of commonly available tools.
Trusting the output of energy modeling tools is another question, one that has been on the lips of more people as stories circulate of new buildings not living up to modeled energy performance. As Mella says, echoing Sheffer’s concerns quoted at the start of this article, “You should never be using a conceptual energy model to get a predictive result.” In addition to the common issue that buildings are not always operated as designed, another problem is simply that these early-stage modeling tools have to make assumptions or use algorithms that don’t line up with the real world. A recent exercise at SmithGroupJJR illustrated how common modeling tools don’t even line up with each other. At the start of work on a building for Western Michigan University, SmithGroupJJR ran initial models in IES, OpenStudio, Revit’s conceptual modeling tool, and TRACE. With the same design and the same zip code, “we found results really do vary a lot between the different engines,” says Mella. The exercise found that predicted energy use intensity (EUI) varied from 66 to 144 kBtu/ft2/year in the different software programs. Mella notes, however—supporting the use of these tools for making comparative design choices—that the relative savings within each tool appear more consistent with each other. Adding sun shading, for example, typically results in a fairly consistent percentage savings across the various tools.
Big Savings That Cost Less
Delivering a better building is great, but delivering a better building with reduced first and operating costs is even better. Can early energy modeling help make that happen?
In the Integrative Design Guide to Green Building, Sheffer and his coauthors discuss the case of the Neptune Midtown Community Elementary School in New Jersey as a classic instance of what Amory Lovins of the Rocky Mountain Institute has called “tunneling through the cost barrier” and that Sheffer more succinctly sums up this way: “BIG savings cost less than small savings.”
Making these savings possible comes from looking at “integrative cost bundles,” says Sheffer, in contrast with simple payback analysis on a line-item basis. If looking at a project on a line-item basis, says Sheffer, “All the stuff you’re incorporating to reduce the loads is going to show up as being more expensive,” while mechanical systems will likely be estimated as a conventional per-square-foot number and not as a lower-cost system that might have paid for those higher-cost items. “You’ve captured all the additions and none of the deductions, and gee, what’s going to happen when you value-engineer?” asks Sheffer, rhetorically. Relying on energy modeling data to show how different system bundles would pan out in terms of costs and energy, the team on the Neptune project was able to justify an additional $125,000 in energy-efficiency measures such as added insulation and triple glazing, which allowed for a 40% load and system capacity reduction and led to $400,000 in savings for the mechanical system (ground-source heat pumps).
Shinn agrees that this use of energy modeling not only saves construction and operational costs but also costs less during design because “the earlier you can make a decision, the less it will cost to implement that decision.” Becoming overly committed to a design before realizing its energy implications can mean either building a lesser building than might have been possible or eating up design fees by retracing steps. While she acknowledges that this doesn’t happen very often, Bogner told EBN, “We really need to be doing the model before the architect shows the owner a rendering of the project.” The owner gets an image of the project, and it becomes their passion, but what if it’s not a good choice in terms of energy?
Although Sheffer doesn’t like the idea of payback periods, preferring to look for those big energy savings that reduce costs, Teske has found that more efficient systems might end up costing more, “so you have to get into operational savings and payback periods.” When that comes up, HKS uses the modeling data to show how the elements being modeled reduce the tonnage of the air-conditioning system, for example. “When we get deeper into the energy modeling, we are looking hard at costs and trying to find ways of revealing what the costs are,” Teske says.
Energy modelers and designers using modeling speak of a feedback process that occurs: results of modeling become feedback not only for the current project but also for future projects and design approaches. Here are a few lessons that designers have picked up along the way—some obvious, some less so.
Massing and orientation
Massing and orientation are often not in the control of the design team due to site or program constraints in infill or campus settings—but when it’s possible to influence, early modeling can provide a lot of help in choosing promising directions with relatively few inputs. A bar-shaped building oriented east-west will typically perform better than the same building oriented north-south, for example—but Teske says not to take anything for granted.
“What’s interesting is that you can go about design in radically different ways and achieve the same target,” he says. “You can put a lot of glass on the outside of a building in the right climate zone if you do the daylight harvesting and lighting power density reductions. You can also achieve the same goal by doing a big, deep box that minimizes its perimeter load, but that is not nearly as attractive.”
In other words, modeling won’t tell you how to design a building—unless you ask it a specific question. That’s what Mella did on an energy model of a new Chesapeake Bay Foundation environmental center targeting net-zero energy and Living Building Challenge certification. Modeling quickly showed that a south-facing bar produced the best energy performance relative to other options, but Mella pushed it further by adding east-facing curvature to the bar to allow early-morning sun to heat office spaces before the staff arrives. He performed iterations modeling the curvature and its precise orientation to get the optimum design—a 1.5-degree curve with slightly reduced bay size so that the overall footprint didn’t grow.
Envelope and glazing
Energy experts agree: energy savings in most buildings should start with the envelope and work inward, considering mechanical systems later.
“Glass placement is the single thing that you as an architect can make sure you do appropriately and that will by far have the biggest impact,” says Bogner. “Once that is set, maybe you’re looking at different solar heat gain coefficients” for glazing, using lower-SHGC glazing to reduce unwanted heat gain, but by then you’re missing the point.”
Strong notes that it’s worth looking at details like the floor-to-floor height. At the Bullitt Center in Seattle, modeling showed that raising the height 14 inches per floor gave all spaces within the floorplate access to daylight—offering a net energy gain.
“It doesn’t take doing more than three or four energy models before you realize that lighting is a huge deal,” says Teske. As a result, “we try to get our designers to not just plop down a 2x4 scheme in the same way you’ve always done it, but really look at your lighting power density and set a target, and get your electrical engineer and your lighting designer to help track that along the way” through design.
Bogner agrees but also notes that energy codes have reduced lighting power densities industry-wide, so finding savings here can take some real work.
Fun with Conceptual Energy Modeling
Getting designers to integrate quick-and-dirty energy modeling runs into their conceptual design decisions is a paradigm shift, argue both Teske of HKS and Shinn of TLC Engineering, and what better way to gently introduce the idea than through a game, combining elements of fun, surprise, and competition?
At Greenbuild 2010, Teske and Shinn unveiled what they call the Wizard Game to do just that. The magical name comes from the Schematic Design Wizard component of eQuest software, a free, quick, and easy tool for comparing design scenarios. Since first sharing the game, Teske, Shinn, and others have brought it to nearly every climate zone in the U.S., by popular demand.
The game starts with the presenters showing an energy model of a hypothetical 50,000 ft2 building that complies with the prescriptive requirements of ASHRAE 90.1-2007. Before being divided into two teams, the audience gets a basic understanding of the base building design, the climate zone, and energy loads. A four-quarter “game” then begins.
Each team selects from alternative massing and orientation choices. The presenters show the modeled energy impacts of each team’s decisions.
Each team has the opportunity to select two of five building-envelope modifications, including changes to roof insulation, wall insulation, window-to-wall ratio, glazing specifications, and exterior sunshade devices. Again the measures are modeled, and the results of both teams are shared and evaluated.
Each team can choose to add daylight harvesting, a reduction in the lighting power density, or both. They can also add a third building-envelope measure from the first quarter instead of one of the two lighting options.
After observing the results of both their own and the other team’s decisions through three quarters, each team has the chance in the fourth quarter to revisit all of their previous decisions.
Winning teams typically achieve 20%–25% reduction in energy compared to the baseline, making very rough design changes and without touching the mechanical systems (except for reductions in loads). Participants usually find they have a lot to discuss; according to Teske and Shinn, here are some common lessons learned:
The energy code has already optimized building insulation, and adding additional insulation to a building dominated by internal loads, particularly in a cooling climate, may be counterproductive.
One of the glazing options is described as “Low E,” and teams often choose it even though its solar heat-gain coefficient is much higher than the base building’s glazing, making it a poor choice in cooling climates.
Exterior sun-shading devices are effective in some climates and envelope designs, but just as often they may be counterproductive on an energy basis, especially in colder climates.
According to Shinn, there is a general tendency during the game, even among commercial architects, to make decisions for the modeled building as if it were a house. But in energy terms, he says, “Houses are from Venus and commercial buildings are from Mars.”
In a statement that could probably have been made by any of the experts EBN spoke with for this article, Shinn sums up many of the lessons learned this way: “If you ignore or are not conscious of the effects of climate, internal loads, and occupancy and operation schedules on a building, you can do some really dumb stuff.”
For more information:
An Architect’s Guide to Integrating Energy Modeling in the Design Process
The American Institute of Architects