One other important factor in this discussion is reference period. If you’re only looking at the first 5 years of a building’s operation, embodied emissions will dominate. The longer the reference period, the more we’ll see operational emissions dominate. This is obvious, of course, but I mention it to get at an important difference between the CLF and Architecture 2030 graphics in the first slide–the CLF describes a single building over a single reference period (probably 60 years?), but the Architecture 2030 graphic describes all building-related emissions between 2020-2050. This means that the longest reference period for a building captured in their analysis is 30 years (buildings built in 2020), but most will be much shorter (the reference period for a building built in 2049 is two years). This, I expect, explains some of why their result skew toward embodied emissions in a way we might not imagine at first glance.
Following up on Kjell’s point about “significant deviation,” here’s a graphic I prepared for a recent webinar about exactly this. Note that the embodied carbon estimates here includes replacements, another twist in the effect of reference period on the operational/embodied split. Still, very wide deviation by geography. These models were built in EPIC, using the Cambium mid-case data over a thirty year reference period, and conservative assumptions for embodied carbon (~960 kgCO2e/sf).
