Data-driven design

Data-driven design uses analysis to shape the design response. When it comes to urban design, analysis is a powerful tool for understanding the effect of context on the site. Advocated by architecture theorists like Kenneth Frampton, contextualism seeks to derive design responses from the urban fabric surrounding the site. Relationships between the climate, people, business, and surrounding structures powerfully shape the design response. While other ideas can drive the design concept, all buildings are affected by their context. 

Good intentions in architecture often lead to regrettable results when designs lack understanding of the effects of the environment such as sun, wind, and light. From selecting the shading on the southern facade to the orienting of a green wall on the southwest entrance, analysis is the key to success. Before widely available computer simulation, it was necessary to approximate on these kinds of decisions by using a sun path diagram. Now we can precisely measure and validate our design moves on the urban and architectural scale using data-driven design. Here, we will walk through the design analysis of a recent project for Emory University’s Rollin’s School of Public Health III, to illustrate. Context was critical to the selection of design options and project budget which in turn shaped the location and sizes of gardens, programs, and shading strategies.

01 Site & Climate

site and climate context

The School of Public Health at Emory extended its campus footprint with the addition of a new state-of-the-art conferencing and distance learning center. The new 140,000 SF building is placed at the center of the densely packed campus. The design responds to many other ongoing campus projects, while still meeting its own program demands. The design team sought to find the most ideal configuration that satisfied a long list of requirements. Identifying massing & orientation, glazing types, shading strategies, and placement of outdoor terraces and gardens helped meet Emory’s student wellness goals. Critically the project site situated the new tower on a small plot of land, previously a modest study lawn, in the immediate proximity of existing buildings of three to seven stories. This context made a critical impact on daylighting and campus experience of the new building showcasing the need for accurate analysis.

climate analysis

02 Maintaining Accuracy

Accurate setup for any simulation is essential to maintain confidence in one's results and data-driven decisions. A project can be analyzed in isolation by itself and produce misleading results. There are usually context buildings around that cast shadows, reflections, and block views which must be considered. Below is an example analysis of the Rollin’s School demonstrating the drastic difference of analysis results when the context is not included (bad practice) versus when it is included (best practice). At its core, analysis is just a decision-making tool. If we did not include context we would make a poor decision. This is particularly noticeable regarding the glare (ASE) and daylight (sDA) analysis showing a massive difference between the two.

daylight and glare analysis

03 Massing & Shadow Studies

During the early stages of design, the design team (SLAM Collaborative), explored a range of massing schemes that would fit on the site and work well in terms of scale and master plan of the campus. The team determined the following 4 massing options were best suited in terms of design and aesthetic to explore further. Modeling to get a quick snapshot of the impact of the massing on energy use, daylight was critical is determining which options should move forward. The comparative study shown below took the team less than  2 hours to create.

massing comparison

With the massing comparison, the first two schemes were removed due to their high energy use and aesthetic. The remaining two massing schemes were further explored, looking at the orientation of the building overall for North/South and East/West. The shadow studies below were completed to provide insight not only into how the new building will connect to the existing ones, but also how the context would obstruct a significant amount of daylight throughout the year. Additionally, not only did this help determine the impact of context on the new building, it helped analyze the impact of the new structure on the existing. The design team chose option two orienting the main footprint North/South to provide shadow on the public terrace year-round with a short amount of direct sunlight in the mornings and during winter.

massing and orientation

04 Impact on Lower Levels

Daylight analysis for the lower level was also intrinsic to the success of the overall building design. The main programs for the lower levels are large lecture areas and sizable conference space at ground and plaza levels. These primary spaces need immediate street access for best serving future events. However, having a large floor plate on the ground level with surrounding context buildings also generated challenges for getting daylight into the space. For this project, the early daylight analysis provided data-driven metrics to guide layouts and orientation.  The design team explored a wide range of facade strategies, glazing percentages, window aspect ratios for daylight, glare and energy to drive iterative decisions . Eventually, the design shifted to allow more glazing on the lower levels knowing these windows were protected by the surrounding buildings. These tests also inspired a stepped massing with ideal locations for outdoor lounging and study areas which was necessary given that the building was displacing the original Rollins Public School of Health lawn. There was also no need to add shading devices (due to shading from context) for these lower levels saving significant project money. Had the same analysis been conducted in the absence of context, the design team would have not maximized the glazing on lower levels and would have struggled to identify a naturally shaded location for an outdoor patio.

daylight and glare analysis

05  Impact on Glazing Options

Façade designs are always a result of finding the balance between design intent, occupant experience, environmental conditions, and budget. Knowing the impact of immediate context buildings helps narrow the scope for window options when one can predict the accurate result of daylight and glare. Running a simulation without context, one may find the same design needs to have less or smaller windows, which obstructs the view out and often disrupts the desired aesthetic. For the Rollin’s School, the attention to more costly glazing options was able to shift away from the ground floors and apply them only to the tower portion. Opening up the windows at the ground provided a large welcoming space and brought in 25% more natural light. The facade design also reflected the impact by stepping out to self-shade along the west and east facades and only requiring shading devices toward the top where the context was not helping. The design team was able to focus on a strategic few locations on the tower to study additional window strategies. The team utilized rapid facade prototyping to test various glazing strategies with real-time analysis with a dynamic testing tool cutting down on analysis time and turnaround. The analysis below took the team 20 minutes.

facade optimization

06 Impact on Energy

Emory’s sustainability goals targeted a minimum LEED Silver while pushing to achieve carbon neutrality. To that end, the team sought to reduce energy and water use while self-generating 10% of its own energy on-site. Using the unique cove.tool optimization feature, the team analyzed thousands of options to identify the best solutions for both energy and cost. The team found the most optimal combination of building options to reduce cost and increase performance. By knowing the key impacts of the site context, the proper analysis resulted in outperforming the national average EUI (Energy Use Intensity in kBTU/sf/year) for education buildings by almost half from a design basis of 93 EUI to 56 EUI. The optimization showed that reducing the EUI to target levels between 50-60 EUI would cost between $20k-$85k per EUI reduction, helping to determine the lowest cost way to meet the targets.

optimization

optimization bundle

Resources

cove.tool Insights