What It Costs to Balance the Great Britain Grid

Electricity cannot be stored in the wires. Whatever is being used at any instant has to be generated at that same instant, and the job of making that true — continuously, across a whole country — falls to the system operator.

Doing it costs money. This page is the official NESO record of how much, what it was spent on, and how it has moved month by month.

What balancing the grid means

The wholesale market settles most of the country’s electricity in advance: generators contract to produce, suppliers contract to buy, and in theory the two match. In practice they never quite do. Demand comes in above or below forecast, a plant trips, the wind arrives an hour late.

So in every half-hour settlement period the system operator makes up the difference, paying generators to turn up, paying others to turn down, and holding some in reserve for the deviation that has not happened yet. It also buys services most people never think about: frequency held within a fraction of a hertz, voltage kept in range, and the ability to restart the system from nothing.

None of that is optional, and none of it is free. The categories in the table below are those purchases, grouped by what the money bought.

The balancing bill

These are complete settlement months only. Where the published data is missing periods, the month is left out rather than charted low — a partial month plotted next to full ones reads as a fall in cost that did not happen.

Constraints: paying wind farms to stop

The largest and most contested part of this bill is not about second-by-second balance at all. It is about geography.

Britain’s best wind resource is in the north and offshore around Scotland. Most of its electricity demand is several hundred miles south. The transmission network in between has a finite capacity, and on a windy day the generation available north of it can exceed what the cables can carry.

When that happens the system operator pays wind farms above the bottleneck to reduce output, and pays generation below it — often gas — to increase. Two payments, for one unit of electricity, one of which is for not producing. It is an uncomfortable thing to read about and a genuinely rational thing to do: the alternative is building enough network capacity that a constraint never occurs, which costs far more than the constraint payments it would avoid.

The uncomfortable part is real all the same, and it is the reason constraint cost is worth watching separately from the rest.

Why the cost has grown

Three structural changes push this bill up, and none of them is a failure of anyone’s day-to-day operation.

• Generation moved away from demand. Coal and gas plants were built near cities and coalfields. Wind is built where the wind is. The distance between supply and demand has grown, and the network was laid out for the old pattern.
• Wind is variable and imperfectly forecast. A forecasting error on a large wind fleet is a large error in megawatts, and someone has to be paid to cover it in real time.
• The old plants provided stability as a by-product. Large spinning turbines held the system’s frequency steady simply by being heavy and turning. Wind and solar do not, so stability now has to be bought as a service rather than arriving free with the electricity.

Set against that, the same shift removed a much larger cost: fuel. These figures are a real expense of decarbonising the system, not an argument against it, and they should be read next to what has happened to the wholesale price of the energy itself on the price cap page.

Who pays, and how it reaches your bill

These costs are recovered through a charge on electricity suppliers, set in proportion to how much electricity their customers use. Nobody is billed for balancing directly; it arrives inside the price of a unit.

Where we show a per-meter figure, it is our own arithmetic and not an official statistic, so the numerator, the denominator and the formula are all published alongside it. Treat it as a sense of scale rather than as anybody’s actual share: because recovery follows consumption, a flat average across every meter overstates the cost to a small household and understates it to a large one.

What actually brings it down

Constraint cost is a symptom of a network that has not caught up with where generation is being built, which points at what fixes it.

More transmission capacity between north and south is the direct answer, and the slowest — new lines and subsea links take the better part of a decade. Storage helps by soaking up generation that would otherwise be constrained off and releasing it later. Demand that can move to when power is plentiful helps for the same reason, which is part of what the flexible tariffs on our Agile review are ultimately for.

The queue of projects waiting for a connection is the other half of this picture: it shows where generation wants to be built, which is where tomorrow’s constraints will be if the network does not follow. That is the grid connection queue, and it is worth reading alongside this page.