A battery in North Scotland is turned down 41 times as often as it is turned up

Elexon publishes every accepted bid and offer in the balancing mechanism with the price it was accepted at. Reading a full year of it, unit by unit, across 6,826 MW of the 7,220 MW that traded at all, shows that a battery in Scotland and a battery in the Midlands are doing two different jobs.

Measured as full power hours a year, weighted by registered capacity so a large unit is not one vote. England and Wales earns 4.6 times what Scotland earns for each kilowatt installed.

What turned up and turned down actually mean

Volume in the balancing mechanism is measured against the unit's own physical notification, not against zero. A battery that has told the system it will charge at 57 MW, and is then instructed to sit at plus 57 MW instead, books a 114 MW upward deviation even though it spent part of that time simply not charging. So turned up covers discharging and charging less, and turned down covers charging and discharging less.

This also sets the arithmetic ceiling. A unit can swing from full charge to full discharge, so the most it can book in half an hour is its registered megawatts expressed in megawatt hours, not half of that. Every unit in the set was tested against that ceiling and one exceeded it, by five per cent, which is within the tolerance of a rounded registration figure.

It is a gradient, not a border

Split by charging zone rather than by country, the pattern is continuous and it runs north to south.

North Scotland is the extreme. Its units spent 1,186 full power hours a year being held down and 29 being called up, and were paid £3.80 for each kilowatt of registered capacity. Assets in South Wales and Gloucester were paid £49.38 for the same kilowatt, 13.0 times as much, for doing the opposite thing.

All 20 zones with a traded unit are in that table. An earlier version of this piece showed ten of them, and the seven it left out were the ones that complicate the story. The gradient survives being shown in full, but it is not a clean line. One English zone sits below the best earning Scottish zone: North Lancashire and The Lakes at £7.66. It holds 74 MW of the 6,826 MW here, so the exception is small and noisy rather than a counter argument.

Three corrections are folded into the figures above. The largest: NESO publishes a postcode with every ancillary auction result, and checking our positions against it found six units at Kilmarnock South, 300 MW between them, sitting 257 km away in Lancashire. Moving them into Ayrshire where they belong changed both sides of this comparison. An earlier version reported six times, and £10.15 against £24.73, calculated on 2,406 MW because the rest of the fleet carried no coordinate; placing Blackhillock, Coalburn, Whitelee, Neilston and Cathkin trebled Scotland's mapped capacity and widened the gap. Separately, the per unit money had been computed over 2025-08-10 to 2026-08-08 while every page printed 2025-08-21 to 2026-08-19. Thirty five of ninety four units carried a figure for the wrong window. Everything here is now computed once, on the window it declares.

What the system is using the asset for

In Scotland the balancing mechanism is using batteries as a sink for wind that cannot get south. In England it is calling them up as a resource and paying them to deliver. The same equipment, two different jobs, decided by which side of the constraint it sits.

That sits underneath a fact about the queue: roughly three quarters of the Gate 2 storage book is Scottish, and Scotland's operating assets are currently earning an eighth of what an English one earns. Whether that is a mispricing or a correct read of a future in which the Scottish constraint is worth more than the English one is a commercial judgement. The tool presents both and resolves neither.

The number is one revenue line, not profit

A battery charges through balancing bids and sells that energy in the wholesale market. The cost leg shows in this data and the income leg does not, so net balancing cash is negative for many assets and that is a purchase rather than a loss. Offer revenue, what the system paid to have the asset discharge, is the clean comparable figure and it is the one reported.

How it was checked before it was published

Summing every unit in the market for 6 to 8 August 2026 gives £43.66m against NESO's published balancing cost of £44.83m for the same three days, a ratio of 0.97. The 3% gap is the non balancing mechanism services NESO's total includes and the settlement stack does not. Across the year to 8 August 2026 the market paid £2,682m in balancing offers, of which batteries took £122.7m, 4.6%.

The zone by zone table is on the tool's Live tab, and every unit behind it has its own page.

See it on the map