Tokenising the Energy Grid: the grid needs a settlement layer. The question is, what kind
Producing energy is no longer an issue for the electricity grid. The challenge lies in coordinating it. Currently, it is coordinated badly, expensively, and final settlement in the UK can lag by more than a year.
A failure of balance, not of supply
On the 28th of April 2025, the Iberian Peninsula went dark. Within seconds, Spain, Portugal and parts of southern France lost power: trains stopped between stations, hospitals dropped onto backup generators, and card terminals went blank for tens of millions of people at once. The initial assumption was that the full weekend-long blackout was caused by lack of electricity. But, when ENTSO-E, the body that coordinates Europe’s transmission operators, published its final report in March 2026, it reached the conclusion that the blackout cause was the exact opposite: there was more than enough electricity generation in the system, but the grid could not hold its frequency steady. Thus, the issue of orchestrating the available assets in real time and balancing the existing electricity led to a cascade of disconnections, resulting in a blackout of an unforeseen scale.
The events of the 28th of April illustrate a fundamental reality. The grid is a machine that has to stay in balance continuously, and evolving energy consumption patterns (more renewables, increased electrification, and fewer large spinning generators to lean on) are making that balance harder to keep.
Frequency is the live signal, and balancing is a market that never closes
The grid runs at fifty hertz, and that number is a real-time readout of whether supply and demand are matched at any given instant: if it drops down, demand is outrunning supply; if it spikes up, there is too much generation. In Britain, the National Energy System Operator (NESO) has set the acceptable margins for frequency oscillation, which are held within a very narrow band. Maintaining that limit is a non-negotiable and underpins the design and operation of an entire market. Unlike almost any market a finance audience deals with, it has to clear in real time, and it never closes.
With current technology electricity cannot be stored at scale. Because of this supply and demand are matched through a sequence of markets running right up to delivery, from forward and futures contracts years ahead down to the balancing and reserve markets where the operator buys flexibility second by second to hold the frequency line. The same commodity moves through all of them, but the closer you get to real time, the question becomes much less about who can generate the most, and more about which asset(s) can be coordinated the fastest. As the big thermal plants retire, that is exactly where the pressure is concentrating. The core challenge of a modern grid is that of coordination, not generation.
Coordinating badly is already expensive
When that coordination fails, the bill is not abstract. Take Scotland, which builds enormous quantities of wind but is connected to demand further south by wires too small to carry it. When the wind is strong, the system cannot move the power to where it is needed, so it pays twice over. It pays once to switch the wind farms off, around £350 million in 2025; then again, it pays to fire up gas elsewhere to replace the power it just discarded, comfortably over a billion. That is roughly £1.5 billion thrown away in a single year, and on NESO’s own projections it could reach £4 to £8 billion a year by 2030. Constraint costs of this kind reached about £1.9 billion in 2024/25, around seventy per cent of Britain’s entire balancing bill, up from forty-four per cent two years earlier.
In 2025, close to forty per cent of northern Scotland’s available wind was curtailed because the grid could not carry it. However, the cost does not fall on the taxpayer, as people often assume; it is recovered through balancing charges on electricity bills, which is to say it falls on anyone who uses power. Ofgem states constraint costs alone added about fifteen pounds to a typical annual bill at the end of 2025, with the network upgrades meant to fix the problem due to add roughly sixty pounds a year by 2030. It is geographically absurd, too: Scotland holds about a third of the UK’s wind capacity but accounts for more than ninety-eight per cent of the power curtailed. This is a coordination failure framed as a generation problem.
The problem does not stop at the border
Europe has long solved cross-border electricity coordination, and it did so not with diplomacy but with an algorithm: the EU’s single day-ahead coupling clears capacity and price across borders automatically, in one continent-wide calculation. Britain is physically plugged into that market through ten interconnectors and 10.3 gigawatts of capacity to seven neighbouring markets, but, after Brexit, it left the coupling and reverted to separate, explicit auctions in which traders commit a day in advance.
The result of leaving has had significant consequences. Inside the coupled market, capacity and price are worked out together, so power always flows from the cheaper side to the pricier one. Outside it, traders have to book interconnector space a day ahead, before the final prices on either side are known, and when they guess wrong, the power physically flows the wrong way. Namely, out of the market that needed it more and into the one that needed it less. That single inefficiency has been estimated to cost somewhere between £120 and £370 million a year, according to Energy UK’s figures. This can also be deemed an avoidable inefficiency as the technology to remove it already exists and has become standard practice in many industries.
Same grid, same commodity, different money
There is a second border problem hiding beneath the first. When electricity crosses a frontier the electron crosses in milliseconds, but the payment crosses a currency boundary and slows right down. A different currency, foreign-exchange (FX) risk, conversion cost, and separate banking rails all lead to a settlement lag the physics never has to face. And it is not a two-currency problem. It is one synchronously connected grid trading in more than ten national currencies. Great Britain’s interconnectors alone span four currencies: pounds, euros, and Norwegian and Danish kroner. Every cross-border electricity trade is therefore also a cross-border payment, which is precisely the problem finance has spent the last few years rebuilding from the ground up with shared-ledger technology.
Finance already rebuilt this layer
The story everyone told about traditional finance and crypto turned out to be wrong. Two years ago, the assumption was that the banks were interested in the crypto yield. This was in fact a much lower secondary priority to the infrastructure. Major asset managers now hold government debt on-chain (BlackRock’s BUIDL fund is the obvious example) not for crypto exposure, but because settlement and audit are far more efficient there. Western Union, a 175-year-old money-transfer company, launched a dollar stablecoin of its own in May 2026 and now settles with its global agent network on a shared ledger, around the clock, weekends included, as an alternative to the legacy interbank rails. And the value is tangible: by RWA.xyz’s count, tokenised real-world assets grew from roughly five billion dollars at the start of 2025 to over thirty billion by the middle of 2026, with credible industry projections of two to four trillion by 2030. The common thread is identical in every case: put the asset and the obligation on a shared ledger, and settlement becomes faster, cheaper, and auditable in real-time.
Fourteen months
The single biggest reason finance moved on-chain was settlement speed. A trade that used to take two days now settles almost instantly, freeing capital that the old settlement window kept locked up. Now ask, how long can it take to fully settle a single half-hour of electricity trading in Britain. The answer is fourteen months. This is not in an unusual case, but the standard course under Elexon’s standard settlement calendar. A half-hour traded this September has its books formally closed in November of next year.
I want to be honest about that figure, because it is easy to misinterpret. The cash does not sit still for fourteen months. Money first moves about a month after the event, at the initial settlement run, and the figures are then reconciled in stages, at roughly two, five and seven months, on progressively more complete data. Fourteen months is the final run, the point at which the numbers are fixed, not the moment money first moves; and if the data is disputed, even that is not the end, because a disputed settlement can run to twenty-eight months. But the underlying picture holds. This is a system that charges and credits some four million contracts a year on partly estimated data, and does not formally close its books on a given half-hour for well over a year. Finance looked at two-day settlement and payment lags and called it intolerable. The system that keeps the lights on closes its books more than a year after the event, and treats that as normal.
Why a ledger, and not simply a better database
The obvious objection is that none of this requires a token; “surely a good, shared database would do”, as many sceptics I’ve met have put it. It would not, for three reasons. First, neutrality is the product: the grid has a transmission operator, distribution operators, balancing parties, generators, aggregators and a regulator, with limited trust between the parties. A shared database needs an owner, who becomes the single point of trust and failure. By contrast, a distributed ledger is jointly verifiable, and no one party can quietly rewrite history. Second, a token can actually be the obligation; transferring it is the settlement, atomic and final, which is the very mechanism that lets tokenised assets clear almost instantly. Third, everyone can audit the same record in real time, so the dispute largely does not arise fourteen months later: the parties already agree, cryptographically, as events happen, and the regulator reads evidence instead of chasing a paper trail.
So when someone says they want to tokenise the energy grid, this is what it ought to mean. It is not a speculative coin bolted onto electricity; it is rather the same thing finance has been building itself -- a neutral, real-time, auditable layer where the value the grid creates, and the obligations between everyone on it, settle as they occur, across many parties, with no central party that has to be trusted.
The only question that matters
The grid is going to get a coordination layer of some kind. It has no real choice; the physics and the economics push the same way, and the pressure rises as the thermal plants retire and the interconnectors multiply. The oustanding question is what kind it gets: open, neutral infrastructure that anyone can build on, or something locked behind market membership and fourteen-month settlement and 28 month reconciliation periods. And the timing is not subtle. The UK and EU concluded exploratory talks on re-coupling their electricity markets at the end of 2025 and are now working towards formal negotiations; at the North Sea Summit in Hamburg in January 2026, the UK and its neighbours committed to a shared ambition of a hundred gigawatts of cooperation projects by 2050. The coordination problem is not going away. It is about to grow by an order of magnitude.