[ Estimated operational and embodied carbon in urban vs. ex-urban areas - click image to enlarge ]

Discussions of carbon footprint are typically focused on emissions---the stuff that gushes out of car exhaust pipes, for example. But there is another important piece of the greenhouse gas puzzle that has so far received much less attention: embodied carbon.

In the case of a building, embodied carbon is the CO2 emitted in the process of construction.  A full accounting of a building's greenhouse gas impact must include both embodied carbon and the emissions from operation.

Likewise, a district can be assesed in terms of both operational carbon and embodied carbon. In this case, embodied carbon consists of the CO2 emitted in the construction of all the infrastructure such as roads, in addition to the buildings.

There has been very little published data on embodied carbon at the district scale. To begin to address this lack of information, GGLO conducted internal research to estimate the embodied carbon in roads in four urban areas with varying degrees of development intensity. This research was presented at the Living Future 2010 conference, and is described in further detail below. 

GGLO Team:
Dave Cutler, Michael Wishkoski, Dan Bertolet

Magnusson Klemencic Associates
Sightline Institute
Stoneway Concrete, Seattle, WA

Further Information:
Request a copy of GGLO's 2010 Living Future presentation.


[ Carbon cycle of a typical building, renovated after 30 years, courtesy Magnusson Klemencic Associates ]


Embodied carbon at the building scale
The carbon cycle of a typical building is shown in the graph above. Over the first 30 years of the building's life, embodied carbon represents about one fourth of the total carbon footprint. The graph assumes a significant renovation at 30 years, which results in an additional spike of embodied carbon, followed by a reduced rate of operational emissions (because the renovation increases energy efficiency). Lastly, at the end of the building's 50-year life, there is a final spike of embodied carbon caused by deconstruction.

The inclusion of embodied carbon in this analysis is critical to informing the decisions about whether to demolish, preserve, or renovate existing buildings. For example in the above graph, if the building was not renovated and operational emissions continued at the same rate, after 50 years total emissions would have been higher. In other cases, the embodied carbon expended to renovate could outweigh the efficiency gains, which would support the case for not investing in a renovation.

And though it may seem counterintuitive, as new buildings become more efficient, the embodied carbon in existing buildings actually becomes less valuable, because reduced operational emissions more quickly offset the embodied carbon sacrificed when a building is demolished. This is not to say existing buildings should not be valued for a whole host of other reasons, but it does underscore the importance of quantifying embodied carbon in buildings.

On the other hand, as new buildings become more energy efficient in operations, their embodied carbon becomes a larger fraction of their lifetime carbon footprint. Therefore we can expect that it will become increasingly important to minimize  embodied carbon in building construction.

Overall, embodied carbon ought to be added to the standard set of considerations that typically determine decisions about what to do with existing buildings, as illustrated below: 

Embodied carbon at the district scale
Given the importance of accounting for embodied carbon in isolated buildings, GGLO set out to investigate how embodied carbon might play a role in informing development decisions at the district scale. The key additional factor is the embodied carbon in urban infrastructure, primarily roadways, sidewalks, and utilities.

Calculations and Assumptions: Embodied Carbon
In our research we did not find any published work that attempted a comprehensive estimate of the embodied carbon in urban infrastructure. But estimates have been made for isolated roadways, ranging from 1300 to 3500 metric tons CO2-e per lane mile. For our calculations, we chose the high end of that range because we wanted to at least roughly account for all the additional infrastructure that typically comes with roads.

To estimate infrastructure embodied carbon, we used GIS to sum the length of the all the streets in the study area, assumed an average of 2.5 lanes per street, and multiplied by 3500 tons/lane-mile. 

For buildings we mined King County's GIS tax record data to sum up the floor area of all the buildings in the study area, and then used the Build Carbon Neutral calculator to estimate embodied carbon.

Embodied carbon for both buildings and roads was ammortized over an assumed lifetime of fifty years to make meaningful comparisons to annual emissions from operations.

Our research revealed another important component of urban infrastructure that is not typically considered: road vehicles. Analysis has shown that the embodied carbon associated with a road vehicle is about one fourth of its total lifetime carbon footprint.

To estimate embodied carbon produced by the manufacture and maintenance of the road vehicle fleet, we multiplied road vehicle CO2 emissions by the ratio of embodied to operational carbon determined by researchers at UC Berkeley. Operational GHGs were determined as described below.

Calculations and Assumptions: Operational Carbon
To help gauge the relative imporantance of embodied carbon, we also estimated operational carbon. For Seattle we used data from the City's GHG inventory
. For other study areas we used the following methods:

  • Road Travel: Household transportion GHGs from the Center For Neighbhorhood Technology's online calculator; add factor for commercial trucks based on Seattle's GHG inventory. 
  • Non-road travel (primarily air): Assumed Seattle's per capita emissions.
  • Buildings: Assumed operational emissions were three times the ammortized embodied carbon.
  • Industry and Waste: Assumed 10 percent of total emissions



Study Areas
Four study areas covering a range of development intensities were analyzed, as illustrated in the map above. Our estimates of emobodied and operational carbon for each study area are summarized in the table below:

Our estimates indicate the following key trends: 

  • Compared to operational carbon, embodied carbon contributes a smaller, but still significant portion of the total carbon footprint of a district.  In our four study areas, embodied carbon ranged from 15 to 22 percent of total carbon, which, interestingly,  is in the same range as would be expected for a typical building.
  • As population density increases, per capita embodied carbon decreases. This would be expected given that low density development requires longer roads to serve a given population, and result in more vehicle miles traveled.
  • Per capita embodied carbon in all buildings is not a strong function of density. This relationship is complicated by the higher prevalence of commercial buildings typically found in more dense development. For residential buildings alone, per capita embodied energy in Capitol Hill is roughy half that in Issaquah.
  • Operational carbon is inversely proportional to density, in consensus with trends that have been observed worldwide. This trend is driven by road transportation, while building operations are not strongly dependent on density.

Embodied Carbon and Growth Management
In general, accounting for embodied carbon adds further support to the case for compact development as a strategy to reduce greenhouse gas emissions, simply because less infrastructure is required. On greenfields it is clear that less embodied carbon will be expended for compact versus sprawling development. But for development in areas with existing infrastracture, embodied carbon expenditure on infrastructure may not even be a significant part of the equation.

Are there new policy and planning mechanisms by which the value of embodied carbon as a shared resource could be leveraged to bring about desired land use patterns? Could a method be devised by which the value of the embodied carbon resource could be exchanged between different districts, such that both parties benefit from their particular assets? The overall concept is illustrated below:

The role of embodied carbon in development decisions at the district scale is more complicated than the case of a single building. A choice between renovating a district and completely rebuilding it is a highly unlikely scenario. In most cases redevelopment will involve infill to an existing built environment.

One potential approach is policy that addresses the equity of per capita embodied carbon. As the data above show, people who live in lower density areas are "allocated" a greater share of the region's total embodied carbon invested in infrastructure. Perhaps municipalities could charge an additional impact fee such that development pays the hidden cost of embodied carbon. Higher-density areas would have a lower, or no fee. This would have the parallel benefit of incentivizing development in more dense areas.

In Seattle and many cities, residents in low-density residential zones often object to new development at higher density. Perhaps there could be policy devised that would offer these residents the choice of either allowing development---which would lower per capita embodied carbon---or,  if they wish to maintian lower density, paying an impact fee to offset maintaining their higher embodied carbon footprint.

A similar mechanism might be possible at the regional scale, based on the the model of transfer of development rights. The relatively high level of embodied carbon that would be expended by development on the ex-urban fringe could be offered up for sale in exchange for increased development rights in urban areas. The ex-urban area would then be off limits to development, and the net expenditure of embodied carbon would be reduced.

This investigation was intended to take a first shot at getting a handle on the importance of considering embodied carbon at the district scale. Our estimates of embodied carbon for a range of study areas is to the best of our knowledge the first published attempt at such analysis.

Our results break some new ground in establishing some important trends regarding embodied carbon in infrastructure and land use patterns.  We recognize that these results are just a first step, and look forward to the expansion and refinement of this this important line of inquiry.

Overall, we view this work as a conversation starter. We have proposed some preliminary ideas for how a better understanding of embodied carbon could inform growth management policy decisions, and welcome collaboration to collectively develop innovative solutions for creating high-performing sustainable communities.