Performance analysis is new to most architects. But fortunately they have a good precedent: in many ways, analysis is a lot like architectural drawings, which move from “fat pen” diagrams to hardline detail. Here are four similarities that can help you get better feedback in less time.


1. A good drawing is only as detailed as necessary.

How much detail is needed for an accurate energy model? How much time needs to be invested in inputs? The answer is similar to a good drawing: the right level of detail depends on the question you’re trying to answer.

As architects, we often approach the early stages of a project with “fat pen” diagrams. When trying to understand the basic organization, concept, and form of a new building, these broad-brush sketches are critical tools for clarifying a design and simplifying the sometimes daunting complexity of a building.

As the design progresses, we naturally move to finer, more detailed types of drawings — tools appropriate for understanding constructability, water management, thermal bridges, etc. Sometimes we “scale jump,” moving back and forth between detail and concept, to understand how one scale informs the other.

The same processes work well for energy and daylight analysis — and for similar reasons. If you’re trying to decide whether Form A or Form B is better, broad brush analysis using standard defaults is sufficient to provide good comparative data. If you’re trying to determine the largest energy loads for a building, you’ll want to make sure you’ve roughed in the occupancy and internal loads. If you’re trying to get to Net Zero Energy and every Watt matters, you will eventually need to dive deeply into each element of the design. The principle is to match the analysis with the scale of the design decision being made.

2. Drawings are for discovery, not just documentation.

Architects use drawings to work through problems — to do the actual work of designing. Drawing is a way to understand whether a design works, to investigate problems, to test solutions.

Similarly, performance analysis provides the most value when it informs a design decision. Analysis should not simply be used to create a snapshot of energy or daylight performance — that’s just documentation. Its real power is discovery: using the model to find the biggest opportunities — to create a strategy, formulate an approach, take a defensible position. What aspects of the design are most important? Where are the biggest opportunities? Which moves provide multiple benefits? These questions help you uncover possibilities.

3. Added detail provides diminishing returns.

The first elements we add to a drawing are usually the most important. As we move from a “design drawing” to a “documentation drawing” we often add a lot of detail — gaskets, connections, notes, hatches. These add context and clarity, but each addition arguably adds less value than the previous one. The early decisions are what define the design.

Similarly, adding detail to an energy model generally provides diminishing returns in terms of accuracy. Early design decisions often have the biggest impact on energy and daylight performance — and these are decisions that can typically be evaluated in comparative terms, using standard defaults.

“Adding detail” to an energy model generally means specifying more precise inputs in three areas: (1) space occupancy & operation, (2) envelope properties, and (3) mechanical systems. Tools like Sefaira start users off with a large number of defaults based upon a small number of inputs. When a user inputs an Office type, Sefaira automatically fills in occupancy rates, lighting loads, plug loads, HVAC setpoints, and operating schedules to be representative of a typical office. Similar defaults are available for envelope properties and system efficiencies. As designers invest additional time to gather data and change these defaults, the results get more accurate but within a narrower band. Default assumptions get the designer roughly 80% of the way there in terms of accuracy, and each additional percentage point requires increasing time & effort. This added time is only “worth it” if the added accuracy is necessary for the scale of the inquiry.

4. Drawings are just representations.

Like drawings, energy and daylight models are representations. They are useful as tools to make and communicate decisions, but even the best rendering will not be identical to the final built project.

Energy models are full of assumptions about weather patterns, construction, and how a building will be operated — assumptions that are bound to differ from reality. Architects must set expectations accordingly with their clients, and are well-advised to present analysis results in terms of percentage improvement, or as a range, rather than an absolute value.

There are good resources out there for improving accuracy — for example, see our post on Predictive Modeling at the Design Stage. But the point is that not all models need to be predictive — on the contrary, like a compelling sketch, some of the most important insights can be captured with a simple model, fit to purpose.

Learning to sketch with analysis tools

For some additional resources on how to use analysis to explore ideas, check out the following resources: