Grids — What’s the Best Green Data?
In our last post, we talked about how the American multinational technology company Google was using data and predictions about green power availability on local grids to help them move to 24/7 carbon-free operations.
Google reckons sustainability requires “access to transparent, granular, high-quality energy data.”
What kind of data are we talking about?
What does good grid energy data look like?
There has been much discussion in recent years about the most useful form of grid energy data. These debates have tended to hinge on whether users should make decisions based on the current average or the current marginal carbon intensity of the local grid.
Marginal carbon intensity
Marginal carbon intensity is about the carbon that would have to be emitted to satisfy an immediate new electricity need. For example, if you ran your dishwasher at lunchtime when there was little congestion on the grid and loads of solar power then it could all be met by lovely, low carbon intensity solar electricity. There would be more carbon emitted if you ran your wash in the evening when the grid was powered by high carbon gas or even coal.
Drawing on power when the marginal carbon intensity of the grid is low minimises carbon released. Hurray! That’s exactly what we want, so it must be the data we need!
Mustn’t it?
Why is everything so counter-intuitive?
The problem with marginal carbon intensity is it subtly combines two different issues. A grid has low marginal carbon intensity if it has enough renewable power to meet new demand. Sounds simple? It isn’t. There are two questions that need to be answered before we understand what low marginal carbon intensity actually means in any given situation:
1) Is the local grid overloaded with demand right now (congested)?
2) If it wasn’t congested, could it meet your demand using renewable power?
A grid can have high marginal carbon intensity if it is currently overloaded OR if it has little or no renewable power to meet demand. These are not the same problem. That confuses things.
Congestion changes the situation on a green grid
It’s always better to avoid placing additional demand on an already congested grid.
Even on a sunny and windy day with loads of renewable power available, an overloaded grid is usually forced to bring a fossil-fuelled power station online if it needs to meet excess demand. Meeting urgent, unexpected demand is almost always done using high carbon intensity electricity. So, high congestion => high carbon intensity => more carbon emitted. So we should pay a lot of attention to congestion!
Or should we?
What if we paid less attention?
There is a somewhat counter-intuitive thing about renewable power. It turns out that a valuable indicator of whether a location or time is good for solar or wind generation is that there’s already a lot of it there.
If you can generate power at a specific time and place it is often possible to scale that generation up. All you need to do is build more farms. NOTE — we need to worry about the space required to do that, so there are issues, but it’s a decent rule of thumb.
Data for planning vs data for reacting
We need to know the answers to our two questions if we want to make good grid planning decisions. Is the problem in a given locus physical availability of renewables or congestion? Marginal carbon intensity doesn’t tell us that. Using it to shape demand is great for the short term but may be less good in the long run because, ideally, in the long term we want to shift demand to where and when there is good renewable potential to meet it.
Marginal carbon intensity could put us off a location just because we haven’t built enough renewable farms there yet. That might not be what we want.
What about average carbon intensity?
Average carbon intensity is a measure based on the current mix of electricity sources available on the local grid, rather than any new demand’s potential marginal impact. Average carbon intensity is less impacted by grid congestion. The UK’s National Grid provides average carbon intensity in the form of an application programming interface (API) and also as a dashboard.
Making the decision to turn on your dishwasher based on average carbon intensity should drive the grid to become greener in the long run because it steers demand towards periods when there is the potential to generate low carbon energy. NOTE — again, we might eventually max out generation capacity in some areas, but we aren’t there yet.
Current thinking is that average carbon intensity may provide the best data for consumers and smart devices to base their consumption decisions on.
What’s in it for me?
However, both marginal and average intensity data have a significant drawback. There is no direct financial benefit to households, businesses, or device manufacturers in using these metrics to green their national grid. What’s in it for them?
Historically, patriotism and individual good intentions are not enough to drive major change. A sustainable grid is more likely to be achieved if the behavioural alteration it requires from users and their convenience and commercial motivations can be aligned. For that, our best data might come from dynamic energy tariffs, where current electricity price matches availability and, often, carbon intensity.
Volts vs Voltaire
The French philosopher Voltaire (who did not invent the battery) said, “the perfect is the enemy of the good”.
Lets face it, Voltaire was no Volta, but on the good vs perfect front he was spot on. None of the green energy metrics above is perfect, but they are all a heck of a lot better than no data. Basing demand choices on any of them is a big move in a right direction. We need data and we need to act on it.
Stepping back for the moment, what is it we are trying to achieve? What is this data for?
We’ll talk about behaviour change and new energy use habits, as well as how grids work, in the next post.
