Bills & Tariffs
The cleanest hour to use electricity is not always the cheapest
Grid carbon intensity varies through the day for reasons that only partly overlap with price, so optimising a household for emissions and optimising it for cost are related but distinct exercises.
By Radhika Iyer3 min read
The mix behind the socket changes by the hour
Electricity arriving at a house is produced by whatever plant happens to be running, and that set changes continuously as demand rises and falls and as wind and solar output varies. The emissions associated with a unit consumed therefore depend on when it is consumed, sometimes by a large factor across a single day.
Several network operators publish this figure, often with a short forecast, and a handful of appliances and controllers can act on it. The data is real and the variation is real. What is less obvious is how it relates to the price signal a household is more used to responding to.
Sometimes they point the same way. Sometimes they don’t, and the disagreement is instructive.
Price and emissions share a cause without sharing a schedule
Both are driven substantially by the balance between demand and available low-cost generation. When wind is plentiful and demand is low, wholesale prices fall and carbon intensity falls with them, because the plant at the margin is cheap and clean at the same time. That is the case where the two signals agree.
They part company for several reasons. Retail prices include network and policy costs that follow demand patterns rather than the generation mix, so an overnight cheap period may be cheap because the network wants load moved there rather than because the mix is clean. Nuclear output is close to constant and low carbon regardless of price. And a windy night and a still night can carry the same tariff.
So a time-of-use tariff is a decent proxy for carbon intensity and not a reliable one. Anyone who cares about the second should look at it directly.
Average and marginal intensity answer different questions
A published intensity figure usually describes the average across everything generating at that moment. What a household actually causes when it switches something on is the response of the plant at the margin, which is the unit that adjusts to meet the extra demand, and that plant is frequently dirtier than the average.
The distinction sounds academic and it changes conclusions. On a system with a large constant low-carbon component, the average may look respectable while the marginal response is a fuel plant ramping up. Consuming more in that hour has a larger effect than the average figure implies.
This is genuinely contested territory, and different published methodologies handle it differently. Treating any single number as authoritative is a mistake, though the direction of the variation across a day is usually clear enough to act on.
A solar household is already doing most of this
Using your own generation at the moment it is produced is both the cheapest and, in most systems, among the lowest-emission things a household can do, because it displaces network supply entirely. That alignment is convenient and it removes the need for any further analysis during daylight hours.
The question becomes interesting outside those hours, and particularly for a battery. A battery charged overnight from a cheap period may be charging from a clean mix or a dirty one depending on the weather that night, and a controller optimising purely on price is indifferent to which.
Some systems can now take an intensity forecast alongside a price forecast, and where both are available the household can decide which one it wants to weight. That is a genuine choice rather than a technical detail.
What is worth acting on and what is not
Moving a large flexible load — a car charge, a dishwasher, an immersion heater — into a cleaner window costs nothing and is worth doing when the tooling makes it easy. These loads are large enough for the shift to mean something and flexible enough for the shift to be painless.
Rearranging small or inflexible loads around an intensity figure is not worth the attention. Boiling a kettle at a different hour is a rounding error, and treating it as a meaningful act risks displacing effort from the measures that actually matter, which are usually about using less rather than about timing.
There is also a limit to what an individual household can achieve here, and it is worth stating. A single home shifting a few units changes nothing measurable by itself, and the value of the behaviour lies in it being common rather than in it being yours. That is an argument for making the tooling automatic and invisible rather than for staring at a dashboard.
The honest summary is that carbon timing is a modest refinement sitting on top of a much larger decision about consumption. Worth doing where it is free. Not worth reorganising a household around.
Common questions
Where does carbon intensity data come from?
Network operators and system operators in a number of countries publish it, calculated from what is generating at the time and often with a short-range forecast attached. Coverage is patchy internationally and methods differ, so a figure from one country cannot be compared directly with one from another. Check whether your operator publishes it before assuming the data exists.
Is exported solar electricity displacing dirty generation?
It displaces whatever is at the margin at that moment, which during a sunny midday on a system with a lot of solar may be relatively clean already. That is one reason the emissions benefit of generation varies with time and location rather than being a fixed figure per unit. It’s still a benefit; it just isn’t a constant one.
Should I charge my car on the cheapest tariff period or the cleanest hour?
If they coincide, the question does not arise, and often they do. Where they differ you are choosing between money and emissions, which is a personal decision rather than a technical one. A practical middle course is to charge within the cheap window but let the controller pick the cleanest hours inside it, which most scheduling systems can now do.
Features writer, Power Your Roof
Radhika writes about solar basics, batteries, bills & tariffs, mostly the parts other people skip and is unreasonably interested in the detail nobody else checks.
