The Distribution Layer · Curtailable connections
The best prospects in Britain already chose to be curtailed.
Transmission curtailment is a Scottish problem, 500 miles from the load. But there is a second, quieter curtailment market on the distribution network — generators who accepted a curtailable connection to get online fast instead of waiting years for reinforcement. They self-selected as willing to trade firmness for speed. And they sit in London, the South East and the East of England, where the demand already is.
Why this changes the mapTwo curtailment markets, not one
Our transmission Index covers 134 wind farms losing 10.98 TWh a year — 99% of it in Scotland. That is the big number, and it is the wrong geography. The distribution layer is smaller per site and far better placed.
| Transmission curtailment (the Index) | Distribution curtailment (this page) | |
|---|---|---|
| Where it happens | 99% Scotland, behind the B4/B6 boundaries | London, South East, East of England, North East, Yorkshire, Merseyside, North Wales |
| Why it happens | Not enough wires south. National constraint. | Local network congestion. The customer agreed to it. |
| Typical site size | 30 MW – 1,075 MW | Sub-MW to tens of MW |
| Technology | Wind only (BM-visible) | Wind, solar, storage, generation of all kinds |
| Is the generator paid to stop? | Yes — ~£26/MWh turn-down bid | No. They lose the revenue entirely. |
| Distance to major load | Hundreds of miles | Often minutes |
| Visible in the Balancing Mechanism? | Yes | No — which is why nobody has counted it |
The row that matters most is the fifth one. A curtailed Scottish wind farm is compensated — it earns roughly £26 for every megawatt-hour it doesn't generate, which sets a floor under any private-wire price we can offer. A curtailed distribution-connected generator on a curtailable connection typically gets nothing. Its opportunity cost is zero. That makes the commercial conversation dramatically easier and the achievable spread wider.
The insight
These customers have already made the decision we're asking for.
UK Power Networks describes the arrangement plainly: when a generator wants a firm connection in a congested part of the network, the network may need reinforcing — which takes time and may cost the customer money. A curtailable connection lets them connect now instead, in exchange for accepting that their access can be restricted when congestion is high.
Speed over firmness
They looked at a reinforcement timeline and a reinforcement bill, and decided that connecting sooner with restrictions beat waiting for a firm connection. That is a documented, revealed commercial preference — not something we have to persuade them of.
Revenue on the restricted hours
A co-located load takes the output the network won't accept. The restriction stops costing them anything. It is the same trade they already made, extended one step — which is why this is the shortest sales conversation in the market.
Every other party approaching these generators is selling them a way out of curtailment: reinforcement, a firm connection upgrade, a battery to time-shift. All of those cost the generator money. We are the only proposition that turns the curtailment itself into income, and we are talking to people who have already signed a contract accepting it.
The data
What the DNOs publish, and what we do with it
| Network | Dataset | Records | Last updated | Licence | Status |
|---|---|---|---|---|---|
| UK Power Networks London · South East · East England | ukpn-curtailment-events- site-specific | 39,570 | 10 Sep 2026 daily | CC BY 4.0 | Key required |
| Northern Powergrid North East · Yorkshire · N Lincolnshire | curtailment-events- site-specific | 348 | 11 Mar 2026 | NPG Open Data v1.0 | Key required |
| SP Energy Networks Central/S Scotland · Merseyside · N Wales | spen_data_quality_ curtailment | — | — | — | To assess |
| SSEN · ENWL · NGED | No site-specific curtailment dataset found | — | — | — | Gap |
The UKPN schema
Northern Powergrid's equivalent adds geo_point_2d, geo_shape, site_class and licence_area — so their events arrive already geolocated and classified.
Honest status. Both datasets are real, current and verified — we have read the schemas and record counts directly from the publishers' APIs. Neither serves records without a registered API key, which is free but must be held by a named person. Until that is in place, this page describes the dataset rather than the findings. No figures on this page are estimated or modelled; every number above comes from the publishers' own metadata.
The analytical work — and the moat
UKPN's own data-governance assessment rates this dataset as publishable "with some restrictions", and mitigates privacy risk by securing DER agreement before naming customers. In practice that means many sites appear under an identifier rather than a name. For most users that makes the dataset frustrating. For us it is the entire opportunity:
- Resolve the anonymised. Cross-reference event patterns and scheme membership against capacity registers, planning records and connection data to name sites the raw file doesn't. Client deliverable.
- Rank by lost revenue, not lost energy. Distribution curtailment is uncompensated, so every kWh in estimated_energy_lost_kwh is revenue the generator never saw. That converts directly into a private-wire offer.
- Overlay the load side. An uncompensated curtailed solar farm within a few kilometres of an industrial estate or a stalled data-centre application is the most actionable pairing in the country. Client deliverable.
- Track the schemes. Curtailable connections cluster in named managed schemes. Whole clusters share a constraint — and a single co-located load can relieve several generators at once.
Anyone can download the file. Turning 39,570 anonymised events into a named, ranked, geolocated list of generators losing real money next door to real demand is a different exercise — and the output of it is released to engaged clients and subscribers, not published here.
What it changes
Three consequences for the business
1 · The addressable market roughly doubles
The transmission Index gives us 134 sites, of which 99 are commercially tractable. The distribution layer adds an unknown but substantial number of smaller sites across six licence areas — including solar, which the Balancing Mechanism cannot see at all. It also removes our single biggest weakness: that almost everything in the Index is in Scotland.
2 · The pitch gets simpler and stronger
No turn-down payment to beat, so no £26/MWh floor under our offer. No CfD basis-risk problem, because these are distribution-connected assets rather than CfD offshore giants. And a counterparty who has already signed a contract accepting curtailment in exchange for speed.
3 · The geography finally works
UKPN's three licence areas contain the London Docklands interconnection cluster and much of the South East — the densest electricity demand in Britain and the historic heart of its data-centre industry. A curtailed generator in Essex or Kent is a fundamentally more attractive private-wire counterparty than one in Caithness, because the load is already there and doesn't need relocating.
The honest caveat
Distribution-connected sites are smaller, so each deal is smaller. A 5 MW solar farm supports a far more modest load than a 240 MW wind farm. This is a higher-volume, lower-ticket segment than the transmission Index — which argues for the productised assessment and subscription tiers rather than bespoke work on every site.
Next
If you hold a curtailable connection, you are probably losing money you haven't counted.
Tell us your site and we will pull your own curtailment history from your network operator's published data, total the energy you lost last year, and tell you what a co-located load would have paid for it. No charge for the first look.
info@strandedpower.co See the transmission Index