Aurora has tracked TB spreads for many years as part of its analysis of flexibility markets, and publishes both forward-looking and backward-looking data for key European and global battery storage markets. The same logic is now moving into market infrastructure and trading platforms in Europe. Since July 2026, EPEX Spot has been publishing top-bottom spread indices alongside its day-ahead auction results, while its parent company EEX is launching TB futures contracts later this month.

This matters because TB spreads are no longer just a way for analysts and investors to describe price shape and volatility. They are now becoming a quotable benchmark, a reference point for investors and financiers, and a potential building block for hedges and financial contracts linked to battery storage economics.

In that sense, the rise of TB metrics looks like the storage equivalent of capture prices for renewables: not a complete model of project economics, but a simple headline indicator that the market can quote, compare and trade around.

The question is whether that shorthand is good enough. A battery earning a living from day-ahead arbitrage alone is very much the exception rather than the rule. Most batteries stack revenues across day-ahead, intraday, balancing and ancillary markets, in proportions that vary significantly by price zone or country. Battery optimisers also re-trade positions and capture value in the spreads between markets, not just the spread within the day-ahead auction.

So as TB metrics become more standardised, more visible and more tradeable, the key question is how well do they actually measure the real-world performance of batteries?

What are TB spreads and what am I comparing them to?

To answer these questions, I pulled data from Aurora’s MCP which contains historical market prices as well as our Flexplorer suite of battery benchmarks. The data relates to the last 12 months to August 2026 with a focus on nine European battery markets – Belgium, Denmark, Finland, France, Germany, Great Britain, the Netherlands, Poland and Sweden.

The first data series is the TB2 spread: the average of the top 2 hours of prices each day (or 4 half hours or 8x 15 min blocks, depending on market), minus the average of the bottom price periods. This can be displayed as a daily data series or averaged over the month. TB2 corresponds to a two-hour duration asset, which is a typical configuration across many of these markets today. It is purely a price metric and says nothing about whether any batteries captured that exact spread; it just measures how wide the opportunity was on any given day.

The second data series is estimated battery margins, taken from Aurora’s Chronos backcast benchmarks — a two-hour battery’s simulated monthly cashflow, using the actual market price outcomes for that month, across every revenue stream it can access: day-ahead, intraday, balancing, and the full range of frequency response and reserve products specific to each market. Where a market has multiple price zones with separate benchmarks (Sweden, Denmark, Great Britain), I’ve taken representative zones and averaged across them where relevant.

If you plot realised margins against TB2 spreads then you get a picture of how much of the variation in battery returns can plausibly be explained by day-ahead price spreads alone.

Great Britain: TB spreads are a reasonable if indirect proxy to BESS margins

Let’s start with Britain, where the TB spread story holds up best. Across the last 12 months, the GB TB2 spread and average two-hour battery margin data move together with a positive correlation and a strong relationship (an R² of 0.65). Wider-spread months such as June 2026 saw materially higher margins than narrow-spread months such as February 2026.

If you were looking for a market to point to and say “TB spreads do track battery returns” then Britain would be a good place to start.

But if you look inside the GB battery revenue stack the story gets more interesting. Day-ahead trading accounts for only around 13% of total battery margins in Britain on average over the last year. Intraday trading accounts for a further 26%, around 30% from the Balancing Mechanism (BM), and the remaining 30% from Dynamic Containment, Dynamic Regulation and Reserve markets combined. In other words, British batteries are mostly not earning the day-ahead TB2 spread exactly, but other revenues which move in roughly the same direction.

Does this matter if the day-ahead, intraday and Balancing Mechanism are all exposed to the same underlying drivers of system stress? A day with a wide day-ahead top-bottom spread is usually a day with a tight or volatile system, and this often also feeds through to high or volatile intraday and Balancing Mechanism prices. Day-ahead prices help shape schedules, opportunity costs and trading positions, which then influence what is left for the intraday and BM markets to resolve. However, real time outcomes in the BM are also influenced by forecast errors, asset availability, network constraints and how market participants choose to re-trade. A hedge based on a broad relationship between markets can work over long periods, but it can still break down on a day-to-day basis when real-time system conditions decouple.

What does the data say about the relationship between day-ahead and BM spreads? There is a clear positive correlation and a reasonable fit with an R² of 0.34. On average, wider day-ahead spreads tend to come with wider BM spreads, but it is a looser correlation than the previous chart.

Some days illustrate how and why the day-ahead and BM spreads can diverge. The first is the 23^rd June - the day-ahead spread was just over £140/MWh, while the Balancing Mechanism spread outturn spiked to almost £690/MWh, a sign of a very stressed system. The 23^rd June might stick in many people’s mind as it is the day when NESO issued an unusual Electricity Margin Notice during the June heatwave. I wrote about how Britain was experiencing its first “Hitzeflaute” as solar production ramped down quickly just as demand was peaking.

The following day saw a mismatch between day-ahead and BM in the opposite direction. The day-ahead spread was nearly £370/MWh (the widest of the year), but in the end the BM spread was under £190/MWh – potentially a sign that the day-ahead market was still pricing cautiously following the previous day’s events.

The relationship between day-ahead spreads and ancillary market revenues can be more complex. Bids into ancillary markets are often priced based on the opportunity cost of trading in other markets, so in principle they can move together, though in practice the strength of the relationship can be influenced by the timing of when you commit to a particular market and what information is available at the time. The plot below shows the relationship between day-ahead TB2 spread and daily average Dynamic Containment (DC) prices for the last 12 months. This shows a stronger relationship than we saw between day-ahead spreads and the BM. This could be explained by the fact that Dynamic Containment is also traded day-ahead (albeit in 4-hourly blocks). DC prices also spiked on both the 23^rd and 24^th June but at a level that was reasonably predictable given the day-ahead prices. Similarly, Dynamic Moderation, Dynamic Regulation, Quick Reserve and Balancing Reserve are all traded day-ahead using the same co-optimisation platform so often move together. It is the Balancing Mechanism which is the outlier.

Zoom out to Europe and TB spreads are far less good as a predictor of BESS returns

I extended the analysis by pulling in data on day-ahead TB2 spreads and 2-hour BESS margins for a further 8 European countries. Broadly speaking there is still a positive relationship, with the lowest BESS margin months seeing low spreads and high margins explained by high spreads. But with the R² dropping to 0.37, the plot indicates that most of the variation in battery margins between countries and across months is not explained by day-ahead spreads alone.

Interestingly, the scatter isn’t randomly distributed around a single line; entire markets sit systematically above or below where the overall trend would put them.

• France is the standout market in this plot. At any given TB2 spread level, French battery margins are well above what the cross-market trend would predict, by an average of over €22/kW/month, the largest positive gap of any market in the sample.

• Germany, the Netherlands and Poland sit close to or modestly above the trend line – displaying high BESS margins but also high TB2 spreads.

• Denmark sits well below the trend line, with BESS margins below what their spread levels would suggest.

These trends are not arbitrary, but rather a function of what each market pays batteries to do.

Why France breaks the trend

French battery margins are high compared with those in other markets, at several times the levels seen in Britain or Finland. But what is particularly striking in Aurora’s analysis is how little of the French BESS revenue stack comes from wholesale trading. Day-ahead and intraday trading combined accounted for just 13% of battery margins over the last year. The remaining 87% came from ancillary services, dominated by aFRR capacity and energy payments.

aFRR is a balancing service with separate payments for capacity and activated energy. RTE procures the required upward and downward aFRR capacity through a national daily tender. French aFRR capacity prices have only a loose relationship with day-ahead spreads - the relationship between average daily aFRR capacity prices and French day-ahead TB2 spreads is weak, with an R² of 0.10. In other words, variation in the TB2 spread explains only around 10% of the observed daily variation in aFRR capacity prices. aFRR capacity prices are mainly determined by the balance between procurement volume and the amount of prequalified capacity available, competition between batteries and other providers, and the opportunity cost of reserving capacity rather than using it in other markets. Day-ahead spreads can affect a provider’s opportunity cost, but they are not the only or main driver of the capacity clearing price.

aFRR energy prices have a more direct link to wholesale prices because providers’ bids reflect their generation costs and wholesale positions. That said, the price of activated aFRR energy is also shaped by real-time system imbalances, the volume and direction of activation, and conditions elsewhere in the European balancing market.

France has traded aFRR energy through the PICASSO platform since April 2025. Bids enter a European merit order, and the marginal activated bid meeting the French requirement may come from France or another participating bidding zone, subject to cross-zonal capacity. This means that French aFRR energy prices are not determined solely by French day-ahead market conditions.

This shows up in the data as a weak relationship between French aFRR energy prices and day-ahead prices (R² of 0.14). This is only slightly stronger than the relationship between day-ahead TB2 spreads and daily average aFRR capacity prices, and substantially weaker than the relationship between day-ahead and Balancing Mechanism prices in Britain. There can sometimes be extreme spikes in aFRR energy prices which have little to no relationship with the day-ahead market (for example on 11^th April 2026, upwards aFRR energy prices hit €4,827 when the corresponding DA price was near zero).

Furthermore, price correlation is not the full story. The aFRR energy revenues earned by a battery also depends on whether a particular bid is selected and activated. So this further weakens the relationship between day-ahead spreads and realised aFRR revenues.

Overall, French day-ahead spreads have at most a weak influence on total battery revenues over the last year. French batteries have earned healthy aFRR payments across both high and low-spread periods, while day-ahead and intraday trading have provided a relatively modest top up in the overall revenue stack.

This explains why France sits above the European trend line shown above. It is not that French batteries are capturing day-ahead spreads more effectively than in other markets. They have benefited from a substantial ancillary-service revenue stream that is only weakly related to French day-ahead spreads.

Why Denmark undershoots

Denmark presents the opposite case to France. Day-ahead spreads can be relatively wide, but the margins captured by Aurora’s battery benchmark are modest, and below the trend relationship between TB spreads and margins seen across Europe earlier. This reflects the composition of the Danish revenue stack rather than an absence of wholesale price volatility.

In the Aurora benchmark for Denmark (DK2), most of the revenue comes from FCR and mFRR, whose prices are determined by reserve-market supply, demand and activation rather than directly by the Danish day-ahead TB spread. These markets have become increasingly competitive as batteries and other flexible providers have entered.

The relevant markets are also larger than Denmark itself. Denmark participates in the Continental European FCR Cooperation, while Danish mFRR capacity has been traded through a common market with Sweden and Finland since 2024. Reserve prices can therefore be set by flexible capacity elsewhere in the connected market, so are not always determined by Danish wholesale market conditions.

For mFRR energy, revenue depends not only on the price but on whether and how much the battery is activated. A large day-ahead spread reflects an anticipated difference between low- and high-price periods; it does not necessarily imply a large real-time system imbalance. Even where balancing prices are high, a battery may earn relatively little if activation volumes are small or its bid is not reached in the common merit order.

This explains both Denmark’s low margins and their weak correlation with day-ahead spreads (the correlation between monthly average day-ahead TB spreads and BESS margins is only 0.2, the weakest correlation of any of the nine markets considered above). Danish batteries are competing in increasingly well-supplied regional reserve markets, while day-ahead arbitrage represents only a modest part of the optimised revenue stack. The market may display attractive spreads, but those spreads do not translate one-for-one into battery earnings after reserve commitments, operating constraints and competition are taken into account.

What this means for TB futures

I don’t want to conclude by being too negative - TB spreads are not a bad metric per se. Their rise is useful because they give the storage market a simple language to describe price volatility, and potential revenues, just as capture prices do for renewables. But they are still a metric for the day-ahead market, not a complete measure of battery economics, and are far from a perfect fit.

All of this matters directly for the new EEX TB futures. A TB future hedges exposure to the day-ahead spread. Where battery revenues are closely linked to wholesale price shape, such as in GB and Germany, that can make TB spreads a useful proxy and partial hedge. Even in Britain the relationship is indirect, because much of the revenue comes from intraday, the Balancing Mechanism and ancillary services rather than day-ahead trading itself; but the different revenue streams are at least exposed to similar system-stress drivers and correlate well with day-ahead spreads.

In markets such as France and Denmark, the case is much weaker. French battery revenues have been dominated by aFRR capacity and energy payments, while Danish margins are heavily shaped by FCR and mFRR markets. These revenues are driven by reserve procurement, competition from other technologies, activation risk, cross-border balancing arrangements and asset-specific opportunity costs — not simply by the day-ahead TB spread. In those markets, a TB future partly hedges one part of the revenue stack, but the basis risk for the full revenue stack is very large.

So overall, TB spreads are a useful broad-brush indicator for BESS revenues but need to be understood and caveated and not simply taken at face value. Before treating a TB spread as a measure of what a battery will earn in a particular price zone or country, you should ask yourself: what is the market actually paying batteries for — and how much of that value is really linked to the day-ahead price?

Today’s announcement by the UK Government is different. It is about the Government getting out of the way — removing red tape so households can generate a little more of their own electricity.

From today, it is legal to sell plug-in solar panels in Britain. Small solar systems of up to 800W can now be plugged directly into a domestic socket without the need for a specialist installation, once the products are certified as safe. After years of regulatory barriers, Britain is joining countries such as Germany where plug-in or ‘balcony solar’ is already becoming commonplace.

At first glance, this might sound like a niche regulatory change, but it could prove significant at consumer and system level.

This extends the democratisation and decentralisation of energy that we’ve already seen with the growth in rooftop solar, EVs, home batteries, and to a lesser extent heat pumps. Plug-in solar is another step in the same direction: making electricity generation accessible to a larger group of households with lower upfront costs and fewer barriers.

This can result in savings for households. Government estimates that households could save up to £110 per year, though as I’ll explain below, the exact savings for a particular household can vary.

But there is a potential kicker that we should watch carefully – uncontrolled growth in behind the meter solar could exacerbate the lack of visibility which system operator NESO has over the grid, which already led to unexpected electricity system issues during the heatwave this summer.

Where can I buy one?

One interesting thing about today’s announcement is that the regulations seem to have moved slightly faster than the market.

Reading some of the coverage, you might assume there is already a flourishing plug-in solar market waiting to serve British consumers. In practice, there appear to be relatively few compliant products available for immediate purchase today.

There is a good overview of what is available and coming soon in this article from the Independent.

I had a quick look for actual products on offer today. The first example was from UKSOL, which is offering an 890W plug-in solar kit priced at £1,089. Amazon seems to have cheaper products available too, such as a 2x500W kit for £549, though the output is clipped with an 800W inverter.

Even if there are few products available today, the situation is likely to change pretty quickly. A range of major retailers have already indicated plans to offer plug-in solar products, including Argos, Currys, Wickes, Screwfix and B&Q. The government’s announcement suggests these products should begin appearing in stores and online over the coming months, which should drive down prices further from where they are today.

How much could you save?

The Government’s headline is that households could save up to £110 per year. As with most energy economics, the reality is a little more complicated than a single number.

There are a few things that matter.

First is the upfront cost. The two examples I gave above have very different costs with the UKSOL offering at £1.20 per watt and the Amazon system at £0.55 per watt.

Second is how much electricity the system generates. That will vary depending on location, orientation, shading and weather conditions.

Third is your usage pattern and level of self-consumption. Every unit of solar electricity that you use directly in the home is a unit of electricity you do not need to purchase from the grid. At the current Ofgem price cap, we can think of this as an avoided cost of around 26.3p/kWh. But households use electricity very differently. Someone who is retired or works from home may consume a high proportion of their solar generation directly, but a family that leave the house during the day might export a greater share of their production to the grid. Some consumers might optimise their EV charging or appliances to run in sunny hours, but not everyone has the option or level of interest to set this up.

The final factor is the export tariff. If you don’t have a smart meter, you may receive nothing at all for exported electricity. With a standard Smart Export Guarantee tariff, you receive something like 4.1p/kWh. With more generous export products, payments can reach around 12p/kWh.

The important point is that two households buying exactly the same solar kit may experience materially different savings depending on how they consume electricity and what export arrangements they have in place.

I created a simple calculator where you can play around with the key variables. A conservative case could be the UKSOL setup with 50% self-consumption, the 4.1p/kWh SEG tariff, the Ofgem retail price, and a lower level of generation due to orientation or shading. This delivers bill savings of £88 per year and a payback period of 12.4 years. An optimistic case could be the Amazon 1kW system, 80% self-consumption and a 12p/kWh export tariff and a higher production. This boosts the savings to £234 per year and reduces the payback to as little as 2.3 years.

Link to live calculator: https://marvelous-heliotrope-1feb2f.netlify.app/

Optimistic case example:

Could plug-in solar reshape the rooftop solar market?

One aspect that so far seems to have received less attention in media commentary is the impact this change could have on the wider domestic solar industry. Plug-in solar is a competitor with rooftop solar – not a perfect competitor, but a competitor nonetheless.

I looked at the Octopus Energy rooftop solar cost calculator to get a sense of the latest costs for rooftop solar. For my home location it quoted a system of 10x 455W panels (so 4.55kW in total) would cost £8,149 installed, equating to £1.79 per watt – significantly above the plug-in systems. It estimates that this would reduce my electricity bill by around £54-64 per month.

What is interesting is the cost structure. The calculator suggests a two-panel rooftop system costs around £6,663, and each panel you add costs £177. This suggests that more than £6,000 of the cost is effectively fixed - relating to scaffolding, labour, electrical works, administration and other installation costs.

This helps explain why rooftop solar tends to work best for larger systems. Once the scaffolding is up and the installer is on site, it often makes economic sense to add more panels. A 2-panel, 900W system for £6,663 does not make sense at all (£7.40/watt!) so households tend towards around the 3-5kW+ range. Though it is worth noting that as you make the system larger and larger, the degree of self-consumption likely drops unless you can alter your usage patterns.

This is why rooftop PV systems are increasingly sold with battery storage, adding perhaps £1,200-£5,300 to costs but significantly improving self-consumption and reducing exports. Rooftop solar (with or without storage) can also pair well with a smart charging EV which can use excess power before it is exported to the grid.

Plug-in solar flips the economics around and opens up the smaller-scale segment of the market. A household can buy an ~800W system for a few hundred pounds, and avoid virtually all installation costs. It also probably wont be long before there are plug-in batteries available in the UK market, which could further improve the economics just as with the larger rooftop offerings. Such batteries are already available in Europe. So far the sale of plug-in batteries has not been legalised in the UK, but it seems that the Government is keeping them under further review.

Overall this suggests that the sweet-spot for rooftop and plug-in solar are quite different. For large households with substantial electricity consumption, a conventional rooftop installation is likely to remain the superior option, especially when the household has an EV or can afford a battery. Plug-in solar caters for smaller households with lower consumption, renters, flat-dwellers, and those who cannot afford the upfront cost of rooftop.

Because plug-in solar is a substitute product, even if only for part of the market, it could place downward pressure on installation for rooftop solar over time. In other words, rooftop solar providers are now not only competing against other installers, but against much cheaper DIY solar kit delivered to your front door.

A glimpse of a bigger future

The real significance of today’s announcement is not what happens in the short term, but what happens as adoption begins to scale.

Germany provides an instructive example. Plug-in solar systems, often marketed as “balcony solar”, have become very popular. As of April 2026 there were 1.3 million officially registered balcony solar systems, totalling 1.3 GW of capacity. It is estimated that 20-28% of these systems are being installed with a battery pack. Whilst registration is legally required, there is no financial incentive or enforcement, meaning that the actual scale of deployment could be higher still.

There is no guarantee Britain follows the same trajectory. There are differences in consumer behaviour, housing stock, and electricity prices. Households are often reluctant to switch retail provider to save £50, so they may not bother to install solar either. But going by the recent growth in rooftop solar in Britain it does seem that consumers are taking action.

Viewed individually, an 800W system is small. But if, say, 20% of the 29 million households in Britain bought an 800W system then this translates to 4.6 GW of capacity! These scales of deployment have very different impacts on the power system…

Be careful what you wish for

This brings us to the more challenging and paradoxical side of this revolution in consumer energy.

The spread of rooftop solar, batteries, electric vehicles and flexible demand is generally a good thing. These technologies reduce bills, lower emissions and can improve energy resilience.

But they also make operating the electricity system more complicated.

As I discussed in my recent article on the UK’s first “Hitzeflaute”, one of the emerging challenges for NESO is visibility. Increasing quantities of generation and flexibility now sit behind the meter, beyond the direct visibility enjoyed in traditional power systems. The issue is not renewable energy itself. The issue is coordinating and forecasting a system increasingly shaped by millions of distributed assets. How do you forecast demand and generation accurately when growing amounts of both sit outside traditional monitoring frameworks?

This challenge already exists today. Rooftop solar, household batteries, EVs and heat pumps are already invisible to the grid operator. Plug-in solar just adds another appliance type and acclerates the issue.

Australia provides a glimpse and warning on where this process can lead. Australia leads the world in per-capita residential solar adoption, boasting over 28 GW of total rooftop capacity. This was fine at first, but as deployment grew it created a problem - in some Australian states such as South Australia, rooftop solar generation alone can exceed total demand during periods of strong sunshine. This can create severe voltage instability – of the kind that contributed to the Iberia blackout, as I discussed in this blog.

To resolve this, policymakers have increasingly explored policy and regulatory mechanisms such as requiring remote kill-switches, cutting solar export payments, requiring retailers to offer attractive Solar Sharer tariffs to encourage daytime power use, and stimulating the deployment of grid scale batteries.

This case study reminds us that successful technologies often create new challenges. The more successful distributed solar becomes, the more important coordination, forecasting and system visibility become. In some respects, that is a nice problem to have but good to call out and essential to manage proactively.

Plug-in solar panels are unlikely to slash everyone’s energy bills or transform Britain’s electricity system overnight. Yet today’s announcement matters because it removes another barrier between consumers and energy. Rooftop solar began that process but plug-in solar extends it to a wider consumer base. If Britain follows the path seen elsewhere, it may eventually be seen as another step in the long transition from an electricity system dominated by large, centralised assets towards one shaped by millions of decisions made at the edge of the grid.

Let’s just make sure that we stay ahead of this from a system operability perspective.

This experience came in contrast to my last EV road trip in France back in 2022. At the time, charging infrastructure felt far less mature. I often found chargers that were unavailable, unreliable, or requiring a specific network payment card rather than a bank card or app. Long-distance EV travel was certainly possible in France back then, but it required significantly more planning than it does today and at times was stressful. I summarised my learnings in this post on X/Twitter.

The difference now is remarkable. France has gone from around 82,000 public charging points at the end of 2022 to over 200,000 today, whilst the UK has also expanded from around 57,000 points in 2022 to more than 120,000 today. Both countries have made enormous progress, but the network in France has grown faster and now has substantially more public charging infrastructure than Britain despite a smaller fleet of EVs on the road: roughly 1.7 million BEVs in France versus more than 2.1 million in the UK.

Public EV charging in France is widely available even in smaller towns and rural areas (arguably better than the UK in this respect). Most chargers can be activated with a bank card or app (I used ChargeMap and Electroverse). From a driver’s perspective, the charging experience now feels at least as good as the UK and arguably better in some respects.

The thing that surprised me most, however, was not the availability of the charging infrastructure. It was the price.

Again and again I found myself looking at charging tariffs and comparing them to what I would expect to pay back home. The difference was consistently large enough that I couldn’t ignore it and starting asking myself: Why is EV charging so much cheaper in France than Britain?

Public EV charging is much cheaper in France

Let’s start by looking at some cost comparisons. For this we need to consider different charging types from at-home to ultra-rapid motorway charging.

For at-home charging, it is a bit of a mixed bag. If we look at standard domestic tariffs then EDF’s regulated Tarif Bleu currently costs around €0.20/kWh, equivalent to about 17p/kWh, compared with the current UK Ofgem price-cap unit rate of 26.1p/kWh. But wherever possible an EV owner would use a smart EV tariff, because that is where the economics really improve. I use the Octopus Intelligent Go tariff at home, which costs 6.57p/kWh – or less than 2 pence per mile at a typical efficiency. Interestingly the equivalent smart EV tariff in France (EDF Vert Electrique Auto tariff) is quite a bit more expensive at €0.13/kWh. So whilst the baseline household electricity price is lower in France, at-home charging can still be very attractive in Britain. As an EV owner with a home-charger, this is where I do the vast majority of my charging (maybe 80-90%).

The more interesting comparison is public charging, which can be much more expensive than at-home charging, particularly in the UK.

The latest UK market averages are around:

• 54p/kWh for standard public charging

• 84p/kWh for rapid and ultra-rapid charging

To put this in context, the 84p/kWh ultra-rapid charging rate equates to over 20 pence per mile, which is more than the ~17 pence per mile for a typical diesel car averaging 50 mpg.

The French equivalents (converted to GBP) are approximately:

• 40p/kWh for standard public charging

• 44p/kWh for rapid charging

• 52p/kWh for ultra-rapid charging

In other words, France is around 14p/kWh (26%) cheaper for standard public charging and around 32-40p/kWh (39-48%) cheaper for rapid and ultra-rapid charging. This also puts the cost of ultra-rapid charging below that of a typical diesel car.

The above figures are indicative mid-points of the range, and there will be some variation by provider and location type. So it is also useful to make some direct like-for-like comparisons:

Fastned charges £0.79/kWh for ultra-rapid charging in Britain. The same company charges €0.61/kWh in France, equivalent to 52p/kWh. That implies a gap of around 27p/kWh: same operator, technology and business model, but a very different price.

Lidl is a useful retail comparison. Lidl France offers DC charging at €0.39/kWh, roughly 33p/kWh, while Lidl UK’s rapid charging is currently 62p-74p/kWh depending on payment method.

IONITY is active in both markets and shows the same pattern. In the UK, direct or app-based charging is currently around 83p/kWh, with subscription rates from 47p/kWh. In France, app/direct charging is €0.62/kWh (or 53p/kWh) with subscription rates as low as €0.31/kWh.

At this point it becomes difficult to argue that the difference is about operator pricing strategy. The same gap appears across multiple operators and charging categories.

What sits underneath the headline EV charging cost?

Most EV drivers see public charging as a simple retail transaction. You plug in, charge and pay a tariff. But the underlying economics are considerably more complicated.

A Charge Point Operator (CPO) is not simply buying electricity and selling it on. Public charging combines electricity procurement, network charges, taxes, charger depreciation, maintenance, payment systems, land/site/concession costs, utilisation risk and margin. An important distinction is that CPOs are commercial electricity users, and the economics they face are shaped by commercial tariff structures rather than household price caps or tariffs.

All of this sits beneath the retail price visible to the driver but drives big differences in the charges they pay.

Wholesale electricity costs explain part of the gap

Wholesale electricity costs are part of the explanation, but not a complete one.

During the energy crisis, France was actually more expensive than Britain: French wholesale power averaged €276/MWh (or £236/MWh) in 2022; higher than GB’s £198/MWh. Since then, however, the comparison has moved clearly in France’s favour. French wholesale power averaged €61/MWh in 2025 and €71/MWh in 2026 year to date, compared with £80/MWh and £97/MWh respectively in GB. At today’s exchange rate, that implies a GB premium of roughly £27–35/MWh, or about 3p/kWh. So cheaper French wholesale power matters, but it is nowhere near large enough to explain a public rapid-charging gap of 14–28p/kWh by itself.

One important nuance is that large electricity consumers in France (including CPOs) have preferential access to long-term nuclear-backed power contracts in France under the Contrats d’Allocation de Production Nucléaire (CAPN). This mechanism targets 10-15 year contracts at around €70/MWh as well as pay as produced contracts for 3-5 years. This scheme replaces the previous ARENH mechanism which provided nuclear power to businesses at a regulated price of €42/MWh.

If we look at Britain specifically it is clear that charging costs have detached from the trend in wholesale prices. Whilst wholesale prices roughly halved from 2022 to now, ultra-rapid charging costs have actually increased by 20-40% over the same period, as shown in this chart from the RAC.

Running a high-power charging hub is not simply about the volume of electricity consumed. It is also about the capacity required to connect to the grid and then deliver that electricity to end users.

A motorway or rapid charging site may need a large connection capable of serving multiple vehicles simultaneously at high charging speeds. The grid connection has to exist whether one or twenty cars are charging simultaneously. From a network perspective, the capacity matters as much as consumption.

As an example, down the road from me is the Energy SuperHub in Oxford, which was touted as Europe’s most powerful EV charging site when it launched in 2022. It currently has 45 charging points but has 10 MW of grid connection capacity with a plan to scale to up to 400 charge points. This required a four-mile underground cable connecting to the high voltage transmission network. The grid connection cost is not in the public domain but the whole project cost £41 million, and this needs to be recovered through charging fees.

The cost of connecting a high-power charging site to the grid can be very high in the UK. A CPO typically pays for the site-specific physical connection to the grid – which can be very material for multiple rapid or ultra-rapid chargers. In addition, they contribute to wider reinforcement and system costs through ongoing network charges, with recent reforms changing how these costs are recovered. Access SCR reduced some upfront reinforcement contributions, but more of the wider network cost is now recovered through ongoing standing, capacity and use-of-system charges. Alongside the Targeted Charging Review, this means that high-power charging hubs can face a large fixed-cost burden even when utilisation is still low. Industry estimates suggest that, for some rapid or ultra-rapid sites, these network and standing charges can add materially to the per-kWh economics — in some cases by 20–30p/kWh, or a significant share of the final retail charging price.

France takes a slightly different approach. Network costs are recovered through the regulated TURPE framework, set by the regulator CRE, and French CPOs still pay network charges. But France has also allowed part of the grid connection cost for some public charging projects to be socialised through the tariff framework, with generous support for motorway charging infrastructure. That does not make network access free, but it reduces the upfront burden of connecting high-power public charging sites and lowers the pressure to recover those costs through each kWh sold to drivers.

Policy costs and taxes add another layer

In Britain, CPOs face non-domestic electricity bills that include policy and system charges such as the Renewables Obligation, Contracts for Difference, Feed-in Tariffs, Capacity Market costs and the Climate Change Levy, which total around 7.6p/kWh before VAT. France also recovers policy and system costs through electricity bills, including the Accise sur l’électricite at closer to 2.7 €cents/kWh. This is not a perfect like for like comparison, but it suggests that UK policy-cost recovery may be adding several pence per kWh more to public charging than the equivalent French component.

VAT is often raised in UK debates around EV charging costs because public charging is taxed at 20%, while domestic electricity is taxed at only 5% (and soon to drop to 0%). That matters for fairness: drivers without access to home charging pay a higher tax rate than those who can charge on their driveway. But it is not an explanation for the France–UK public-charging gap, because public charging in France is also priced including standard VAT.

Utilisation is another important factor. Public charging infrastructure is expensive to build, and sites constructed ahead of demand can spend years operating below their eventual utilisation levels. Those costs still need to be recovered. But utilisation does not explain the France–UK price gap. If anything, France’s larger public charging network relative to its BEV fleet likely means lower utilisation and upward pressure on costs.

Looking across the components, it seems that lower French wholesale prices could explain around ~3p/kWh of the gap, policy costs perhaps another 5p/kWh but the network charges may be where the biggest differences are found in the cost stack (if readers have hard data on this then please share in the comments!)

Is the UK EV charging market as competitive as France?

Competition is potentially part of the story, particularly for motorway service station chargers.

In the early stages of EV adoption in the UK, Ecotricity’s Electric Highway network played an important role in establishing national motorway charging coverage. At a time when EV numbers were small and the commercial case was uncertain, that investment arguably helped create a viable charging network far earlier than might otherwise have happened.

However, the consequence was that much of the motorway network became tied to a single provider through long-term exclusivity arrangements. When Gridserve acquired Electric Highway, it inherited many of those arrangements.

The Competition and Markets Authority subsequently examined EV charging at and near motorway service areas. It found that long-term exclusivity arrangements covered a large share of UK motorway service areas and had the potential to restrict competition. In 2022, the CMA secured commitments preventing those exclusivity rights from being extended or enforced beyond November 2026, which over time may allow greater competition at motorway charging sites.

France is not a perfect example of open competition either. Motorway charging provision is often awarded through concession processes which may result in a single operator per site, though I saw examples of direct side-by-side competition on my recent travels. The market has developed around a broader mix of providers including TotalEnergies, IONITY, Engie and Tesla and has not historically been dominated by a single network in the same way.

Competition matters and we should ensure that the UK market does open up, but it is not the primary explanation for the charging cost gap. The UK-France EV charging price gap is simply too large, and the gap exists both for motorway charging and standard public charging.

The elephant in the room: electricity is expensive in Britain

What started as an article about EV charging costs ends up raising a much broader set of questions about end-user electricity prices as a barrier to electrification.

France has managed to build a large public EV charging network while maintaining significantly lower charging prices than Britain. The explanation is not a single policy, a different business model or a market feature. It reflects the cumulative effect of lower wholesale prices, commercial tariff structures, network charging arrangements and policy costs (and in the case of motorway service area charging may have been exacerbated by a lack of competition).

A driver looking at an 85p/kWh tariff is not thinking about all of these different components of the cost stack, yet this is what they end up seeing on the app or screen.

EV owners in Britain face materially higher prices for public charging. For EV owners with their own charger and driveway it probably doesn’t matter that much, as the en route charging is a small part of their charging behaviour. Around two-thirds of households have a driveway in Britain. But if we really want to see wide adoption of EVs across ALL household types, then we need to address the high cost of public charging.

In a way this brings us to the central argument of the recent report by Onward, which I critiqued in this LinkedIn post. It argued that the real prize for decarbonisation is not to chase renewables targets, but to double down on electrification – yet as electricity in Britain is so expensive then this discourages individuals and businesses from switching to cleaner solutions. I did not buy into Onward’s Alternative Policy Pathway or the assumptions sitting behind it, but they were absolutely right to point out the elephant in the room.

That is why the EV charging price gap matters. It is not just an EV story, but rather a reminder that unless we make cleaner solutions also cheap and a compelling customer proposition, then this will hold back take-up. In Britain, this remains the unresolved part of our decarbonisation plan.

Today, those questions feel much less contested than they once did. As I discussed recently with Sonia Dunlop from the Global Solar Council, solar PV has now surpassed 3 TW of installed capacity globally, a milestone that would have seemed extraordinary only a few years ago.

The more interesting question is no longer whether solar and wind can scale, but whether the rest of the system can keep pace.

That theme ran through my latest podcast, a special edition to mark the 300th episode of Energy Unplugged. I was joined by Aurora’s five regional Managing Directors: Ana Barillas covering Iberia and Latin America, Hugo Batten (Asia-Pacific), Hanns Koenig (Europe), Dan Monzani (GB, Ireland and Africa), and Oliver Kerr (North America).

What struck me most was how often the discussion converged on the same underlying challenges in very different parts of the world: solar cannibalisation in Chile and Europe, transmission congestion in China and Germany, connection queues in Britain and the US. The conversation kept returning to a set of common themes. The challenge for the energy transition is no longer simply building clean generation. It is whether grids, markets, regulation, flexibility and demand can keep pace with the speed of change.

From renewables scarcity to abundance

The first part of the discussion focused on a trend that is becoming increasingly visible across global power markets: renewables abundance.

In the early part of the energy transition, the problem was a scarcity of renewables and clean power. Solar and wind were expensive, deployment was subsidy-driven, and the main policy challenges were how to bring costs down, scale supply chains and persuade investors that these technologies could become mainstream. This has changed very quickly. In market after market, renewables have moved from marginal technologies to major sources of generation, with build rates that would have looked implausible a decade ago. Clean energy investment now significantly outpaces investment in fossil fuels: the IEA expects around $2.2 trillion to flow into clean energy in 2025, roughly twice the amount going into oil, gas and coal.

The result is new kinds of growing pains: power systems are now having to absorb large volumes of weather-dependent generation before grids, storage and market design have fully adapted.

Ana Barillas gave some of the clearest examples of what this shift looks like in practice. Across Iberia and Latin America, the story is no longer simply that renewables need to scale faster; in several markets they already have. Spain has built renewables quickly, but demand growth, electrification and grid expansion have not kept pace. In northern Chile, solar build-out has been so strong that capture prices have fallen to levels that would have looked almost unimaginable a few years ago, putting serious pressure on generator profitability. Brazil, meanwhile, has seen extraordinary growth in distributed solar, raising harder questions around cost allocation, system efficiency and network use. Her broader point was that markets can move from scarcity to abundance much faster than investors expect — and that the industry is often better at forecasting renewables build-out than at forecasting when that build-out will start to undermine its own economics.

The same transition is visible in Asia, albeit at a vastly different scale, as Hugo Batten explored through a deep dive on China. I challenged Hugo that there is an apparent slowdown in renewables deployment in 2026. He responded that there is a tendency in Western commentary to interpret any moderation in Chinese renewable deployment as evidence that the market is slowing. Yet China installed around 400 GW of wind and solar capacity in 2025 alone; even a fall from that level would still imply deployment on a scale unmatched anywhere else. The more important story is not a loss of momentum, but a shift in focus. China is moving beyond the old model of guaranteed prices, build-anywhere incentives and the assumption that the grid will take whatever is connected. New projects are being pushed towards greater market exposure, with more output sold through auctions, contracts for difference and spot markets. At the same time, China is still pursuing transmission expansion at extraordinary scale, including further west-to-east HVDC corridors. Even in the world’s largest renewables market, the challenge is becoming less about whether assets can be built and more about whether flexibility, inter-province trade, transmission and market design can integrate them effectively.

Readers of this Substack will recognise this theme from my recent pieces on duck curves, negative pricing and solar cannibalisation. The discussion with Ana and Hugo reinforced the same points. The next decade may be defined less by the challenge of building renewable generation and more by the challenge of extracting value from increasingly abundant renewable generation.

The growing importance of grids

The second theme was grids — and the extent to which they are becoming the binding constraint for the energy transition.

One statistic from the episode stood out. Aurora’s recent analysis found that 72 TWh of mainly renewable generation was curtailed in Europe during 2024 because of grid constraints, roughly equivalent to the annual electricity consumption of Austria. Congestion management costs reached almost €9 billion — a reminder that grid constraints are no longer just an engineering problem, but a material cost of the transition itself.

In the early phases of the energy transition, the grid was not usually the binding constraint. Networks generally worked well enough to absorb the volumes of renewable power being built, so the main challenge was to accelerate deployment and bring clean technologies down the cost curve. That assumption is now breaking down. The grid is increasingly where the success of the transition becomes a constraint: renewable output being curtailed because the wires cannot move it, developers facing radically different economics depending on where they connect, and system operators having to run markets more cautiously because stability services have not kept pace with the generation mix.

Dan Monzani’s perspective was that grids matter, but they should not be treated in isolation from wider system design. In Britain, the connection queue has become a major blocker, but reforms to rationalise it have created their own trade-offs: a battery co-located with a solar project can be pushed behind an older standalone project simply because of how the rules are structured. His broader argument was that network build-out, market reform and locational signals are two sides of the same coin. The harder political problem may increasingly be cost allocation rather than target-setting. If the cost of new infrastructure is simply piled onto electricity bills, it can weaken the economics of electrification — precisely at the point when higher electricity demand is needed to make better use of the networks being built.

Hanns Koenig explored the same issues from a European perspective. Europe is already one of the regions where the side effects of renewables success are most visible, particularly in terms of negative prices, congestion and curtailment. Hanns drew an important distinction between these side effects. In his view, negative pricing may prove more temporary, reflecting subsidy schemes designed for an era when renewables were niche technologies and now being slowly corrected – but network constraints are harder to imagine away. In high-renewables systems, congestion and curtailment are likely to become structural features unless market design evolves. Germany is an instructive case study: network operators have been overwhelmed by connection requests from renewables and batteries, while industrial electrification is also creating new demand for access. At the same time, the old assumption of a national “copper plate” is reaching its limits, with growing pressure for stronger local incentives through price-zone reform, dynamic grid fees or other locational signals. For investors, the key message is that grid risk is no longer just a physical issue. It is becoming a market-design and political-economy issue as well.

Oliver Kerr highlighted a different manifestation of the same issue in North America. The problem is not that generation projects do not exist; in many US markets, there are large volumes of capacity sitting in interconnection queues, waiting to connect. What has changed is the pressure on the system from the demand side. Rising demand from data centres has exposed study processes and utility procedures that were designed for a world of slower load growth and far fewer projects trying to connect at once. North America’s nodal markets are better at making congestion visible than Europe’s zonal markets, but that does not remove the underlying challenge. Congestion can still destroy project economics very quickly, and investors have to manage much more granular location-specific risk. Oliver’s broader point was that the next phase of grid reform is less about inventing entirely new technologies and more about updating the rules, incentives and interconnection processes so that both new generation and large new loads can connect at the pace the system now requires.

Flexibility is key

If renewables abundance creates the problem, and grids reveal where it hits hardest, flexibility is one of the main ways the system starts to respond.

In the earlier phases of the energy transition, batteries, demand response, smart charging and software optimisation were often treated as useful but secondary: helpful for smoothing the edges, but not central to the task of building clean generation and networks.

That view is becoming much harder to sustain. As more weather-dependent generation enters the system, value increasingly shifts towards technologies and business models that can move demand, shift output, manage volatility and keep the system balanced in real time.

Hugo Batten and Ana Barillas described markets where storage is moving from value enhancement to necessity. In Australia, Hugo noted that batteries have grown from roughly 1 GW to 15 GW in a system with peak demand of around 30–35 GW, making them a central source of firming rather than a niche technology. India is moving quickly too, with a battery pipeline above 50 GW and contract structures such as FDRE (Firm Dispatchable Renewables Electricity) and deviation settlement mechanisms creating direct incentives to pair renewables with storage.

Ana made the same point from the perspective of markets she covers. In Chile, financing standalone solar has become extremely difficult because so much value is lost during peak solar generation. Spain has moved in a similar direction as solar build-out continues without sufficient growth in storage, flexibility or demand. As I explored in a recent blog, Spain sees a high level of renewables cannibalisation; but batteries and smart demand may soon come to the rescue.

Our discussion also highlighted the limits of relying on flexibility. Batteries can shift generation into higher-value hours, reduce exposure to curtailment and negative prices, improve renewable revenues and, in some cases, defer grid investment. But they cannot permanently compensate for chronic underinvestment in networks, weak market design or insufficient demand growth. Ana was particularly clear that if every solar developer responds to low capture prices by adding batteries, battery revenues themselves can cannibalise quickly. Oliver made a parallel point on flexible load in North America. Data centres may be able to help the system by turning down during the tightest 30 or 50 hours of the year, but for the other 8,700 hours they still add substantial demand. Flexibility is therefore not just a technology story. It is a market-design story about creating the right incentives for storage, demand and software to respond when and where the system needs them.

The next phase of the transition

The final part of the podcast looked ahead to 2035, which feels like the next meaningful checkpoint for the transition. By then, success will be judged less by whether the world has added another large tranche of renewable capacity, and more by whether electricity has become a materially larger part of our energy system.

That means electrification has to move from policy ambitions and targets to reality. More transport, heating, industry and digital infrastructure will need to depend on the power system, at the same time as the system becomes more weather-dependent. That is why the conversation kept coming back to the same ingredients: control costs, build and use networks better, send clearer locational signals, speed up grid connections, value flexibility properly and give investors enough confidence to build the right assets in the right places.

I asked each of my guests what had to go right by 2035 for the energy transition to be a success. They each brought unique perspectives from their regions:

• Dan’s test for GB and Ireland was cost control through better system design, because public support will not survive if the transition is seen as technically impressive but needlessly expensive.

• Hanns argued that Europe will struggle to solve congestion and curtailment without stronger locational pricing or something close to it, even if the politics remain difficult.

• Oliver’s North American version of the problem was more procedural: simpler, smarter interconnection for both generation and load, plus regulation that rewards flexibility rather than just more capital expenditure.

• Hugo put the APAC challenge most bluntly: in many markets, the limiting factor is still the speed and quality of transmission build-out. His line from an Australian client — “build transmission and then get out of my way” — captured his sentiment well. Market design matters, but there are still places where the basic physical infrastructure is the gating factor.

• Ana’s answer from Iberia and Latin America was the other side of the same point: markets and regulation have to evolve as quickly as technologies do. Investors need to know what the system needs, where it needs it and when it needs it. Otherwise capital will keep flowing into assets that may make sense individually, but do not maximise system value.

So the 2035 test is not simply whether we can build more renewables. It is a system test. Can power markets absorb much more clean generation without destroying its economics? Can grids connect and move power fast enough? Can demand become more responsive? Can regulation reward the assets that make the system work, rather than just the assets that are easiest to build? And can all of this happen at a cost that maintains political consent?

That is why the debate now feels different from when Energy Unplugged started in 2019. Back then, many conversations still circled around whether clean technologies could become cheap enough, scalable enough and mainstream enough to matter. After 300 episodes, that question has not disappeared, but it is no longer the main pre-occupation of leading thinkers, companies and governments in the space.

Renewables are deploying faster than anyone expected. The harder question now is whether the rest of the system can keep up. That means wires as well as wind farms, demand as well as supply, market rules as well as megawatts, and political consent as well as project finance.

The next phase of the transition be less about proving that clean technologies can work, and more about making the whole machine work around them. If the next decade goes well, the story in 2035 will not simply be that we built more renewable capacity. It will be that electricity became a larger, cleaner and more dependable foundation for the economy.

That relationship is breaking down in many markets as solar cannibalises itself. As solar penetration rises, this puts downward pressure on residual demand and market prices, as I explored in my recent article. The effect is that the hours of highest solar generation are increasingly becoming the hours of lowest value. In some European markets, a growing share of solar output is produced when wholesale prices are close to zero, or even negative. Meanwhile, some of the highest-value hours occur when solar generation is low or zero.

Solar can become a victim of its own success, unless there is sufficient flexibility in the market to absorb the excess.

The effect is not uniform across Europe. Britain, Germany and Spain sit at very different points on the solar saturation curve, making them useful case studies.

In this piece I look at H1 2026 power prices and solar generation across all three markets. More importantly, I look at when a solar assets actually earn revenue. The results suggest that for solar investors and asset owners, the key question is no longer simply where the best irradiation can be found, but where a market sits on the saturation curve, how quickly it is moving, and whether/when batteries and demand flexibility will come to the rescue.

How to read the charts

For each market I’ve shown two things.

First, a scatter plot of solar generation against the day-ahead price across H1 2026. This shows how solar affects prices. The solar generation is only front of meter with rooftop / behind the meter excluded.

Second, a normalised revenue-per-GW curve. This shows how the average hourly earnings of a solar asset vary through the day and across the year, indexed so that 1.0 equals the H1 average. Earnings are a function of both price and generation. The revenue curves show the feedback loop from solar depressing wholesale prices and how this effects what they earn.

Great Britain: not yet a duck

As I pointed out in the previous article, GB is an outlier in that it doesn’t yet have a strong duck curve. The more moderate solar deployment and insolation is not yet sufficient to have a strong effect on market prices.

The scatter plot shows this clearly: prices cluster in a broad band of roughly £75–150/MWh across the full range of solar output, with weak correlation. Prices do drop sometimes when solar output exceeds around 5 GW, but these are intermittent events rather than the structural midday price suppression seen elsewhere.

The revenue curve confirms this. GB solar earnings are broadly correlated with solar output – the more you generate the more you earn. This means that earnings peak in the middle of the day from 9:00-15:00, and that the sunnier months of April-June see higher revenues than January-March.

This reflects where GB sits in the saturation cycle. Solar penetration is still modest enough relative to grid size that the fleet does not consistently suppress prices. We’re starting to see early signs of price cannibalisation, with those price drops above 5 GW solar output, but this hasn’t become structural yet as it has in Germany or Spain.

For now, standalone solar in GB still works on its own merits, provided you can secure a grid connection. Colocation with batteries enhances returns by capturing spreads and ancillary revenues, but it isn’t yet the fundamental requirement it has become in Spain or Germany. Expect that to change as the GB solar fleet continues to grow.

Germany: the bimodal revenue day

The level of solar PV saturation in Germany sits somewhere between Britain and Spain.

With over 120 GWs of capacity, Germany has largest solar PV fleet in Europe. But Germany has much more extreme price outcomes than Britain or Spain, with day-ahead prices ranging from -€499/MWh to +€666/MWh in H1 2026. That spread dwarfs most other European markets.

When I look at the scatter plot a few things jump out:

1. a large cluster around the Y axis when solar generation is zero and prices can be high

2. a cluster of dots from top left to bottom right which shows the impact of increasing solar generation on reducing wholesale prices, and

3. quite a large number of sometimes extreme negative prices which only occur when solar generation is high. The most extreme negative prices happened in April and May.

The revenue-per-GW profile has a strong bimodal shape. Revenue peaks in the morning ramp up of 8:00-10:00 at around three times the daily average, while the solar fleet is still ramping up and prices are high. From there it drops through the middle of the day before reaching a second revenue bump in the evening. The revenue shapes are very different across months. January and February have a bell-shaped distribution as solar is insufficient to saturate the market. April and May saw extreme negative prices which are sufficient to drag the monthly revenue numbers negative in the middle of the day.

The practical implication of this is that for a south-facing fixed-tilt array — which generates peak output between 10:00 and 15:00 — a significant chunk of peak generation is produced in the hours of lowest revenue. This points towards upside and investment opportunities for solar trackers and East-West axis deployments. The strong within day spreads underpin an attractive business model for standalone batteries – with some of the highest IRRs across Europe.

It’s worth noting that the vast majority of operating German solar projects carry feed-in premiums or CfD-style support, so don’t always face the full impact of negative prices and wholesale market saturation. But new merchant projects, or those rolling off subsidies see this in full.

Spain: peak solar saturation?

In the previous piece I described Spain as the “undisputed champion duck” — midday prices averaging around a quarter of the H1 average, an evening peak roughly double the average, and a peak-to-trough ratio of around 7:1. In proportional terms the spread in prices is wide, though in absolute terms it is much smaller than that of Germany.

The charts show us what that price shape means for a solar asset operating in a market with this level of solar saturation and the picture is stark.

On the scatter plot we see that the highest price hours are often when solar PV generation is low or zero; whilst in most hours when solar PV generation is above 5-10 GW, the prices collapse towards zero. Spain has alot of day ahead prices just above, exactly at, or just below zero (as explored here). There are exceptions when prices and solar generation are both high but the main clusters sit near the axes.

The daily revenue graph for H1 shows that the highest revenues per hour are made in the morning and evening peaks. The hours of peak generation — 11:00 to 15:00 — are not where you make the most money. Going to monthly revenue profiles adds further nuance: in January solar PV makes some healthy returns despite much lower production with a bell-shaped curve across the day. Prices were exceptionally low across February due to a combination of high wind and hydro. April and May show the most extreme duck curve with close to zero revenue in the sunniest hours but healthy returns in the morning and evening peaks.

This all has implications for site configuration and future investment. Solar trackers — panels that follow the sun across the sky — are more common in Iberia than elsewhere in Europe. In the early years, trackers could capture higher shoulder prices. But when many plants in the market are trackers, they all compete against each other in the same windows.

The price shape also opens up the opportunity, if not imperative, for investment in batteries. Spain currently has very limited battery storage capacity (<100MWs) relative to the scale of its midday solar surplus. The current market outcomes create the incentive for standalone batteries and colocation (both new build and retrofit). Additionally as EV deployment scales up, then smart charging EVs should also soak up the solar excess (though as I explored in this article, grid charging disincentives the shift in consumption towards the sunniest hours). Together these factors should rebalance the market, absorb the midday solar excess and compress the trough, and dampen the evening spike as batteries discharge into the market. The outlook for solar PV very much depends on how quickly the battery storage and EVs arrive. Whilst Spain currently sees very high levels of cannibalisation, I would not see this as a permanent equilibrium.

The solar saturation spectrum

These three markets illustrate the trajectory that every European solar market is following, just at different speeds.

GB is broadly where Germany was five or six years ago — solar earns its keep, colocation is a bonus. Germany is deep in the transition — configuration and colocation matter enormously, and merchant standalone solar faces a real cannibalisation challenge. Spain is highly saturated with solar — the midday hours in sunnier months have stopped generating meaningful revenue. The market should rebalance as more batteries and smart charging EVs enter the market.

So the central question for solar investors in Europe is no longer just “where is the irradiance?”, it is “where is this market in the saturation curve — and how fast is it moving?”

Data: Aurora Energy Research EOS API and REE ESIOS API, H1 2026 historical generation and day-ahead prices. Revenue analysis normalised to H1 average hourly earnings per GW of installed capacity.

There is a lot of technical detail behind these debates, but ultimately the media headlines pointed to Net Zero and decarbonisation as the culprits. The argument made is roughly that renewables are making the grid harder to run, so Net Zero is failing.

It is easy to see how the headlines end up there, but the conclusion is too simplistic.

To be very clear - renewables are definitely part of the story. A power system dominated by wind, solar, batteries and distributed energy resources behaves differently from one built around coal, gas and nuclear power stations. Renewable penetration changes power flows, the shape of generation, and makes forecasting more difficult. It changes how operators must manage the grid to ensure stability.

We absolutely need to acknowledge and tackle these operational challenges, but this is not the same as concluding that Net Zero is to blame, or that decarbonisation is failing overall.

In many ways, the current debate points to a deeper shift in the energy transition. For the last twenty years or so, the central challenge has been how to generate more clean electricity. Government interventions and subsidies played a major role in getting wind and solar built. But solar PV, batteries and EVs are no longer government-backed technologies waiting for support – falling costs mean that they are economically attractive in their own right. In other words, this is not just a Net Zero policy project any more, it is a market reality.

The deployment challenge has not gone away, but the real challenge now is not simply how to deploy renewables - it is how to coordinate power markets and systems around millions of low carbon devices.

Reliability and decarbonisation are not mutually exclusive

The debate in the media often treats reliability and decarbonisation as if they are fundamentally in conflict or mutually exclusive.

If that were true, we would expect the cleanest power systems to be the least reliable, and the highest carbon systems to be the most reliable - but this is not the case.

Some of Europe’s most reliable electricity systems are also those with high deployment of renewables and/or low carbon emissions – for examples Finland, Switzerland, Germany and France. All four countries rank highly in indices relating to reliability of electricity supplies. The technology mix varies across the examples: Finland, France and Switzerland all have a high proportion of generation from nuclear and renewables; whilst Germany has a very large fleet of solar and wind but no nuclear. All of these examples are strongly interconnected with neighbouring markets.

On the other hand, there are examples of grid stability issues in systems which still have high proportions of thermal, fossil fuel generation - such as ERCOT, SPP and PJM in the US. Neither set of examples categorically confirms or disproves the relationship between decarbonisation and reliability, but illustrates that there are a range of outcomes on both metrics.

Academic research points in the same direction with more nuance. Studies of high renewables systems consistently find that variable renewables create real engineering challenges including lower inertia, greater forecasting uncertainty, steeper ramping requirements and a greater need for flexibility, reserves and balancing services. But the literature does not support the claim that higher renewables penetration or lower emissions automatically means more blackouts. The conclusion is that systems with high renewables need different operational tools, market arrangements and flexibility resources if they are to remain reliable.

That distinction is crucial. Operational challenges do not prove that Net Zero is fundamentally flawed. They do show that power markets and systems must evolve as they decarbonise in order to remain reliable.

So the question for Britain is not whether Net Zero and reliability are compatible - the evidence from other countries suggests they can be. The real question is whether Britain is putting in place the data visibility, flexible resources, market arrangements and control systems needed to make a low-carbon system work well.

The Iberian blackout: a cautionary tale

Whilst I give examples above where low carbon and reliability go hand in hand, this is not to say that this is always the case.

The Iberian blackout in April 2025 was a powerful reminder of what can happen when grids fail: tens of millions of households were left without power for around 18 hours.

I explored the causes of the Iberian blackout in a recent episode of Energy Unplugged with my colleague Christina Rentell.

When the blackout happened, the first reaction in many quarters was to point the finger at renewables. That reaction was understandable, as renewable penetration was high in the moments leading up to the event. But that is not the same as saying renewables caused the blackout - the reality was more nuanced. High renewable penetration was a necessary condition for the operating challenge the system faced, but it was not a sufficient explanation for the collapse that followed, as similarly high renewables penetration was seen on other days in Iberia and happens routinely in other grids.

The real issue was that the system did not have the flexibility or capability to manage the high renewables penetration. The system experienced oscillations in voltage and frequency which the system operator did not adequately address. The oscillations caused generation plants and equipment to trip and disconnect from the system, and this quickly led to a cascading failure.

The key lesson from the incident was that a power system with very high renewable penetration has to be planned, operated and regulated in a different way. It needs enough flexibility to balance the system in real time against variable renewables output; explicit stability services for frequency and voltage control; and the right regulations and operational procedures to manage the physical characteristics of the system it has become.

Since the blackout, the Iberian system has been operated more cautiously. The grid operator has procured more balancing reserves, keeping gas on the system to provide inertia. The event has accelerated the debate about the need for more system flexibility. Grid-scale battery capacity was very limited at the time of the blackout (less than 100 MW), but it is now expected to build out rapidly over the next few years.

The Iberia example illustrates what can go wrong if system operation and market design do not keep up with the changing generation mix. GB had already learnt some of these lessons on the back of the blackout in 2019 which affected around 1 million people, in which there was also a cascading failure as a result of insufficient flexibility and operational limits for plant being set too tightly. Since then, the system operator has overhauled frequency control markets and we have seen a surge in battery capacity in GB.

This does not mean Britain is now immune from grid reliability challenges, but there are practical lessons here about how the approach to system operation needs to adapt for a high-renewables mix.

The Hitzeflaute: a new flexibility challenge in Britain

The recent debate on grid stability in Britain kicked off with a stability issue which occurred on the 23^rd June during the summer heatwave.

I wrote an article at the time describing the phenomenon as a “Hitzeflaute” — a German word literally meaning a “heat lull”. Many people in energy circles are familiar with the term “Dunkelflaute” - the cold, still and dark periods that create stress in winter power systems. The Hitzeflaute is its summer counterpart.

The Hitzeflaute creates two distinct system challenges on hot, sunny days. The first is that as solar generation ramps up in the morning and peaks around noon, other generators are pushed out of the system, leading to a high instantaneous renewables penetration and potentially issues of low inertia. In GB this issue is limited simply by the amount of solar we have on the system – it is not sufficient to meet total demand as it is in some other markets such as Iberia (as illustrated in my recent substack on Duck curves). The second issue is potentially more problematic: in the evening demand remains high, exacerbated by air conditioning demand on hot days, but solar output falls away rapidly.

NESO issued an Electricity Margin Notice for the evening of the 23^rd June, which was later cancelled. At first glance it seemed a bit unusual to have a period of system stress during the summer – but it is a new reality which many system operators are having to manage as solar capacity continues to surge, and air conditioning demand increases as the climate is warming.

The issue was not a shortage of generation, as daytime solar output was strong. The problem was whether the system had enough flexibility to manage the changes in generation across the day, and whether the right resources were available to respond quickly enough at the right moment. In other words, it is about flexibility.

The jump from there to “Net Zero is making the grid unsafe” was quite predictable, but analytically too simplistic.

The NESO debate is really about visibility

Since the heatwave incident, there has been increasing scrutiny in the press around how the incident was handled, including reports from whistleblowers. Then in the last few days the media cited a leaked report on the difficulties that NESO is facing in managing the grid which was written before the heatwave incident.

I have not seen the report itself, but the key points reported in the media seem to be:

• concerns about visibility in a system increasingly made up of smaller and more distributed assets

• impact of this on the ability of NESO to produce accurate forecasts

• over-reliance on manual grid management by engineers not automated processes

These are reasonable concerns but need to be interpreted correctly.

The challenge is not simply that Britain has more renewable generation or is pursuing Net Zero. It is that Britain increasingly has millions of assets rather than the hundreds of assets which used to make up the grid. The GB power system was originally designed around a relatively small number of large thermal generators – coal, nuclear and gas. Coal is gone from the system, as is most of the nuclear. The system increasingly includes solar panels, batteries, EVs, heat pumps and smart devices spread through homes and businesses across the country. That creates some very different operational problems.

This shift matters because much of the growth at the edge of the system is no longer waiting for a new round of subsidy or direct government intervention. Falling technology costs mean that rooftop solar, batteries and EVs are increasingly adopted because they make economic sense for households and businesses. EVs have hit upfront cost parity whilst running costs are way below ICE vehicles. Rooftop solar is already economic for households and business, and about to become MUCH cheaper once ‘plug-in’ solar panels will be allowed to be sold in the UK from next month (for a deep dive on solar and grid integration, check out my podcast with Sonia Dunlop from the Global Solar Council). Last year we saw nearly 200,000 solar installations in the UK versus 46,000 in 2021. EVs went from 1% of new car sales in 2018 to 25% last year.

So this is no longer about Net Zero per se – it is simply about consumer choice. The question is not whether these assets will appear on the system, but when they will appear and whether the system can see them and coordinate them.

This challenge already sits at the heart of the current debate on GB market design. In my recent article on GB market design, I noted that NESO’s proposed reforms to balancing and dispatch are increasingly focused on improving visibility of distributed resources and reducing operational uncertainty. NESO is consulting on reforms intended to provide greater visibility and transparency of smaller assets, potentially down towards the 1 MW level. One of NESO’s central concerns is that growing volumes of batteries, embedded generation and flexible demand sit outside the visibility and control frameworks designed for an earlier generation of power systems.

The debate is often presented in the media as an argument about renewable energy or Net Zero. If the diagnosis was simply that renewables are the problem, the proposed solution would be fewer renewables. Instead, the proposed solution is more visibility and improved coordination. But is visibility down to megawatt scale enough?

Recently I recorded an episode of Energy Unplugged with Devrim Celal, Chief Flexibility Officer at Kraken. We talked about a system in which we can harness vast amounts of information at the edge of the grid, about distributed rooftop solar, household batteries, EV charging and flexible demand at a level of granularity that would have been unimaginable a decade ago. This is not something that might happen in the future – it is the type of data that Kraken already handles today on its platform. Whilst NESO is consulting on reforms that would improve visibility of assets down to the megawatt scale, Kraken already has operational data at the level of individual homes and devices.

So the problem is not a lack of data in general – it is about who has access to which data, and what they do with it.

At the moment, data on power generation, consumption and flows is distributed across generators, retailers, software platforms such as Kraken, aggregators, distribution networks, and the transmission system operator NESO. Due to privacy concerns, the household-level data on behind the meter assets and smart charging is only available to retailers / software platforms, not to grid operators.

In the report leaked to the media, the proposed solutions are that NESO should have more access to data on distributed and behind the meter installations, and that it should use this to improve its forecasting ability.

It is hard to argue with these points in general. But I would make a few observations about the direction of travel and policy choices:

• Firstly, the idea of ‘demand forecasting’ needs to reflect the new market reality. System operators used to think of demand as being relatively fixed in the short term and organise the supply side to meet this demand. The new reality is that an increasing proportion of demand responds to prices in real time – EVs, grid scale and behind the meter batteries, heat pumps. The notion that NESO can ‘forecast demand’ only works if it reflects this dynamism and the linkage to prices and markets.

• Secondly, it is one thing to get access to data, and another to act on it. NESO already has bids and offers from grid scale batteries, but they are systematically under-utilised in the Balancing Mechanism as these assets are often ‘skipped’ in favour of traditional technologies. This issue has been recognised for some time but progress to act on it has been slow. Here we’re talking about a hundred or so assets in the tens-hundreds of MW per asset… the data challenges would go up several orders of magnitude if we want to utilise data on millions of assets down to household scale.

• Thirdly, there is a real policy choice about whether we want the system to be more centrally coordinated or market orientated – questions I explored in my recent article on market design. A centrally coordinated system means granting the system operator ever more visibility and control. The alternative is to open up liquid and transparent markets for energy and system services and enable market participants to reveal information through their bids and offers. This is ultimately a policy choice about who owns and manages the risks, and their ability to do so.

Linking the threads together – wouldn’t we expect markets to solve demand and supply more cleanly than an administrative process?

The next phase of the energy transition

For much of the last twenty years, the central question of the energy transition was how to get more clean power built. That was a world of targets, support schemes, contracts, subsidies and government-backed deployment. Those interventions helped take wind, solar, batteries and EVs from expensive alternatives to mainstream technologies.

We are now entering a different phase. Low carbon technologies are increasingly the most economic choices to make for households and businesses. Rooftop solar, batteries and EVs are spreading across the system not because policy says they should, but because costs have fallen and the commercial case is becoming stronger.

That is good news for decarbonisation, but it changes the nature of the problem. The harder challenge is no longer how to deploy clean technologies – it is how to integrate them into a power system that was not designed around millions of small, flexible, price-responsive assets.

It is not an easy set of challenges to solve. The Iberian blackout highlighted what can happen when system operation does not keep pace with a changing generation mix. The heatwave in Britain highlighted a new flexibility challenge created by hot, sunny days when solar output rises and then falls away quickly. The NESO debate highlights the challenge created by millions of smaller assets sitting at the edge of the system.

Part of the solution is to improve visibility and forecasting. But we should also look at the broader questions of how to design markets that reveal the value of flexibility, and how to create operational tools to manage a system comprised of millions of distributed assets. If we are still reliant on manual processes in the control room, then it is right to call out the risks of the current approach.

But this is where the current debate risks focusing on the wrong questions. The issue is not whether Net Zero inevitably makes the grid unreliable - international experience shows that low-carbon power systems can also be reliable. The issue is whether we are adapting the rules, markets, data flows and operating practices quickly enough for the system we are building.

That is a harder, messier and more technical challenge than simply subsidising more clean power onto the system. But it is also the challenge that will determine whether the next phase of the energy transition succeeds.

The “duck curve” is a term borrowed from the California grid operator CAISO, which first plotted the ‘net load’ or residual demand curve (total demand minus solar output) and noticed it looked like a duck in profile — a long belly in the middle of the day, and a steep neck rising in the evening.

The equivalent shape also shows up in day-ahead wholesale power prices. When the sun is shining and solar generation is high, wholesale prices are suppressed — sometimes significantly. When solar output drops in the late afternoon and evening, demand peaks and prices spike. The bigger and faster the solar build-out, the more exaggerated this shape becomes.

This matters for anyone with a financial exposure to electricity prices: owners and developers of solar PV assets, batteries, industrial consumers, retailers, and investors. A solar farm with peak output in the middle of the day is increasingly squeezed out by other solar generators. A battery that buys cheap around lunchtime and sells in the evening is exploiting exactly this shape and spread.

So which European markets currently see the strongest duck-shaped power prices?

Notes on methodology: I pulled out day-ahead prices across H1 2026 from Aurora’s EOS API for the largest eight European power markets. For each market I calculated the average price by hour of day across all days in the period (amalgamating 15 or 30 minute blocks), then indexed each hourly value against the H1 average — giving a plot that captures the shape, independent of price level. I’ve also shown absolute prices below. I’ve adjusted for daylight saving time, and converted the GB and Poland prices to Euros.

Two caveats are worth keeping in mind. First, this is an average across six months, so it smooths extreme short-lived events and changes between individual months. Second, while solar is a key driver of price shape, it is not the only one: demand patterns, interconnection, fuel prices and market structure all play a role.

Spain: the undisputed champion duck

Spain exhibits by far the most pronounced duck-shaped price profile.

On a typical day, midday prices fall to around a quarter of the H1 average before rising to roughly double that level in the evening — a peak-to-trough ratio of more than 7:1. No other European market comes close.

Several factors combine to produce this profile. Spain has one of the highest installed solar capacities in Europe relative to its market size. Insolation levels are far higher than northern Europe — more sunshine means more generation. Solar trackers (panels that follow the sun across the sky) are far more prevalent in Iberia than elsewhere, which widens and flattens the solar generation curve rather than producing a sharper spike around noon. The result is a longer, deeper price low from 10am to 5pm (longer still if we were to isolate Spring/Summer months).

The late evening price peak (around 9pm) also reflects unique factors influencing the demand pattern in Iberia. The hottest part of a Spanish summer day arrives well into the early evening, shifting cooling demand later. Cultural habits amplify this further — no-one goes for dinner before 9pm; streets come alive after dark. And Spain is on CEST even though it is roughly due south of the UK, meaning that as the sun moves over through the day the ‘local time’ is 1 hour later.

France: the duck is getting fatter

France shows a similar but somewhat less extreme profile — midday prices averaging around 30% of the average, with an evening peak of around 160-170%. Solar penetration is growing rapidly, and the shape is becoming more pronounced year on year. The morning peak (around 7–8am) is a French-market feature worth noting — partly driven by heating demand in winter months included in the H1 average.

Germany, Netherlands and Poland: three ducklings

Germany and the Netherlands track each other almost exactly due to their tight market coupling. Both show a moderate duck shape: midday at roughly 40% of average, evening peak around 150-160%. Poland shows a recognisable duck profile (midday at ~50%, evening at ~150%) but with less depth than its western neighbours — reflecting lower solar penetration and a grid still dominated by coal.

GB: not yet a duck

GB’s midday trough is shallower still, and the evening peak more modest. GB has proportionately less solar capacity, combined with lower insolation - limiting solar output and the duck-curve effect. Though solar penetration is ramping up, so expect this pattern to strengthen.

Norway: not a place for (price curve) ducks

Norway stands apart. Prices move within a relatively narrow band across the day, with little sign of a duck curve. This is unsurprising in a system dominated by flexible hydro generation and with limited solar capacity. Without large volumes of midday solar, there is no persistent structural pressure pushing prices down during daylight hours.

In effect, Norway gives a useful counterfactual: this is what daily price shapes look like in the absence of significant solar penetration.

Italy: the duck lives in the South

Italy is the most surprising result. With its southern location and growing solar capacity, you’d expect a pronounced duck-shape to its power prices. At the national level it doesn’t appear — prices move in a relatively compressed band. The explanation lies in its market structure: Italy’s national price is a weighted average across eight regional zones, and the northern industrial zones (which dominate the National price) have less solar and more baseload than the south.

Drill into the prices in Sicily, Sardinia and Southern Italy and a clearer duck shape begins to emerge — the prices around midday fall further than the national average. As solar continues to build in Italy, the national price shape will increasingly reflect it.

Shape versus level: a different perspective

Looking at price shape alone only tells part of the story.

Italy, despite its relatively flat shape, is by far the most expensive wholesale market in absolute terms — averaging €127/MWh across H1, with the midday trough still at €88/MWh. Spain, the most extreme duck shape in relative terms, is the cheapest — averaging just €50/MWh, with midday prices averaging in the low teens and falling to just above/below zero in many sunny periods. Germany, Poland and the Netherlands have the largest peak to trough spread in absolute terms.

What this means for solar and battery projects

The price duck curves matter enormously for asset economics, especially for solar and batteries.

A solar PV plant in Spain produces its maximum output in periods where the price is often low, zero or slightly negative. It earns most of its money in the morning and evening shoulders, when solar output is ramping up/down, demand is high, and prices spike up. Installing trackers gave some upside for those who got in early, but even this opportunity is cannibalising itself. Now the industry looks towards colocation of solar + BESS.

Batteries make money by trading the arbitrage between low and high price periods, so what matters is the size of the spread, the duration of the extreme high/low prices, and whether the daily shape lends itself to a 1 or 2 cycle strategy (plus a host of balancing and ancillary market dynamics).

The few batteries in Spain operate in a market with generally low prices. Spreads are high in relative terms but moderate in absolute terms. Batteries in Italy operate in a high-priced market but with compressed spreads. Of the markets analysed above, the biggest absolute price spreads are to be found in Germany, Poland and the Netherlands.

There is plenty to unpack here so it deserves a whole article to itself… I will return to it in the coming days!

In the meantime - I hope you’ve found this article interesting. If you’d like the data for another market not yet shown, then let me know and I can pull it out (add a comment to tell me which market and share the post).

P.S. cover photo is courtesy of Aurora’s modelling team who have amassed an impressive collection of rubber ducks…

The political dividing lines are pretty clear: the Conservatives and Reform tend to argue for continued North Sea production, whilst the Greens and Lib Dems oppose it outright. Labour seems the most conflicted. It wants to maintain credibility on climate, and committed in its manifesto to a moratorium on new North Sea licensing. But it also has deep political and economic reasons to care about jobs, industrial resilience and the cost of living.

The debate burst back into life this weekend. First came reports that incoming prime minister Andy Burnham was considering a more pragmatic stance on the North Sea, allowing production from already-licensed but contested fields such as Rosebank and Jackdaw, and potentially loosening the policy around tiebacks (incremental drilling adjacent to existing fields). Then today came the clarification from Deputy Leader, Lucy Powell that there has been no formal policy change, just a “change of emphasis”.

In practice, that leaves the issue exactly where it so often ends up — politically live and wide open to being caricatured by both sides.

The debate on North Sea oil and gas has been lost at sea

The deeper problem is that much of the debate for or against North Sea oil and gas production is still made on weak grounds. We hear claims about lower bills, greater price stability, or the inconsistency of importing gas when renewables are cheaper. These points are politically very salient, but they are often overstated or simply wrong. Worse still, they are easy to knock down and so distract from the actual debates we should be having. Often there is also a confusion between supply side policies such as licensing, and demand side policies such as electrification - they are not the same thing.

In my view there is a serious case to be considered for a limited continuation of North Sea oil and gas production, it just isn’t the one that is usually made.

The case I would make is a simple one: even on ambitious decarbonisation pathways, the UK will continue to use oil and gas into the 2030s and 2040s. If that is true, the policy question is not whether we approve of fossil fuels in general or not. It is how the remaining demand is met — and with what economic and environmental consequences.

Clearing away the misleading claims

Let’s start by clearing away some of the misleading claims which are often made but are on shaky ground or tangential to the debate.

Lower prices

A common argument in favour of North Sea oil and gas production is that it will lower UK energy prices. It is an attractive claim because it speaks directly to the cost of living crisis that we have lived through in the past few years.

But under current market arrangements, it is largely false to think that opening up the North Sea will do much to prices. Oil is priced globally. UK production does not meaningfully determine what UK drivers pay at the petrol pump. Gas has historically been more regional, but it is now increasingly shaped by interconnected European and global markets, especially as LNG has become more important. Additional domestic production can change where the UK sources its fuel from, but it does not change the price that consumers pay.

There is an important caveat to make here: the government could choose to contract for domestic gas at fixed prices, exactly as it does for renewable electricity through Contracts for Difference. This could be a condition of licensing and is exactly how oil and gas licensing is done in some parts of the world – we are just not that familiar with this model in a UK or European context. In that setup, domestic production could be used to stabilise prices, but this would be an active choice to reallocate risk to the state. It is not an automatic benefit of having gas under the North Sea, and it is not how the UK gas market works today. If we pursued this idea we would have to accept that there would be high price periods when such a contract could look attractive, but also low price periods where it is locking us into higher prices.

Either way, the case for continued production should not lean on implied bill savings that the current market design does not deliver.

Lower volatility

A slightly subtler version of the price argument is that domestic production may not lower prices, but it at least makes them less volatile. This also sounds plausible but under current market design it does not really work either.

Domestic oil and gas is sold into wholesale markets at prevailing prices. UK consumers remain exposed to global price shocks, regardless of where the physical molecules originate. Geography alone does not insulate households or businesses from volatility. Again, we could intervene more strongly in the market to fix prices in order to reduce volatility, but this is not a decision about licensing.

“Renewables are cheaper, so why gas?”

An argument often made by opponents of continued or new oil and gas production, is that gas makes little sense when renewables are cheaper.

This is again persuasive, but it is not an apples for apples comparison.

Claims about cheap renewables are usually made on the basis of the cost of electricity generation per megawatt hour. On this basis renewables win hands down – wind and solar are cheaper than a gas power station at producing each MWh of power. But wind and solar do not provide every service the system needs. Even in a highly decarbonised power system, there is still a need for firm and flexible capacity to manage periods when wind and solar output is variable or low, including multi-day or seasonal shortfalls. The share of GB power generation coming from gas has already shrunk from over 40% in 2016 to around 25% in 2025. But gas is still required in many hours of the year to fulfil demand; and to provide inertia and reactive power.

The role of gas will shrink as storage, interconnection and demand-side flexibility expand. But it has not disappeared yet and will not disappear completely that quickly. Decarbonisation pathways, including those from the Climate Change Committee discussed below, still envisage some role for gas-based generation — increasingly with carbon capture, or potentially replaced by hydrogen — as part of a balanced system.

And electricity generation is only part of the story. Gas is still used extensively in heating and in parts of industry where renewables are not a direct substitute. Replacing that gas depends on electrification, new infrastructure, changes to buildings, and millions of consumer and business decisions. That transition is happening, and I would love it to happen quicker, but with the best will in the world, gas will still play a role for some time to come.

So “renewables are cheaper than gas” is too simplistic as a statement. Without specifying which service is being provided and in which sector, it glosses over a lot of complexity.

It does not answer the real question: how does the UK manage the residual role of gas as the energy system transitions?

Resetting the debate: the UK will continue to use oil and gas for decades, even in Net Zero scenarios

Once the weaker arguments are set aside, the debate comes into focus. A serious discussion about North Sea oil and gas has to start with a simple premise: the UK will continue to use oil and gas for some time.

This is not a rejection of Net Zero as an end game. In fact, it is embedded in the UK’s own official analysis of what the transition to Net Zero looks like.

The Climate Change Committee’s pathways and the government’s energy demand projections show fossil fuel use falling, but not vanishing in the near or medium term. Specifically:

CCC’s 7^th carbon budget (published February 2025)

• This shows gas demand halving from ~720 TWh in 2025 to ~360 TWh in 2035, reaching ~200 TWh by 2050 with a residual role mainly in the power sector.

• Oil demand also roughly halves from ~750 TWh in 2025 to ~375 TWh in 2035 with a more sustained decline to ~120 TWh by 2050, with the remainder mainly focused in aviation

The Department for Energy Security and Net Zero, Energy and Emissions projections (published February 2026, Annex E) – the ‘reference scenario’ is a representation of current and planned Government policies, and as such is not aligned to the 7^th carbon budget. It contains much more bullish figures for oil and gas use.

• In this projection, primary oil demand decreases by 42% from 2025 to 2050

• Primary demand for gas increases by 21% from 2025 to 2050 (I had to triple check this myself as it seems like a surprising trajectory, yet it is DESNZ’s latest official projection)

The gap between the CCC pathway and DESNZ’s reference scenario is a clear representation of the ‘policy gap’ to the carbon budget. The CCC shows the trajectory required to stay aligned with the UK’s carbon budgets: rapid falls in oil and gas demand, driven by faster electrification, efficiency improvements and fuel switching. Even then the CCC forecasts a residual demand for oil in transport and parts of industry in the 2030s and in aviation into the 2040s; whilst gas remains widespread in heating into the 2030s and in the power sector into the 2040s. DESNZ’s reference case, by contrast, reflects the world implied by current and planned policies. On that basis, oil and gas demand falls more slowly — and in the case of gas, may not fall at all.

This matters enormously for the North Sea debate. It shows that central assumption of UK climate policy is not immediate elimination of oil and gas, it is a managed decline. it also shows that the residual role of oil and gas demand is highly sensitive to government policies and to delivery.

If the UK wants the lower-fossil-fuel future implied by the CCC pathway, it has to close the delivery gap on heat, industry, transport and firm low-carbon power. If it does not, the country will need more oil and gas, for longer. But even in the CCC pathway we will need oil and gas into the 2040s.

None of this is an argument against ambition. But we should be clear about the scenario we are planning for and delivering rather than assume that the Net Zero scenario has already been delivered.

Overall, the relevant question is not whether the UK should use oil and gas at all - we will do even in the most ambitious decarbonisation scenarios. The key question is how the remaining demand is met during the transition.

The real choice: domestic production or imports?

Once the residual demand for oil and gas is acknowledged, the policy choice becomes clear. The UK is not choosing between producing oil and gas and eliminating its use. It is choosing between meeting remaining demand through domestic production or through imports.

The public debate often rests on the assumption that supply decisions and demand outcomes move together, but in reality they do not. The UK can choose to stop producing oil and gas domestically, but this does not make oil and gas consumption disappear - it changes the source of supply. If the UK continues to consume oil and gas during the transition — as its own planning assumptions suggest — then restricting domestic production increases reliance on imports.

The choice between domestic and imports has consequences. Imports mean more value flowing overseas, fewer jobs, and lower tax receipts. None of this is an argument for maximising production at any cost. It is simply to say that domestic production and imports are not equivalent.

Balance of payments and economic leakage

Put simply, domestic production reduces the need for imports, and imports represent an outflow of value to other countries. For a country that still uses large volumes of oil and gas, replacing domestic supply with imported supply means sending more money overseas for essentially the same energy service. That worsens the trade balance relative to the counterfactual in which more of that value is created domestically. All else equal, a larger import bill adds pressure to the current account and makes Sterling less attractive than it would otherwise be.

Jobs and skills

The same logic applies to employment and skills. North Sea oil and gas production supports large numbers of high‑skilled jobs across a complex supply chain, many of them concentrated in specific regions. As production declines, employment in the sector will fall, and so too will associated tax receipts. Forgoing domestic production does not eliminate demand; it changes where the economic activity associated with that demand is captured.

Skilled workers trained in offshore operations, engineering, subsea services and project delivery will not automatically move into new sectors. If domestic oil and gas production falls away too quickly, then this can hit local economies hard. A more orderly decline gives the UK a better chance of carrying those skills across to other industries. Many are relevant to offshore wind, carbon capture and other infrastructure — but that transfer of skills and expertise needs time and effort.

Fiscal value

Domestic oil and gas activity also generates tax receipts – a way of the UK capturing some of the value from allowing firms to extract these resources. These tax receipts fluctuate with prices and production levels, so cannot be guaranteed year to year, but they are real and matter during a period of significant fiscal pressure.

This is not an argument about maximising extraction to fund public services. It is about capturing value from activity that will otherwise occur abroad if demand persists and domestic supply declines. Choosing imports over domestic production means giving up tax receipts.

Environmental footprint of domestic versus imported gas

Another factor relevant to the debate around domestic gas is its environmental footprint versus imports. Gas is a fossil fuel, and its combustion produces emissions (around 320kgCO_2e per barrel of oil equivalent). That remains true regardless of where it is produced.

But if we are considering domestically produced versus imported gas then we need to look at the full lifecycle emissions not just combustion emissions. UK gas production typically has a lower lifecycle emissions footprint than imported Liquefied Natural Gas (LNG), which is the marginal supply of gas into the UK and Europe. The emissions associated with production, processing and transportation are 28 kgCO_2e/boe for domestically produced gas, and 85 kgCO_2e/boe for LNG imports. Taken together this means that the full lifecycle emissions of imported LNG are around 16% higher than that of domestically produced gas (e.g. 320+85 compared to 320+28).

This does not mean that domestic gas is ‘clean’, nor does it justify open‑ended production. It simply recognises that if gas is being used anyway, it is preferable to source from the North Sea than from LNG, from a global emissions perspective.

Conclusion: towards a more rational North Sea debate

The case for continued North Sea production should not be made on the basis that it will cut bills, tame volatility or make the UK energy independent. Those claims are too easy to overstate, and too easy to knock down.

The debate needs to be more rational. The UK will continue to use oil and gas for some time, even on an ambitious decarbonisation pathway. If that demand exists, the source of supply matters. Domestic production is not the same as imports once the economic value, jobs and skills, fiscal revenues and environmental footprint are taken into account.

That does not mean maximising extraction or giving up on the energy transition. It means managing the decline properly: with clear limits, strong regulation, and a serious plan to reduce demand through electrification, efficiency and low-carbon alternatives.

In short, a more rational North Sea policy would stop pretending that policy on licensing will deliver demand reduction and eliminate oil and gas use. The UK should move away from oil and gas as quickly as it credibly can. But while demand remains, the source of supply matters for economic and environmental reasons.

Managing that reality is not a retreat from Net Zero, it is part of making the transition work.

We deploy solar panels, wind turbines, batteries, EVs and grid transformers. The discussion is framed in gigawatts, terawatt-hours, capacity factors and power prices.

And yet, a dimension that is often overlooked is the shift from molecules of oil and gas towards tonnes of metals and materials.

Copper, aluminium, nickel, lithium, steel. The physical materials that have to be mined, processed, manufactured and transported before a single electron can flow through a clean energy system.

This was the theme of a recent episode of Energy Unplugged, where I was joined by Georgie Hallett, Chief Sustainability Officer and Head of Physical Markets at the London Metal Exchange (LME).

A couple of headline statistics from our discussion really stood out for me:

• In 2025, around US$21 trillion worth of metals were traded through the London Metal Exchange. This is larger than the annual GDP of China! Yet many in the energy sector probably have not heard of the LME.

• The trades amount to four billion tonnes of notional volume - equivalent to roughly 500kg for every person on the planet (note: the metals are re-traded so this is way more than the total physical volume)

The statistics matter not because of the LME itself - there are other trading platforms out there. They hint at the sheer scale of the materials production system that underpins the global economy. Metals markets support everything from buildings and bridges to consumer goods, manufacturing and transport infrastructure.

The energy sector is still only a relatively small share of that demand (hard to say, but in millions not billions of tonnes). But materials demand is rising as the energy transition implies not only a shift from molecules to electrons, but also to more CAPEX and materials intensive forms of energy production.

The scale of the materials challenge for the energy transition

Let me give you a few examples of the physical reality of the energy transition in terms of metals and materials:

• A gigawatt of solar PV requires roughly 2,800 tonnes of copper and around 4,000 tonnes of silicon, alongside large quantities of aluminium, steel and glass.

• A gigawatt of onshore wind requires roughly 3,000 tonnes of copper and around 120,000 tonnes of steel. Offshore wind requires even more copper, steel and concrete.

• A 1 GW / 4 GWh battery system contains thousands of tonnes of copper and aluminium, thousands of tonnes of lithium compounds, and vast quantities of graphite and other materials.

• An Electric Vehicle contains up to 4 times as much copper as a conventional ICE vehicle.

The transition is therefore not simply a challenge of generating clean energy. It is also a challenge of sourcing large quantities of materials. Are these secure? Are they sustainable?

The new energy security

For the last few decades, energy security meant access to molecules – oil, gas and coal. The big geopolitical questions of the energy sector revolved around resource-rich nations, shipping routes and fuel import dependence.

The energy transition has not eliminated those questions – as we’ve seen through the energy crisis and Iran war. But it adds a new layer.

Increasingly, governments are concerned about minerals security and publishing strategies around this. Industrial policy is becoming entwined with resource security. Supply chains are being scrutinised not just for cost, but for resilience and geopolitical risk. For example, five of the metals traded on the LME appear on the UK’s Critical Minerals List.

Can we secure enough of the required materials for the energy transition (and more broadly)? Can production keep pace with demand in particular materials like copper and lithium? How concentrated are supply chains?

Most importantly - are we replacing one set of strategic dependencies with another?

In the podcast conversation, Georgie seemed less worried about absolute resource availability. Her concern seemed to be less whether the industry can ultimately scale supply, and more how that supply is delivered: who produces it, under what standards, and with what trade-offs between sustainability, resilience and cost.

Can we have more mining and less environmental impact?

This was a central question and tension that kept coming up throughout our discussion. The energy transition requires more metals and materials. At the same time, society is demanding higher environmental and social standards.

Can we have both at the same time?

The mining and metals industry has historically struggled with public perception around environmental and social standards. Yet it is also difficult to imagine a world of electrification, renewable energy and battery storage without an expansion of mining activity and metals production.

One of the biggest challenges has been encouraging the sector to be more transparent about both its shortcomings and its progress. For decades, much of the metals industry operated with limited visibility around provenance, environmental performance or social impacts.

That is beginning to change, and Georgie explained the LME’s role in shaping this change. The LME marketplace has defined a set of minimum standards which all metals products muct adhere to in order to be listed on the exchange – some products were dropped due to failing to meet these standards. The LME also created a ‘passport’ system which also facilitates disclosure and traceability.

Transparency does not solve every problem, and I am sure that many stakeholders would advocate for the standards to be tighter. But it does at least allow markets to make informed choices – making sustainability something that can be measured, challenged and priced.

I brought into the discussion the parallel between sustainability measurement and market based mechanisms in metals and the power sector. In power there are numerous green certificate schemes around the world, either created through government regulation or voluntary schemes.

Will anyone pay for sustainable metals?

This led me to a question to Georgie: If two tonnes of aluminium look identical, but one was produced with lower emissions and stronger social standards, should it be worth more?

The metals industry is increasingly grappling with the same challenge that electricity markets have faced for years about how to identify green electrons and attach a green premium.

The answer matters enormously. Because sustainable outcomes usually require incentives (or tough regulatory approaches).

If markets cannot distinguish between better and worse production practices, investment signals remain weak and nothing changes.

If they can distinguish, and crucially if buyers are willing to pay a premium, the economics begin to shift.

Is the metals industry facing a ‘trilemma’? What does success look like?

One way of thinking about the challenge in the metals industry is through a familiar tension in energy – the trilemma. How can we have energy which is secure, sustainable and affordable?

Increasingly, a similar challenge is emerging in metals and minerals. We want supply chains that are secure, sustainable and affordable.

As governments place greater emphasis on resilience and strategic autonomy on metals and materials, does sustainability risk taking a back seat?

There is no easy answer, but Georgie was hopeful that the industry would resolve or avoid these tensions. Or to put it differently, she thought that the metals industry could only credibly supply to the energy transition if it is itself sustainable.

The energy transition is often presented as a shift to electrons; a technology deployment story. But it may also be judged on how those technologies are delivered, in other words: Can we provide enough materials for the energy transition whilst simultaneously raising environmental and social standards?

I hope so.

A better measure of optimisation skill is the Margin per MW under management - assuming we can make a fair comparison.
I've looked at margins for all 2-hour assets over the last 12 months, excluding commissioning periods, and it shows:
- BP consistently ranked at or near the top (albeit with a small fleet)
- Tesla is consistently in the top 5
- EDF had some great months but sometimes dropped down the rankings
- Statkraft and Field's relative performance has improved over the year
👉 As an asset owner do you choose the optimiser that smashes it in selected months or is consistently close to the mark?

Outcomes aren’t only about optimiser strategy — location matters too. Scottish assets tend to outperform those in England & Wales. So I compared real world performance of each asset against the respective Aurora regional backcast benchmark.
The range of performance is wide — from well above 200% to well below 100% of the benchmark.
Tesla, BP and Field had assets consistently ahead; others had sites running behind.
👉 Site selection and design matter at least as much the choice of optimiser.
👉 When choosing optimiser, be sure to compare against comparable peers, not the fleet as a whole

Put those three together and you get three different leaderboards.
- Scale tells you about established players and who the market trusts
- Absolute margins shows who is consistently extracting the most value
- Comparing to regional benchmarks starts to isolate optimiser skill

Each is useful, but together they show that there is not a single winner overall.

So if you’re choosing an optimiser, what matters most?
I'm curious how others think about it - so let me know in the comments!

Note: Data is taken from the Aurora Flexplorer API.

This article was first published on LinkedIn: https://www.linkedin.com/posts/richard-howard-25261023_which-is-the-best-battery-optimiser-in-gb-ugcPost-7481004911858532352-T917/

But as we’ve added more and more renewables to power systems, this has started to lead to periods of excess renewables production - where renewables (plus nuclear and any other must run plant if present in a system) exceeds total demand.

How does the market clear in these situations? We’ve all heard about the surge of negative power prices in European power markets over the last few years. I thought I would dig into the data for Spain and Germany, using the Aurora Energy Research API.

The price distribution tells two very different stories in Spain and Germany about the incidence and depth of negative prices, and market clearing around exactly €0/MWh.

Spain sees lots of low and negative price periods. This was particularly so in February 2026 where around 45% of periods cleared at a price below €1/MWh and around 32% at or below zero. Around 12% of intervals settled at exactly €0/MWh. The market is repeatedly clearing at the boundary between assets that curtail below zero and those that curtail at zero. When prices do go negative, they only go slightly negative - as there is a sufficient volume of merchant projects in the system that would rather cut their output than deliver into a system at a negative price.

Germany on the other hand sees fewer negative-price events, but much deeper ones. In April, over 8% of price periods fell below -€10/MWh, and more than 1.5% below -€100/MWh. The market rarely pauses at €0 — it falls straight through it.

The difference reflects renewables business models and the prevalence of legacy subsidy support schemes. Spain's primarily merchant fleet has a strong incentive to curtail when prices turn negative, whilst many German assets still have subsidy protection that allows them to keep generating profitably even when prices go strongly negative (though this was amended for more recent German auction rounds).

For many PPA and CfD projects and contracts, the price point in the market which matters the most is exactly €0/MWh. A single character in the offtake contract can determine whether generation is paid or not, economically curtailed or not. That can be worth millions and make or break a project.

Is grid operation keeping up with the pace of renewables?
What caused the Iberian blackout last year — and what can we learn?

These are the key questions we unpack in the latest episode of #EnergyUnplugged.

The energy transition is no longer just about building renewables —
it’s about designing and operating power systems very differently to how we did in the past.

As solar and wind scale, the system is quietly losing what used to keep it stable:
→ synchronous inertia
→ voltage support
→ predictable behaviour

Unless we rethink how we design and operate grids for high renewable penetration, the risk of blackouts will rise.
Framing the Iberian blackout as a “renewables problem” misses the point.
It happened during a period of high renewable generation — but the root causes were system design, system operation, and a lack of flexibility.

This was a system design stress test. And more markets will face the same challenge.

The implication is clear: reliability is shifting from generation adequacy to system operability.

This changes where value sits — and what needs to get built.
Batteries, synchronous compensators, grid-forming technologies, demand-side flexibility, new frequency and voltage markets, updated grid codes —
not add-ons, but core to stable power systems.

In this episode, Christina Rentell and I break down what caused the blackout, what’s changed since, and why it matters.

🎥 Watch on Youtube here: https://lnkd.in/ea6g4Uja
🎙 Watch or listen on Apple Podcasts: https://lnkd.in/eeqhDjc5

Subscribe now

That sounds counterintuitive, but look at the chart - a highlight from an Aurora Energy Research roundtable I attended in Madrid this week.

Spain's network tariffs and charges are largely based on when demand has historically been highest. As a result, consumers often face higher non-wholesale charges during the middle of the day.

That might have made sense in a pre-renewables world but today it increasingly clashes with reality. Midday is now often when solar generation is abundant, wholesale prices are lowest, and the system is struggling to absorb excess renewable output. In many solar hours, generation is curtailed because there is simply not enough demand.

Yet the tariff signal still encourages consumers to do the opposite: consume less during solar hours.

This raises a broader question about how we recover the cost of the electricity network.
Historically, basing grid charges on peak demand was logical because network investment was driven by the need to meet demand peaks.

But in a system increasingly dominated by renewables, is all demand created equal?
Should a consumer drawing power during a sunny afternoon in a region with abundant local solar face the same network charges as somebody consuming during a congested evening peak and relying on energy transported hundreds of kilometres across the grid?

The future likely lies in tariffs that are not just dynamic in time, but also in location.
Network tariffs in GB reflect spatial dynamics - e.g. whether the location is demand or generation dominated - but are temporally less dynamic.
Germany is exploring forms of dynamic grid charges that vary according to local network conditions. The underlying idea is simple: consumers should face price signals that better reflect the costs (or benefits) they create for the system.
Ultimately, this problem can also be solved through Locational Marginal Pricing, where network constraints are fully reflected in the wholesale price.

The energy transition is forcing us to rethink many assumptions.
"Use less at peak times" made sense when the challenge was meeting peak demand. But in high renewables systems the challenge increasingly becomes finding productive uses for abundant zero-carbon electricity!

High solar penetration is fundamentally changing the wholesale price shape in Germany — and the signals are becoming impossible to ignore.

This chart shows the average daily price shape across Q2 seasons from 2020 to 2026. As solar capacity has scaled up year after year, it has progressively flooded the midday market, dragging prices down during the hours of peak solar generation. What started as a modest suppression has become a full collapse: midday prices turned negative in Q2 2025 and crashed further negative in April 2026.

But the flip side is just as dramatic. With solar contribution dropping off in the evening, a sharp ramp in residual demand hits a market with insufficient flexible capacity. The result is an increasingly strong evening price spike. During the heatwave last week, the average evening peak exceeded 350 EUR/MWh — surpassing even the 2022 energy crisis - and individual days saw higher prices still.

The spread between midday trough and evening peak is now huge. In a sense the market is sending a clear message: standalone fixed-tilt solar makes less and less revenue during the sunniest hours. The economics are pushing developers firmly towards co-location with battery storage or tracking solar. With this scale of intraday volatility, and the corresponding boost to BESS margins, it's no surprise that interest in grid-scale BESS is at an all-time high.

Solar PV has now passed 3 TW of installed capacity globally.
That's an astonishing milestone (having only passed 2TW in 2024!)
Just a few years ago, the idea that solar would become the dominant source of new electricity generation worldwide seemed ambitious.

Today, the question is no longer whether solar can scale.
The question is: can our grids keep up?

In the latest episode of Energy Unplugged, I was delighted to be joined by Sonia Dunlop, CEO of the Global Solar Council (GSC), for a wide-ranging discussion on what comes next for the world's fastest-growing energy technology.

Among the topics we explored:
☀️ Why solar deployment continues to grow rapidly despite the short-term slowdown in China
🔋 Why batteries are becoming the essential partner technology for solar
⚡ How curtailment, congestion and system flexibility are emerging as the next major challenges
🏠 The growing role of distributed energy resources, rooftop solar and household batteries
🌍 Why market design, grid investment and regulatory reform will be critical as renewable penetration continues to rise

Solar's first 3 TW were impressive.
The race to the next 3 TW may depend less on panels—and more on grids, storage and flexibility.

🎥 Watch the full episode on YouTube: https://lnkd.in/eR9EVyYG
🎙 You can also watch or listen on Apple Podcasts: https://lnkd.in/e55nap_p

That matters because this was not one institution having a bad week. It was multiple parts of the country’s operating system reacting to the same stress. The heatwave did not stop Britain functioning altogether (and in this article I will be careful not to be too dramatic). But it did show how quickly schools, transport, healthcare and parts of the energy system fall back on precaution or restriction once temperatures move outside the historical comfort range for which they were built.

It is worth being clear about what we consider ‘hot’ here. These temperatures were severe for Britain. The Met Office issued a rare red extreme heat warning for 24th and 25th June, forecasting at least 39°C and warning that this might be exceeded in places. The outcome was slightly lower but still set new records: Gosport hit 36.1°C on Wednesday 24th, before the June temperature record increased further to 36.7°C on Thursday 25th.

But these temperatures are not extraordinary by wider European standards. In the same heatwave, Paris hit a June record of 40.9°C, while France as a whole had its hottest day ever nationwide. Temperatures hit a peak of 45°C in Spain and 42.7°C in Portugal, both below their hottest days ever which exceeded 47°C. Statistically, the high temperatures in GB and France were more of an anomaly and more ‘exceptional’ than those in Iberia where very hot days are more common. The disruption was not limited to Britain: France also closed schools at scale under severe alerts, with 13,500 schools reported to be closed or put on special schedules on Thursday.

These comparisons are important because they help us to avoid two inaccurate conclusions. The first is that Britain has somehow uniquely “failed” while the continent carries on as normal - it did not. The second is that disruption in Britain proves the heat was unprecedented in a wider sense - it does not. Both views are too simplistic. The more useful conclusion is that the temperatures that now begin to destabilise British systems are already occurring in a warming Europe, and they are increasingly plausible here too.

Schools are the clearest example. Here the institutional logic is easy enough to understand. There is no legal upper temperature limit for classrooms, and many schools were dealing with buildings that trap heat with poor ventilation, little shading and very limited cooling. Against that backdrop, closing early or shutting for a day can appear the prudent decision from a school’s perspective. It was reported that around 1,000 schools across England and Wales closed or shortened the school day, while teaching union said some members had passed out in class when temperatures in some classrooms reached 40°C.

But it is also where the weakness of the current approach becomes most visible. Closing schools may derisk the situation from a school’s perspective. It does not remove the risk for children or families. Some parents can adjust, others cannot. Some children go home to cooler, supervised environments; others may be in overheated homes, left with improvised childcare arrangements, or simply lose structure and support at short notice. So while the school closure is understandable as an institutional response, it is harder to defend as a properly joined-up social one. In that sense, closure can be read not simply as caution but as a form of maladaptation: a way of transferring risk when systems are not equipped to absorb it properly.

Transport showed the same pattern in a different form. Network Rail urged passengers travelling to, from or within the extreme heat zones to travel only if absolutely necessary, warning that trains would run at reduced speeds and on amended timetables, with a heightened risk of delays and cancellations. There were also incidents which made the warnings tangible: a Great Northern service between Moorgate and Stevenage broke down near Drayton Park, with passengers eventually evacuated along the track in very hot conditions.

The same defensive approach spread into wider public life. Hospitals cancelled appointments and, in some cases, declared critical incidents as cooling systems and services came under pressure. Public attractions and ceremonial events also closed or adapted. Even Wimbledon qualifying was briefly disrupted when its electronic line-calling system failed in the heat.

Taken together, this was not a national shutdown. But it was a visible shift into defensive operating mode, somewhat reminiscent of the COVID lockdown era. That in turn points to the deeper issue: the UK still too often treats heat as a temporary emergency to be managed around, rather than as an increasingly normal condition to be designed for.

The power system is a particularly revealing example because, unlike schools or rail, it did not generate the same visible public drama. Yet the signal it sent may be just as important. On 24 June, NESO issued an Electricity Margin Notice for the 7pm to 10pm evening peak, then cancelled it later that day. NESO is clear that such notices are routine operational tools and do not mean electricity supply is at risk. In other words, this was not a blackout warning. But it was an indication that margins were tighter than the operator would like, and that the market was being asked for more headroom.

What made the episode interesting was the pattern behind it. The tight margins were due to extremely high temperatures affecting Great Britain and the continent, combined with low wind. Solar covered roughly half of UK daytime demand on the Tuesday, but that this flipped by evening, when the UK was importing power from Europe as solar output fell and demand remained elevated. NESO had sought an extra 1,900 MW for the evening peak.

That is important because it points to a newer kind of stress. For years, the standard shorthand for the stress point in renewables-heavy systems has been the winter Kalte Dunkelflaute (my favourite German word): a dark, windless period in which both wind and solar generation are weak. That problem is very real, and it remains central to debates about adequacy, flexibility and security of supply in a renewables-dominated system.

But this week points to a different, increasingly relevant problem: the summer version, referred to as Hitzeflaute. The label is useful because it captures a specific operational shape. In a hot, still weather pattern, daytime solar generation can be very strong. But the challenge comes later in the day. Solar falls away in the evening just as cooling demand and the broader system ramp can remain elevated. That creates a sharp intraday imbalance: not too little solar overall, but too much at the wrong time of day, combined with low wind and tighter evening margins.

Seen like that, the NESO episode was not an oddity but a warning shot. Britain’s electricity system must now increasingly be operable across both kinds of weather stress. It still has to handle the winter Kalte Dunkelflaute problem of cold, low wind and low solar. But it also needs to handle the newer summer Hitzeflaute problem of heat, low wind, strong daytime solar and difficult evening ramps. For a country trying to electrify more of heat, transport and industry, that is not a niche operational curiosity - it goes to the heart of what system resilience now means.

This reveals the common thread across the week’s disruption. In schools, precaution meant closing or shortening the day. In rail, it meant reduced speeds, service cuts and discouraging travel. In hospitals, it meant postponing planned care when cooling systems failed or pressure rose. In electricity, it meant seeking extra operational margin ahead of a tight evening peak. None of these responses was irrational on its own terms. But all of them point to systems leaning harder on precaution because adaptation has lagged behind the climate reality they increasingly face.

So what would a more serious adaptation agenda look like?

First, it would start with buildings. The UK has been far too slow to take overheating seriously as a mainstream design problem. Research on overheating risk has already identified the most effective passive adaptation measures: external shading, solar control glass and improved ventilation. For schools in particular, work by Arup for London found that 98% of schools reported overheating as an issue, and set out a broad set of resilience measures. Yet the gap between that diagnosis and widespread retrofit remains huge.

Second, it would treat cooling as a legitimate adaptation need rather than an afterthought. Policy has only started to edge in that direction. The Government has had a heat decarbonisation strategy for a while now, but talks a lot less about cooling. In November 2025, the government expanded the Boiler Upgrade Scheme to include air-to-air heat pumps, explicitly acknowledging technologies that can keep homes warm in winter AND cool in summer. That is a useful step, but also a tellingly late one. Britain is only beginning to act as though summer cooling might be part of climate adaptation policy rather than a continental luxury.

Third, public institutions need more coherent operating principles under heat stress. Schools should not be left to improvise threshold decisions in largely unsuitable buildings with no legal maximum classroom temperature. Transport systems need clearer long-term pathways for retrofitting infrastructure that was built for a cooler climate. Hospitals and public buildings need more resilient cooling and ventilation if essential services are to continue reliably under heat stress. A 2023 UK heat adaptation policy brief argued that building regulations and wider frameworks remain fragmented, with particular weakness in protecting the existing stock rather than just new build.

And fourth, adaptation needs to include the electricity system explicitly. Adaptation cannot just be a story about shading schools and planting trees, important though those things are. It must also be about how NESO plans operability in a system with more solar, more electrification and more frequent climate extremes. In practice that means recognising two stress cases at once: not only the winter Kalte Dunkelflaute, but also the summer Hitzeflaute. Both are now real system problems. Both require flexibility, storage, interconnection, clearer demand-side response and credible arrangements for maintaining margins when renewable output and demand shapes do not line up neatly.

That is, ultimately, what this week has taught us. The UK does not just need to survive occasional heatwaves. It needs schools, transport, hospitals and electricity systems that can continue to function through them. At the moment, too many institutions still respond to heat by stepping back, scaling down or passing risk elsewhere. That may be understandable in the moment, but it is not a credible long-term strategy for a warming country.

The deeper choice is not between panic and “just making do”. It is between a country that treats heat as a freak interruption to normal life, and one that redesigns normal life so it can carry on in hotter conditions. Britain has made far more progress on recognising the need to mitigate climate change than on adapting to it. This week’s disruption suggests that the second task can no longer remain secondary.

It was a great pleasure to host on #EnergyUnplugged !

In this in depth conversation we cover:
- Why electrification of nearly everything is so critical to the energy transition
- The 'primary energy fallacy' - why focusing on primary energy overstates the scale of the challenge for the energy transition
- Deep dives on the status and outlook for electrification in transport, heating & cooling, and industry

🎥 Watch the full episode on YouTube: https://lnkd.in/epjN-EDh

🎙 You can also watch or listen on Apple Podcasts: https://lnkd.in/enP_qCvu

📖 For a deeper dive into the topic, read Jan Rosenow's accompanying article: https://lnkd.in/eZpyRU6V

There is a paradox at the heart of the energy transition.

We are spending hundreds of billions on new infrastructure: wind farms, solar parks, batteries, EV charging networks and grid reinforcement. Yet much of the infrastructure we already have remains surprisingly underutilised. Networks are congested in one location and idle in another. Renewables are curtailed while consumers pay high prices. Flexible assets sit dormant when they could be helping to balance the system.

What if the next great breakthrough in energy wasn’t building more infrastructure, but using existing infrastructure far more intelligently?

That was the subject of a conversation on Aurora’s Energy Unplugged podcast, where I spoke with Devrim Celal, Chief Marketing & Flexibility Officer at Kraken, about how AI, flexibility and digitalisation could transform electricity systems over the coming decades. The discussion explored how power systems are evolving from centrally managed networks into highly dynamic platforms coordinated by millions of connected devices.

From power stations to platforms

For most of the last century, electricity systems followed a relatively simple model.

Large power stations generated electricity, transmission operators moved it across the country, and consumers passively consumed whatever arrived at the socket.

That model is disappearing.

Today’s power system increasingly consists of millions of distributed assets: rooftop solar panels, home batteries, electric vehicles, heat pumps and smart appliances. Individually they are small. Collectively they represent an enormous resource.

The challenge is coordination.

An electricity grid with millions of active participants cannot be managed using twentieth-century approaches. The amount of data becomes overwhelming. Decisions need to be made in seconds rather than hours. Optimisation must happen continuously across countless devices.

This is where AI starts to become interesting—not as a futuristic add-on, but as a practical necessity.

The flexibility revolution

One of my long-standing observations about energy markets is that most discussions focus on generation.

When policymakers worry about security of supply, they ask where the next power station will come from.

When developers invest capital, they ask which technology will produce the next megawatt-hour.

But increasingly the more important question may be: when will electricity be consumed?

As renewable penetration grows, flexibility becomes more valuable. Electric vehicles can charge overnight rather than during the evening peak. Batteries can absorb excess solar generation and discharge later. Heat pumps can shift demand without consumers noticing any difference.

The economic logic is compelling.

Every unit of flexibility reduces the need for costly network reinforcement, lowers balancing costs and enables higher levels of renewable generation. As discussed in the podcast, distributed flexibility is becoming a critical ingredient in renewable-heavy power systems. [youtube.com]

The energy transition is therefore not simply a story about replacing fossil fuels with clean generation. It is also a story about making demand more intelligent.

The hidden bottleneck

One of the most fascinating parts of the discussion concerned visibility.

Most people assume network operators have a detailed understanding of everything happening on the grid. In reality, low-voltage networks often remain surprisingly opaque. Operators can see what happens on major transmission assets but have far less visibility of activity on residential streets and local distribution networks. Low-voltage grid visibility is increasingly emerging as one of the industry’s biggest challenges. [youtube.com]

A useful analogy is road traffic.

Imagine trying to manage a national motorway network while having no information about local roads. Congestion would appear seemingly from nowhere. Bottlenecks would be difficult to predict. Investments would be harder to prioritise.

That is broadly where power systems find themselves today.

The proliferation of smart devices creates an opportunity to solve this problem. More sensors, more connectivity and more granular data allow operators to understand what is happening in real time.

The result is not just a more efficient grid. It is a more investable grid, where decisions can be based on evidence rather than approximation.

Towards the autonomous grid

Perhaps the most provocative idea explored in the conversation was the concept of an “autonomous grid”.

The phrase may sound futuristic, but the underlying trend is already underway.

Across the economy, we are moving from systems that are directly operated by humans to systems supervised by humans but increasingly executed by software. Financial markets provide an obvious example. Supply chains offer another.

Energy appears to be heading in the same direction.

AI systems can already optimise batteries, forecast demand, schedule charging events and participate in electricity markets. As data quality improves and interoperability standards mature, the scope for automation expands rapidly. The discussion explored how AI, real-time optimisation and intelligent orchestration could significantly increase utilisation of existing network infrastructure. [youtube.com]

That does not mean humans disappear from the process.

Rather, their role changes. Instead of making thousands of operational decisions, they set objectives, oversee outcomes and intervene when necessary.

The destination is not a fully autonomous system. It is a system where humans and machines each focus on what they do best.

A reminder that software matters

For much of the energy transition, the spotlight has been on physical assets.

Wind turbines. Solar panels. Batteries. Transmission lines.

All are essential.

Yet increasingly, software is becoming just as important as steel and concrete.

The difference between a highly efficient power system and an inefficient one may not be the quantity of infrastructure deployed. It may be the quality of intelligence coordinating it.

The energy transition therefore has two dimensions. The first is physical electrification. The second is digital optimisation.

We’ve spent the past decade focused on the former.

The next decade may belong to the latter.

And if that proves true, the future winners in energy will not simply be those who build the most infrastructure. They will be those who learn how to orchestrate it most effectively.

My biggest takeaway? The energy transition isn’t just becoming cleaner. It’s becoming smarter. And that may ultimately prove just as important.

In the second blog, I set out the broad spectrum between market-led systems, coordination-led systems, and the hybrid models that sit between them. I outlined ways in which a hybrid system can either be a coherent package with the best of both worlds, or a messy compromise in between. The risk with hybrid systems is not that they combine markets and coordination. It is that they do so without being clear about who is responsible for what, how investors are meant to respond, and where risks ultimately sit.

To put it more provocatively - who is to blame if/when things go wrong? A market-led system tends to fail in ways that are very visible and attributable: price spikes, scarcity events, or even loss of load. A more coordinated system can fail more diffusely: through procurement errors, rising fixed charges, delayed network build or incremental cost escalation that only becomes obvious over time. That does not make coordination-led failure less real — but it can make it more politically tolerable, precisely because the costs are harder to trace back to any single decision.

That is the lens through which I want to examine Great Britain as a case study in this third blog. GB is far from market-led but is not (yet) moving fully back to straightforward central planning. It is trying to hold onto the benefits of both – though arguably the latest policy proposals move it much further towards the planned end of the spectrum. GB already has amongst the highest industrial electricity prices in Europe and the OECD, which underlines that diffuse cost accumulation is not a theoretical concern.

Can GB keep a coherent package, or does it end up in the messy middle, with further costs accumulating along the way?

Some brief context…

As we saw in the first blog in this series, GB was one of the first power markets to privatise and liberalise globally – and in its early guise it performed well in dispatching thermal assets and sending investment signals. Arguably at this point it was at the market-led end of the spectrum.

GB has been on an ambitious pathway to decarbonise its power system – introducing renewables subsidies in the early 2000s, creating a system of carbon budgets under the 2008 Climate Change Act, setting a 2050 economy-wide net zero goal in 2019, and clean power targets to 2030. It has been a testing ground for how the net zero energy transition alters power market dynamics and a wide range of policy changes have been considered along the way.

The first major overhaul of power market design was Electricity Market Reform (EMR) in 2011-13. This introduced an additional carbon price (we already had the EUETS but the price was too low to have much effect), transitioned renewables subsidies to the Contracts for Difference model, and introduced a Capacity Mechanism (CM). The 2010s saw a massive uptick in renewables deployment – subsidy costs went up in absolute terms but competition and technology improvements drove the costs per MWh down. The CM performed well in keeping the lights on – mainly by building peakers in the early days (later lots of batteries) not the CCGTs that were once envisaged.

Fast forward to 2022 and the Government kicked off another major Review of Electricity Market Arrangements (REMA). At its core, REMA began as a review of whether GB’s electricity market arrangements were still fit for purpose in a decarbonising power system. The original objective was not simply to revisit wholesale pricing, but to ask a broader question: how should the market evolve to deliver a decarbonised power system that remains secure, cost-effective and investable?

In the early phases of REMA, pretty much everything was on the table. Aurora held various workshops with Government to bring our ideas and insights, including examples from other markets – in fact this is where I first developed the market design framework I introduced in the first blog in this series.

Locational pricing was a huge topic in the early part of the REMA process. Zonal and nodal models were debated extensively, heavily modelled, and treated seriously as possible ways to sharpen dispatch and siting signals. The attraction was obvious: in theory, more granular pricing could reduce constraint costs, improve locational decisions and lower total system costs. But the evidence was unclear and often conditional rather than decisive. Aurora produced a public report on locational pricing models which made this point clearly: the modelled savings could be real but relatively modest, and they were highly sensitive to the wider design and transition path. If locational pricing raised investor uncertainty and nudged up the cost of capital for renewables and other capital-intensive assets, those theoretical benefits could disappear very quickly — or even turn negative.

By the time government reached its 2025 decision, that trade-off had become central to the debate: whatever the textbook efficiency gains, ministers were not willing to undermine investor confidence by introducing a high-complexity transition at the very moment the system needed unprecedented volumes of capital. (note: hold that thought for later).

Reformed National Pricing: the return of the very visible hand

The REMA summer update in July 2025 set out a clear direction of travel for GB: reject zonal pricing, retain a single national wholesale market, and pursue a more coordinated and strategically planned electricity system. The Reformed National Pricing (RNP) package is built around the Strategic Spatial Energy Plan, network build, connections reform, charging reform, and operational changes to balancing and constraint management.

In a sense, by rejecting a more market-led pathway involving LMP, the alternative became (even) more central planning. The REMA 2025 update document is unusually explicit that the state holds many of the levers needed to solve the siting problem — spatial planning, seabed leasing, procurement, network planning, charging, price controls — and argues that more/better coordination is required.

This is a coherent thesis on paper. But the harder question is whether it can be turned into a coherent implementation framework, and crucially whether it is preferable in the end to the LMP model which was rejected.

The shift towards central planning is also visible in the evolution of the CfD scheme. It began as a revenue-stabilisation mechanism with a competitive allocation. At the time, ministers presented the level of intervention as something to be reduced over time as renewables reached maturity and grid parity. Instead the CfD has become the key instrument to drive renewables deployment in the UK, with Government making increasingly granular choices about what gets built – the latest announcements around CfD Allocation Round 8 go even further in this direction.

Siting and investment levers: big questions, unclear destination

In March 2026, Ofgem opened its call for input on locational charges and regulatory siting levers under RNP. The paper intentionally explores options ranging from incremental refinements to more transformative approaches, with reforms anticipated by 2029, and it emphasises the need to support predictability, fairness and investor confidence during the transition.

The consultation is asking a fair and important question: if GB is not going to use locational wholesale prices, where should the locational signals for investment come from?

It reopens some fundamental market-design choices. It sets out how locational signals can come from coordinated access and permissions; from price-based signals such as locational TNUOS grid charges; or from some combination of the two. TNUOS charges are currently set such that generators pay the marginal cost of transmitting electrons added to the current system. The document floats an alternative where TNUOS would be set based on the marginal cost of transmission against the future, planned system. For new and existing wind assets connecting in Scotland, this means that TNUOS charges might be high and rising (as they are now), somewhat lower (if the cost of planned grid build is socialised) or zero.

This level of uncertainty could easily make or break an investment. Once you tell investors that locational signals may be re‑written through a combination of charging reform, connections reform, and potentially new siting levers, you change the risk profile of projects in development. Even more problematic – it is not just new assets but also existing assets that are subject to these risks, hence grandfathering arrangements become critical.

Having rejected Locational Marginal Pricing as too risky for investors it is not yet obvious that the alternative now proposed performs better against the same test.

Dispatch and balancing reforms: incremental fixes — or a move toward central coordination?

The other major strand of Reformed National Pricing is operational: how to make the system more operable and reduce the cost of balancing while retaining a national wholesale price. Interestingly, whilst the reforms are couched in terms of thermal grid constraints, the reforms are not about sharpening locational incentives – but making dispatch work better in general. It is questionable whether this is the biggest issue that the current system faces right now.

In early 2026, NESO launched its call for input on balancing, settlement and dispatch reform under RNP. The starting point is familiar: rising redispatch and congestion, insufficient visibility of balancing resources, and market structures and incentives that can cause inefficiencies.

It proposed the following package of proposals to tweak balancing and settlement:

• lowering the mandatory Balancing Mechanism participation threshold (to potentially as low as 1MW) to increase visibility of smaller and distribution-connected assets;

• aligning the market trading deadline with Gate Closure, so trading and balancing actions do not overlap;

• requiring Final Physical Notifications to match traded positions;

• moving towards more granular, unit-level bidding;

• shortening settlement periods to 15 or 5 minutes;

Some of these changes seem logical in isolation but taken together they constitute a significant shift in the DNA of the market - providing the system operator with much more visibility and strengthening its role.

The much bigger question raised in this document is whether GB should move away from self-dispatch toward more centralised or ‘hybrid’ approaches to scheduling and dispatch. This matters a lot because it is not just a technical tweak to balancing rules. It would shift who decides which assets run, how closely financial and physical positions must align, and how much discretion sits with the system operator rather than market participants.

Balance of opinions?

Industry responses to this package of reforms have been mixed at best.

Most respondents seem aligned on reducing the BM threshold to bring more visibility into the system, though past a point this adds significant administrative costs onto smaller players. Dropping the threshold to 1MW would bring thousands more assets into the fold, so it is unclear whether this strikes the right balance of costs and benefits. Similarly, most respondents supported shorter settlement periods but highlighted implementation costs and complexities (though most of Europe has moved to 15 minutes already without issue). So far so good. Though doing both together creates orders of magnitude more data for the system operator to manage – it has taken NESO a long while to make progress on harnessing smallish units already in the BM; this would be a completely different ball game.

There was far more pushback on the proposals to align the trading deadline with gate closure and to force alignment of FPNs to traded positions. Market participants flagged that this cuts across how assets and portfolios are currently traded and optimised, and that limiting near-term trading reduces flexibility. Recent Aurora analysis shows that FPNs are a poor predictor of outturn metered output for renewables and batteries. For renewables this is due to inherent forecasting error. It might be right to expose the true costs of these imbalances, but the rule change adds complexity and could result in more conservative bidding behaviour. Batteries continue to trade after submitting FPNs, and the system wants to harness this short-term flexibility. Unit level bidding received similarly strong pushback by market participants.

Although the proposals around hybrid/centralised dispatch were positioned as ‘exploratory’ they received significant pushback from respondents. Centralised dispatch had been dropped by DESNZ earlier in the REMA process on grounds of deliverability, investors confidence and value for money – so many respondents questioned why a move to centralised/hybrid dispatch is back on the table. Unpacking this, it seems that some parties feared that under this model a generator cannot reliably predict how much of its output will be taken up by the system operator. It is another area where replacing market risk with policy risk can deter investors. Others cited how this would complicate cross-border trading if Europe did not follow a similar path.

How does this weigh up against alternatives such as LMP?

Overall, it seems that industry was in favour of incremental tweaks under RNP, but cautious of implementation complexity, and strongly opposed to a major overhaul.

It is instructive to refer back to the reasons that LMP was rejected and look at the new RNP package through the same lens. The reasons government gave for rejecting zonal pricing in 2025 were: concerns about investor uncertainty and resulting increases in the cost of capital; fairness / distributional issues; and delivery complexity.

Implementing the full set of RNP proposals carries many of the same risks. The reforms to siting and investment levers create significant policy and regulatory uncertainty. There is significant delivery complexity. Governance and risk management are potentially moving around in ways that are hard for investors to read.

This is why ‘coherence’ is a harder test than it first appears. A system can be coherent in principle, yet still drift into a less accountable form of coordination if layers of small interventions accumulate without a clear allocation of responsibility. In that world, no single reform is decisive on its own, but the combined effect is to weaken market signals and make it harder to say who is accountable for the resulting costs, or if things go wrong.

None of this means that RNP is wrong or that we should not reform at all. But I think we need to recognise that the proposed package has reopened some fundamental questions and created huge uncertainties. There is now a very wide range of outcomes in terms of where GB will land in the markets vs coordination spectrum, though it seems likely to lean much further towards coordination.

The real risk at the coordinated end is policy and political commitment

A more coordinated system can absolutely deliver faster build-out, clearer siting decisions and fewer constraints. But it does not make the hard choices disappear; it moves them from being market outcomes to coordination and policy decisions.

The risks are familiar: planners still have to act under uncertainty; procurement and classification decisions become politically contested; and once the state takes a stronger hand in shaping the system, it becomes easier to lock in today’s assumptions and harder to adapt when technologies, costs or politics change. That is also why failure in a more coordinated system often becomes slower-moving and harder to attribute: the costs accumulate through planning, procurement and fixed charges rather than arriving in one visible market event.

Under the latest proposals, the role of markets is being further squeezed out: under a largely planned and hybrid dispatch system the ‘market’ which remains is mainly a cost of capital game linked to capacity auctions.

For this to work well – delivering efficient outcomes and investment at pace - the system needs to be disciplined, adaptable and above all credible. That, in the end, is the thread running through all three blogs. Power market design is not just about prices, contracts or institutions in isolation. It is about whether the system is clear about what is being coordinated, how risks are allocated, and whether the framework remains believable when it is put under pressure

Arguably the GB system is being put under A LOT of pressure at the moment! If gilt yields are anything to go by then we are living through turbulent times. The Reform and Green parties surged in recent elections; whilst support for Labour has dropped. Sir Keir Starmer’s premiership is under threat, with key contender Andy Burnham favouring a reversal of privatisation. On the other hand, polls point towards Reform winning the most seats in the next UK general election, with a policy platform to scrap net zero altogether.

A more coordinated system depends more heavily on political durability. But in a politically volatile environment, that is tough to deliver.

If policy commitment cannot be taken for granted, and GB is going to lean further into coordination, then a few principles matter:

First, judge RNP against the same tests used to reject zonal pricing. If zonal pricing was rejected because it risked unstable long-term signals, investor uncertainty, higher financing costs, distributional tensions and delivery complexity, then the preferred alternative should be held to the same standard. Otherwise the debate risks becoming inconsistent: rejecting one form of uncertainty while introducing another through a complex package of coordinated interventions.

Second, be explicit about the direction of travel. If GB is moving towards a more coordinated model, then just say so explicitly. Trying to preserve the language of markets while shifting more and more decisions into planning, procurement and system operator decisions makes it harder to read. Investors can adapt to many things but ambiguity about the underlying philosophy and model is one of the hardest.

Third, worry as much about the transition as the end-state. Opening up too many options at once creates uncertainty, and investors will price that uncertainty immediately. High-confidence, high-value measures should be separated early from the more speculative or disruptive ones. Otherwise a package intended to reduce system costs could end up raising financing costs instead.

Fourth, keep the framework legible. Investors do not need every risk removed, but they do need to understand which instrument is doing what - and who is accountable as more layers of intervention accumulate over time. Under RNP some of this is becoming unclear, with grid access and grid fees potentially both influencing siting decisions. A more coordinated model only works if roles, risks and responsibilities are allocated clearly enough that market participants can still form a view.

Fifth, leave room for innovation and demand-side flexibility. A risk in more coordinated systems is that they become very good at supporting what planners already understand, and worse at discovering new technologies, business models or forms of flexibility. That matters particularly on the demand side, where EVs, batteries and smart tariffs could become a huge source of system value. Any reform path should be careful not to hard-wire solutions so tightly that it squeezes out experimentation or leaves too little room for demand-side response to do its job.

Closing thoughts

The question running through this whole series is not whether markets are “good” and planning is “bad”, or vice versa. It is what power markets are now being asked to do in a net zero system, and whether any given market design is honest and coherent about where coordination sits. GB is a live test of that broader question. If a system wants to rely less on prices to guide outcomes, then it needs to be clear about what markets are still for, which decisions are now being taken administratively, and how that shift can remain credible over time.

If you’ve made it this far in the series then… frankly well done! This started as a small side project, expanded to 3 blogs then ballooned to more or less a thesis!

I hope you have enjoyed reading it as much as I enjoyed writing it. Please do comment, repost and connect. If there are questions left unanswered, or you’d like me to cover a specific market next time then get in touch.