Average or marginal emissions factors

Hi James, In general, average emissions should be used for annual carbon estimates. Marginal emissions factors should be used for comparing two scenarios either at a moment or by using hourly emissions factors. Since buildings are long-lived assets, use a long-run marginal emissions rate if it is available. NREL Cambium is what we use, going to the balancing region (aka balancing authority) (or state) scale and using the hourly factors. If you ask an energy modelers for their 8760 hourly file, you can use a Cambium download or use the ACP from USGBC/GridOptimal to compare scenarios. https://www.usgbc.org/credits/gridoptimal-152-v4?view=resources LMN is release a post as part of our Path To Zero Carbon series in the next week on this subject. -Kjell

There's another distinction to be made here between SRMER (short range marginal emission rates)and LRMER (long range marginal emission rates). The former describe the emissions of "one more" MWh at any particular time and the latter describes the emissions of having "one more" MWh of load on the grid over many years. The differences in emissions due to peaking/dispatchability/ramping rates that James refers to describes the former; the construction of additional generation capacity to meet that load consistently over time that Kjell describes is the latter.  NREL's Cambium data includes both SRMER and LRMER metrics. Important to note too that (at least in California and likely elsewhere too) Cambium data (and the ReEDS model behind it) are already out of date given recent policy changes. There's also a slew of economic assumptions behind the model that describe the import/export of power between balancing authroities so that the SRMER and LRMER metrics in a given region do not necessarily match the emissions induced by a load in that region (i.e. some of the energy to meet that load will be imported from elsewhere). We still use the data, but this caveat feels important. IN our last correspondence, Pieter Gagnon (Cambium's primary author) mentioned it as a reason to approach state-level estimates with some caution.  In terms of this choice in the context of LCA, almost all wbLCA is 'attributional' and has assumptions that track more closely with the use of AER. So if the energy-related emissions are being reported in phase B6 of a wbLCA, I would imagine that AER are preferable.  In some of our projects, we're presenting LRMER and AER results side by side. The next version of EPIC will also include the ability to toggle between LRMER and AER! For more info on LRMER, here's the article by Gagnon and Cole describing them:
https://doi.org/10.1016/j.isci.2022.103915

I’m likely too late to the party, but will chime in just in case.   For all the reasons described above, there is an element of "It depends on what question you are trying to answer" when deciding on the most appropriate emissions data set.  I suspect we all have a slightly different approach and likely all are a bit right and a bit wrong.  For what it’s worth, here is how I think of it:  - If you are trying to understand the operating emissions impact of an existing building, then the average profile is a good bet.  This is (theoretically) the profile your building is currently operating under.   - If you are trying to understand the impact of building electrification vs natural gas, then long-run marginal (LRM) is likely the most applicable.  This is (theoretically) how the grid will adjust long term to the additional baseload your new electric building is putting on to the grid.   But this one is a bit tricky because in the short term, it’s possible the short run marginal (SRM) is more accurate.  The challenge here is that most SRM models show all new loads being met with high emissions generation sources (i.e. fossil fuels) so the emissions are super high, as your data found.  This can lead to a misleading conclusions on the annual operating emissions of an all-electric vs natural gas comparison.  The reality is that (most) utilities are adjusting to the new baseload of building electrification with long term additions of low emissions resources so the long-term outlook of building electrification is more in-line with LRM.  - If you are trying to understand the impact a DER (distributed energy resource) or flexible load will have on your operating emissions, then it’s a bit of a toss up between the three.  I often will go with short run marginal (SRM) because it’s one way to consider the near term SRM vs LRM conundrum of electrification.  If its an existing building, then average might be appropriate.   The punchline is that it is all still a bit squishy and I imagine that our collective thinking on this will evolve greatly in the coming years!  (Side note to the group - If anyone happens to be attending Greenbuild in a couple weeks, Z Smith, Clark Brockman and Joshua Radoff and I will be attempting to provide some guiding principles for designing buildings ready for the future of energy in one of the sessions.  We would love to have the BG community in the audience asking the tough questions!) It’s hard to really know what the grids will look like in 10 years, but we do know the role buildings will play in larger grid decarbonization will look much different than it does today!