We know from Chapter 1 that assessing building emissions is an important part of reaching net zero carbon, and that the hourly emissions are key. This section will establish the technical foundation for the subsequent discussion sections.

Part 1: Overview of Data Options

For an energy modeler, understanding that grid emissions are time-varying is the easiest part of this journey. Integrating them into modeling and analysis practice is another story. Enter the emissions multiverse. dramatic music

The first element to understand when entering the hourly emission multiverse is there are a couple different ways that the data can be sliced. I’ll be focusing on the NREL Cambium tool as this is likely what standards will reference since it’s NREL-supported and open source. I encourage folks read up on the other approaches that are out there, such as WattTime.

The descriptions of the options below reflect my understanding based on years of reading and conversations, but I’m always learning and deeply welcome input from others!

Future Scenarios

The Cambium documentation provides more information, but in general each scenario is envisioning different future scenarios for the electricity sector in terms of generation mixes and adoption rates of renewables. I lean towards using the mid-case scenario to ensure a conservative estimate, although the low-cost renewables future does seem more possible by the day.

Outputs

I’m only focusing on CO2e. If you want to learn about the other options, the NREL documentation for Cambium provides more details.

Grid Areas

Cambium has two main ways to consider a geographic area in terms of electric regions – state and balancing authority. There is another conversation to be had on utility vs. grid region, balancing authority, and imbalance market, but that’s a wormhole for another day.

State of the Grid

The question of average vs long-run vs short-run marginal is perhaps the topic that has the most chatter around it. For me, the answer ultimately comes down to what question are we trying to answer. The historic emissions of our existing electric load? Future emissions of our new electric load? Impact of building electrification? Benefits of onsite DERs (Distributed Energy Resources such as PV, batteries or grid-interactive flexible loads)? Something else entirely? The data set we use is influenced by these questions.

Here is a quick primer on the three options for hourly emissions modeling – average (AVG), short run marginal emissions rates (SRMER) and long run marginal emission rates (LRMER):

Average:
NREL Cambium definition: Average emissions rate of all generation in a given region.
The AVG universe is in theory our current grid reality – modeled emissions based on what is happening with loads and generation on the grid today. This can be useful when considering the current electric load of an existing building now or in the future if nothing in our current setup were to change.

Short-run marginal emission rate (SRMER):
NREL Cambium definition: Emission rate of the next unit of electricity considering the grid’s structure as fixed.

In the SRMER universe, the grid generation systems are as they exist today but with new loads that have surprised their systems. The grids need to bring additional generation resources online to support these new loads that weren’t in their long-term planning. It’s possible some electrification will fall into this category in the short term (ex: transportation). But generally, this scenario is going to apply to atypical or uncommon events- heat waves, winter storms, higher ventilation loads because of closed windows due to poor air quality, etc. In the current system, these will almost always be high emissions fossil fuel sources such as natural gas peaker plants.

Long-run marginal emission rate (LRMER):
NREL Cambium definition: Emission rate of the next unit of electricity considering the grid’s structure as variable.

The LRMER universe is the future where the grid knew there were new loads coming and planned for it in their base load generation resources. This is more likely what the impact of electrification will be. It might take a while to ramp up, but with proper planning the grids will be able to handle new electric loads within their base load generation resources. In theory, these generation sources will mostly be new renewable and BESS installations.

Part 2: Dear goodness that is confusing…now what?

The question of which data set to use is often framed in terms of which data is the most true data. Initially, most of the hourly emissions modeling was focused on SRMER (in fact, the original pre-Cambium term was simply “marginal emissions” which refers to SRMER). There is now a shift towards LRMER, although the debate continues between the various camps.

My perspective is it all depends on what question we’re trying to answer. For example, the challenge with SRMER is that it is looking at the past to predict the future. I hope in ten years we’ll look back on this moment in time as an inflection point when the energy industry went through a once-in-a-generation transition. If we see the rapid transition to grid-scale renewables and storage that suddenly seems possible, then historic models considering coal or natural gas as a cheap and viable resource will seem antiquated.

On the flip side, SRMER can help explain the importance of DERs and grid decarbonization in today’s reality. SRMER is essentially the ghosts of our energy future if we don’t change our peak time-of-use fossil fuel ways. When calculating the impact of a DER deployed today, before the larger systems have changed, I think it’s a fair argument that SRMER is the right approach since it likely is the real emissions avoided by DERs, at least in the near term.
Where my emissions co-pilots Forest Tanier-Gesner, Jules Earley, and I have generally landed is:

• AVG is appropriate for existing buildings and short-term outlooks (with a possible exception for parts of the NW…more on this in the next chapter)
• LRMER is the appropriate for electrification and long-term operating emissions analysis
• SRMER can be used for near term DER and flex load calculations and LRMER for the longer-term outlook.

As will be a theme throughout this series, I welcome input and hope to get a good dialogue going on, especially on this topic of data sets. Hourly emissions is in its early days so I suspect all the data is little bit right and a little bit wrong. As modelers, we should be comfortable triangulating a bit between all the options knowing that the full story likely lives somewhere in between them. Not always easy to do in practice, but it does always keep us learning!

Next up: With Great Power Comes Great Responsibility: Data from the Northwest

Additional resources:
https://www.youtube.com/watch?v=ZieoR-_IPnc
https://www.nrel.gov/docs/fy22osti/81611.pdf
https://www.watttime.org/app/uploads/2022/10/WattTime-MOER-modeling-20221004.pdf

Also published on LinkedIn by Karina Hershberg

Chapter 1: Not All That Glitters is Zero Carbon
Chapter 2: The Emissions Multiverse
Chapter 3: With Great Power Comes Great Responsibility
Chapter 4: Dr Strange Data, or, How I Learned to Stop Worrying and Love the Hourly Emissions
Chapter 5: Wear Nice Glasses and Design Beautiful Spaces
Chapter 6: No One is Net Zero Until Everyone is Net Zero