France Is Repeating April's Forecourt Shortage. The Market Behind It Is Tighter.

OilWatch Network Analysis — seventeen per cent of French filling stations are missing at least one fuel. France reached 18% in April under almost exactly the same retail conditions. The forecourt disruption is therefore not the warning. The warning is that the global system supplying diesel now has fewer export barrels, depleted seasonal inventories and refineries already running close to their limits.

France is not running out of diesel.

The strongest evidence for that conclusion is not the size of France's strategic reserve. It is what is happening at filling stations outside TotalEnergies.

On Monday morning, Economy Minister Roland Lescure said 17% of French filling stations were missing at least one fuel, against 16% over the weekend. But 91% of the affected stations were TotalEnergies sites. Total currently caps diesel at €2.25 per litre, while the French average was about €2.41 on Sunday. Lescure described the concentration as the direct consequence of Total's price ceiling rather than evidence of a national supply failure.

Five months ago, France ran almost the same experiment.

On 7 April, around 18% of French stations were missing at least one fuel and 83% of the affected sites were TotalEnergies stations. The government's explanation then was essentially identical: France did not have a refinery or national-stock shortage; cheaper Total fuel had redirected demand into one network faster than its logistics system could replenish individual forecourts.

That April episode is the base rate against which September has to be judged.

Normalise the numbers and September looks less alarming

Raw percentages such as 83% and 91% can mislead because they depend on the size of the affected population.

Normalising them by network size produces a much more useful comparison.

TotalEnergies says its metropolitan French network contains approximately 3,300 stations. The government monitors roughly 9,900 stations nationally.

These are OilWatch calculations from rounded national figures, not official network-specific counts. They also measure stations missing at least one fuel, not diesel ruptures specifically.

But the result is striking.

Total's implied network disruption is almost identical in the two episodes: roughly 45% in April and 46% now.

Outside Total, the opposite is true.

The implied disruption rate falls from about 4.6% in April to 2.3% in September.

That is arguably the strongest evidence available that France does not currently have a general national fuel-supply problem.

The finding also survives reasonable uncertainty in the inputs.

If the reported Total share is varied by ±2 percentage points, September's implied non-Total rate ranges from approximately 1.8% to 2.8%. Applying the same test to April produces a range of roughly 4.1% to 5.1%.

The ranges do not overlap.

There is also some ambiguity over whether Total's wider French network should be described as around 3,300 or around 3,600 stations. That barely affects the finding that matters. Using 3,600 Total sites changes the central non-Total estimates only to roughly 2.4% in September and 4.8% in April.

The exact Total penetration percentage changes more, but the conclusion about the rest of the French network does not.

Non-Total disruption today appears to be roughly half its April level.

In April, the French government independently described the shortage rate outside Total as around 4%, which provides a useful cross-check on the method.

So the national headline — 17% of stations affected — needs to be read carefully.

It is overwhelmingly a TotalEnergies event.

The price-cap mechanism still fits

The pricing incentive was enormous in April.

TotalEnergies capped diesel at €2.09/litre through 7 April across its 3,300 metropolitan stations. It then raised the diesel ceiling to €2.25 while extending its wider fuel-price policy through the end of April.

Against the national diesel price being reported immediately before the April peak, that represented a discount of roughly 22 cents per litre.

Today Total's diesel ceiling is again €2.25/litre, while the government/AFP national figure is about €2.41. That is a headline discount of approximately 16 cents per litre.

EuroOilWatch's station tracker independently shows French diesel averaging €2.419/litre, with a €2.425 median, in its latest 20 September dataset. It currently has 8,272 qualifying diesel-price observations across 8,755 active stations after excluding prices more than 14 days old.

The April and September national spread figures are not completely like-for-like. The April figure came from an AFP calculation over government station data; EuroOilWatch applies its own freshness rules today. A methodological difference of a few cents could consume much of the apparent narrowing.

So we should not claim yet that Total is suffering the same disruption despite a definitively smaller discount.

The defensible conclusion is simpler:

Both episodes combine a very large Total discount with almost identical implied disruption across the Total network.

The proper OilWatch test is local rather than national.

For each Total station, compare its diesel price with the nearest viable non-Total competitors using one consistent dataset, then compare that local discount with confirmed rupture status.

If the probability of a Total rupture increases with the size of the local price gap, the government's explanation can be demonstrated independently.

If it does not, something else is contributing.

The retail event is a replay. The wholesale market is not.

Several important stresses were already present during April.

Hormuz was disrupted. Gulf product exports had already fallen sharply. Ukrainian attacks were already damaging Russian refineries. European import patterns were already changing.

Those are not September developments and should not be presented as though they are.

What has changed since April is that the original shock has been followed by additional export restrictions, further inventory depletion and a refining system operating extremely hard without rebuilding adequate middle-distillate stocks.

The current signals are better summarised with their baselines made explicit:

S&P Global estimates that global diesel exports averaged only 5.85 million barrels per day in August, 25% below August 2025. It now expects global diesel cracks to average about $84/bbl for the rest of 2026, $31 above its previous expectation.

US refineries, meanwhile, have approached 97% utilisation this summer. S&P describes the remaining unconstrained refining system as operating close to maximum levels as autumn maintenance begins.

US inventory data make the point even more clearly.

In the week ending 21 August, distillate inventories stood at 103.4 million barrels — 14% below their five-year seasonal average and the lowest ever recorded for that point in the year — despite refinery utilisation of 97.4%.

That may be the cleanest single expression of the current diesel problem:

Refineries are running close to their limits and still cannot rebuild normal stocks.

Crude became cheaper while diesel became more valuable

This is the central market signal.

North Sea Dated crude averaged $120.36/bbl in April.

By August it averaged $91.00/bbl.

That is a like-for-like monthly decline of approximately 24%.

If crude scarcity were still the principal constraint on diesel, that decline should have provided substantial relief to refiners and product markets.

It did not.

The IEA says Atlantic Basin refining margins instead reached record levels in August, led by sharply higher diesel cracks. It also reports that global refinery throughput in August was still 4.2 mb/d below the previous year, despite rising to its summer peak.

OilWatch's cross-check of the OPEC/Argus Rotterdam gasoil series shows the monthly diesel crack rising from roughly $50/bbl in April to more than $72/bbl by July. A separate Platts physical ARA assessment subsequently averaged around $80/bbl in August.

The methodologies are not identical, so the July-to-August step should not be treated as one continuous statistical series.

The direction, however, is unambiguous.

The raw material became substantially cheaper while the value attached to converting that raw material into diesel increased.

That is the signature of a constraint that has moved downstream.

The question is no longer simply whether Europe can obtain crude.

It is whether the global refinery system can convert enough crude into the middle distillates Europe needs.

This is why France can have crude arriving at its ports, substantial petroleum stocks and functioning refineries while diesel itself remains exceptionally expensive.

Oil is not diesel.

Russia deteriorated sharply after April — but not in a straight line

Russia provides one of the clearest post-April changes.

Russian seaborne diesel and gasoil exports in April were around 3.25 million tonnes for the month.

Using approximately 7.45 barrels per tonne, that is equivalent to roughly 0.8 million barrels per day.

By the first ten days of July, Russian diesel and gasoil loadings had fallen to around 234,000 bpd.

That is roughly a 70% decline from the April daily equivalent.

Two separate mechanisms were responsible.

One was physical: repeated Ukrainian attacks reduced refinery operations.

The other was policy: Moscow imposed restrictions on diesel exports to protect its domestic market.

Those two risks must not be treated as interchangeable.

An export ban can be relaxed by government decision.

A destroyed or heavily damaged hydrocracker cannot.

The position has also improved in some respects since the July low. Russian total seaborne product exports rose 16.4% month-on-month in August to 4.57 million tonnes, as some refinery operations recovered.

But that rebound does not mean the diesel problem disappeared. August product exports remained dramatically below normal year-earlier levels, and the product mix included substantial volumes of fuel oil and naphtha rather than unrestricted diesel exports.

The IEA's September assessment captures the aggregate result: combined net diesel/gasoil exports from Russia and the Gulf were 1.6 mb/d lower in August than in February, when those two sources supplied almost 45% of global seaborne diesel trade.

The correct description is therefore not relentless collapse.

It is partial operational recovery inside a much more constrained export system.

Europe's stock buffer has been consumed

Comparing raw April and September ARA stocks alone risks a seasonal objection because gasoil inventories normally move through a yearly cycle.

The stronger evidence is their position relative to recent history.

Independently held ARA gasoil inventories — including diesel and heating oil — fell to 11.90 million barrels in August, their lowest level in four years.

September has shown some rebuilding rather than continued straight-line decline.

That matters. The deterioration is not monotonic.

But the market is rebuilding from an unusually depleted starting point.

Prompt ARA fuel availability has remained sufficiently tight that market participants have been advised to allow five to seven days to obtain competitive supply offers.

The US picture removes much of the seasonality argument altogether: inventories are below the normal seasonal range while refinery utilisation is already around 97%.

Together, those observations tell us what "less spare margin" actually means.

It is not a metaphor.

Maximum or near-maximum output is failing to restore normal inventories before winter.

Europe can rearrange barrels without creating new ones

The import picture requires similar care.

Cargo data show the sources feeding Northwest Europe shifting repeatedly as traders respond to arbitrage and disruption.

But a barrel loaded in Germany or the UK and delivered into ARA is not necessarily new supply for Europe.

It may simply represent the movement of an existing European barrel from one part of the regional system to another.

Loading origin also does not necessarily identify where a product was refined, so port statistics cannot quantify this perfectly.

Nevertheless, the distinction matters.

Europe can keep Rotterdam supplied by moving product around Europe without increasing Europe's total diesel availability.

Redistribution can solve a local shortage.

It cannot, by itself, rebuild the continental buffer.

That is why the most important replacement-supply number may be the global one: diesel exports of 5.85 mb/d in August, down 25% year-on-year.

That is the pool from which Europe, the Americas and Asia are all trying to replace missing barrels.

Jet fuel is now competing for the same refinery system

Europe also faces another middle-distillate demand.

Energy Aspects estimates that Europe could run a 510,000 bpd jet-fuel deficit during the fourth quarter.

South Korean jet shipments to Europe have reached approximately 129,000 bpd in September, their highest level since October 2022, while ARA jet inventories have fallen to a seven-year low.

Diesel and jet fuel are different finished products.

Refineries have some ability to alter yields between them.

But both draw from the middle-distillate part of the barrel, and that flexibility has limits.

A refinery complex already operating close to maximum throughput cannot independently maximise every product.

That is why the current European problem increasingly looks broader than diesel.

It is becoming a middle-distillate capacity problem.

Hormuz is being worked around, not restored

There is genuine adaptation in the Gulf.

But the numbers have to be defined carefully.

Kpler estimates that around 2.5 million barrels per day of crude will be loaded through ship-to-ship transfers in the Gulf of Oman during September, up from about 1.4 mb/d in August.

The scale is remarkable because the system was rarely used before the war.

It was developed as an emergency response after ordinary Hormuz shipping became too dangerous and expensive.

The workaround has prevented a considerably worse crude shock.

It is not a return to normal.

Current benchmark VLCC freight from the Gulf to China has exceeded $30 per barrel, the highest level on record in the cited LSEG series. With crude around $105 at the time of the calculation, freight alone represented more than a quarter of the barrel's value, compared with only 2%-3% before the war.

Earlier IEA data put comparable Middle East-to-Asia VLCC freight at around $15.65/bbl on an April monthly-average basis.

The comparison is not average-to-average — today's number is a current market level — but it shows the scale: current tanker cost is roughly twice April's monthly average.

This is adaptation at enormous cost.

It also reinforces the crude-versus-diesel distinction.

The global system has become increasingly effective at finding ways to move crude around the disruption.

Diesel economics have nevertheless become more extreme.

Restoring access to crude is therefore not restoring the missing refining margin.

France's 88.2 days belong to June

EuroOilWatch currently shows 88.2 days of diesel cover for France.

But the label matters.

The underlying stock observation is from June 2026, and the denominator is June diesel consumption.

EuroOilWatch explicitly describes the figure as "Stock observed Jun 2026 · monthly consumption, same month" and notes that its 90-day reference line is days of consumption, not the EU Directive's net-import calculation.

So the defensible statement is:

France had 88.2 days of diesel cover on EuroOilWatch's latest measured June data.

Not:

France currently has 88.2 days of diesel.

The distinction is especially important in an article arguing that buffers have been consumed through the summer.

The French stock-cover series is now almost three months behind the forecourt events being discussed.

Its next observation matters more than its June level.

The figure is also not a competing estimate of France's strategic petroleum reserve.

EuroOilWatch's measure is diesel-specific: measured diesel stock divided by same-month diesel consumption.

France's statutory and international strategic-stock obligations cover broader petroleum categories and use different methodologies.

They answer different questions.

The 88.2-day number is therefore a baseline, not a countdown.

Demand is helping France — and price is destroying some of it

French diesel demand has also declined substantially.

August road-diesel deliveries were around 2.217 million cubic metres, down 9.2% year-on-year. Diesel represented approximately 61.9% of French road-fuel consumption, while diesel deliveries over January-August were down about 7.6%.

Physically, this helps.

A litre not consumed is a litre that does not have to be refined, imported or withdrawn from storage.

But it would be misleading to describe the decline entirely as a protective demand trend.

At diesel prices around €2.40 per litre, some consumption is being rationed economically.

Households reduce journeys. Freight businesses adjust routes. Marginal economic activity becomes less attractive.

Demand destruction helps rebalance the physical system.

It is also one of the costs of the shortage.

The government's six-to-eight-week assurance is useful — but it is not a published model

Lescure said Monday that France's strategic stocks were full, that the government had "very clear visibility" over the next six to eight weeks, and that it had no immediate supply concern over roughly the next two months.

That information should be reported.

It should not be confused with a disclosed stock-and-flow forecast.

The government has not published, alongside the statement, a reproducible calculation showing refinery throughput, expected imports, consumption, terminal stocks and inventory draws for each of those six to eight weeks.

It is therefore best treated as an official operational assurance, not an independently testable estimate.

More significantly, Lescure also acknowledged the global issue.

He said the world is drawing from stocks and warned that if the disruption continues, an eventual supply problem is possible.

That distinction is broadly consistent with the evidence.

France does not have a demonstrated national shortage today.

The international system replacing the barrels France and Europe consume has less room for error.

The French tripwire is now measurable

The national shortage headline is no longer the best indicator.

April gives us a control case.

At the April peak:

18% nationally affected, 83% Total, implied non-Total disruption ~4.6%.

Today:

17% nationally affected, 91% Total, implied non-Total disruption ~2.3%.

Even after varying the rounded Total shares by two percentage points in either direction, the implied ranges remain separate.

That gives EuroOilWatch a quantitative threshold to monitor.

If the national shortage percentage rises but the non-Total estimate remains around 2%-3%, while roughly nine in ten affected stations remain Total, the price-cap/logistics explanation continues to fit.

If the national percentage rises because non-Total disruption moves back through 4%, 5%, 6% and higher, while Total's share of affected stations falls, the interpretation changes.

That would be evidence that the disruption is escaping the discounted network.

The second test is the one only a station-level tracker can perform properly:

Total's local discount versus confirmed rupture probability.

Not Total versus a national average.

Total versus the nearest comparable non-Total stations.

Day by day.

If a larger local discount reliably predicts a higher probability of rupture, the government's explanation is quantitatively supported.

If stations rupture independently of that spread, the supply story becomes more interesting.

That is the next OilWatch analysis.

France is not the shortage. France is the stress test.

The French forecourt event itself is less alarming than the headline suggests.

September is reproducing April's retail pattern at almost exactly the same implied penetration of Total's network, while the rest of the French station system appears to be performing materially better.

But the market behind those stations is different.

Global diesel exports averaged 5.85 mb/d in August, 25% below a year earlier.

Russian diesel availability has been reduced by both refinery damage and policy restrictions.

ARA gasoil inventories reached a four-year low during the summer.

US distillate inventories fell 14% below their five-year seasonal average despite refinery utilisation above 97%.

Europe faces a further 510,000 bpd jet-fuel deficit in the fourth quarter.

S&P Global has increased its diesel-crack forecast by $31/bbl, to around $84 for the rest of 2026.

And perhaps most importantly, North Sea Dated crude was around 24% cheaper in August than in April while diesel refining margins moved in the opposite direction.

That divergence tells us where the constraint has moved.

The global system has become increasingly successful at finding crude oil and finding ways to move it.

It has not restored the spare refining capacity required to turn enough of that crude into diesel, jet fuel and other middle distillates.

April showed what happens when a large retail discount overwhelms one company's delivery network.

September is running the same experiment against a global system with fewer export barrels, thinner seasonal inventories, extraordinarily expensive shipping and very little idle refining capacity.

The French stations are not yet telling us that France is short of diesel.

They are showing us where the failure will appear first if Europe can no longer replace what it consumes.

France is not the shortage. France is the stress test.