Overview
TRACE (Tool for Rapid Assessment of Carbon Emissions) is a structural engineering design tool intended to help consulting design engineers estimate total embodied carbon in the gravity system of their building and compare it to project benchmarks as well as industry standards.
Often times, early in the lifecycle of a new development project, structural engineers must make a decision about what gravity and lateral system they want to use. At the same time, an owner/develop may have specified carbon goals they want to hit for LEED, SE2050, or related programs. This tool is intended to guide these designers through the decision-making process and lead them to making a data-driven decision about the smartest choice for their building.
How it works
The tool operates by taking a specified set of user inputs, described below, and conducts preliminary beam sizing based on industry-standard logic and rules-of-thumb, and then calculates embodied carbon based on standard metrics. It then compares the totals to benchmarks. The tool also constructs a visual representation of the building to aid in schematic design.
In specific, the underlying logic behind the design is as follows:
- Estimate beam/column sizes from span rules of thumb (e.g. span/20 for beam depth, 16” square for concrete columns)
- Calculate member volumes from geometry
- Multiply by material density to get mass
- Multiply by published embodied carbon factors (kgCO₂e/kg), which come from standard databases and investigative papers
Primary Inputs (Design Variables)
The tool takes several inputs:
- Bay size in feet (X and Y span)
- Number of bays (X and Y)
- Number of floors
- Floor-to-floor height
- Mechanical, typical floor, and typical roof loads (PSF)
- Structural system choice (from structural steel, reinforced concrete, and mass timber/glulam
Primary Outputs
- 3D representation of structure designed
- Total embodied carbon in kgCO₂e
- Embodied carbon on a per-square-foot basis (kgCO₂e/sqft)
- SE2050 performance level
- Material performance comparison
System Requirements
- Mac/Windows machine running Rhino 8
- Local Python 3 environment
- No additional requirements!
How to Use
- Open Rhino, and run the Grasshopper command to open a separate window. It is highly recommended to split screen both programs so that the user can see the visual feedback of their parameter adjustment in real time.
- Open the TRACE file. The sample structure will automatically generate in the Rhino window.
- Provide input geometry based on preliminary design information available to you as the designer. As you update the geometry inputs, the structure preview will automatically update in the Rhino window.
- Update loading values as needed. The program assumes a mechanical loading on the first floor, roof load on the top floor, and typical floor load on all others.
- Select your material of choice. The structure in the Rhino window will update to match your choice.
- Scroll to the output section of the tool to see estimates for the total kg CO2e of your building for all material options, along with the per-square-foot equivalent and the performance against SE2050 benchmarks.
A video tutorial can be seen below; this video is intended to help users familiarize themselves with the TRACE environment and the different inputs.
The full Grasshopper logic used can be seen below.
Some key assumptions were made in the development of the tool.
For the embodied carbon associated with structural steel, I opted to use the 2025 AISC (American Institute of Steel Construction) EPD (Environmental Product Declaration) for Hot-Rolled Structural Steel Sections, issued 10/14/25. This report cites a total global warming potential (GWP) of 0.898 kg CO2e per kg for the A1-A3 stages (for all intents and purposes, these are the stages associated with production).
For concrete, I chose to use the NRMCA LCA Report, published in July 2022. For stages A1-A3, for 4001-5000psi concrete (what would typically be used for a purpose like this), it reports minimum and maximum values of 288.10 and 468.89 kg CO2e per cubic meter of concrete. If we use the average of these two values at 378.495 kg CO2e/m^3, and assume an industry-standard concrete density of 2400 kg/m^3 (150 pcf), we obtain a value of 0.1577 kg CO2e/kg.
For mass timber option, I chose to use a study by the firm Arup on Mass timber embodied carbon factors, which presented average A1-A3 embodied carbon factors of 0.25 and 0.28 kg CO2e/kg for CLT and glulam respectively. I opted to use 0.28 for the purpose of this tool, as glulam is more commonly used in beams.
For the slab, I assumed a conventional 6” thick slab at all floors. 6” = 0.5 ft, and at 150 lb/ft^3, this means 75 lb per sqft. Assuming this means 34.02 kg/ft^2, using our 0.1577 kg CO2e / kg factor from earlier, we assume 5.365 kgCOe/sqft.
For the conversion between column volume and weight, I assumed standard density values of 150 pcf for concrete, 490 pcf for steel, and 35 pcf for mass timber/glulam.
Finally, to decide on benchmark embodied carbon values, I utilized values from the 2024 SE2050 industry report on embodied carbon. This report stated that a recommended maximum was 350 kgCO2e/m2, and the industry average was 270 kgCO2e/m2. As such, I included categories for within maximum, industry average, high performance (75% of average) and ultra high performance (50% of average). It is worth noting that the capabilities of this tool only include estimates for gravity system elements (slabs, beams, columns); if I was to include lateral system and other components, we could directly use these values. However, I reduced these values by 50% to make the comparison more accurate, as most estimates say between 40-60% of carbon in a building comes from the gravity system alone.
A preview of results can be seen below.
The tool has several potential areas for expansion in the future. For one, a feature to implement project-specifc information about location, proximity to production plants, etc. could be useful to more accurately determine A1-A3 carbon.
The visualization capabilities could be enhanced to more accurately display member sizes, although this would be somewhat complicated. However, it could be worth exploring.
Additionally, better functionality to display analysis results could be included. Right now, panels are used, which are functional, but not super aesthetically pleasing.
Also, the current analysis only includes capabilities to analyze the gravity system (beams, columns, slab); future iterations could include more customizability of these elements, inclusion of lateral system or nonstructural components, etc.
