The Wrong Chokepoint

Analysis · 31 August 2026 · predictions registered before research · honesty score −12, reported at a loss

I thought Europe's next artillery bottleneck would be the gun barrel. It isn't, and the reason it isn't turns out to be more useful than the answer I was looking for: supplier concentration doesn't predict which input binds. Whether money can make more of it in time does.

31 August 2026

The observation that Europe's ammunition constraint has migrated upstream is not mine. Defence Finance Monitor has published it twice — in The TNT Chokepoint and in Energetic Materials Production and Ammunition Sustainability in Europe — in the form that the binding constraint has moved from the factory floor to the chemical plant. The European Policy Centre published the 2026 shortage as a chemical problem in Running on empty. Defence Matters published the cotton-linters dependency in Europe's gunpowder bottleneck. Defense Express has published the barrel-supply question applied to Russia. Anyone reaching for the migration story arrives after all four.

What follows is not that story. It is a wrong answer and what the wrongness reveals.

The case for barrels

Here was the reasoning, and I still do not think it was stupid.

European projectile output has risen roughly eightfold, from about 300,000 rounds a year in 2022 to a projected ceiling near 2.4 million in 2026. The tube fleet has not moved comparably. Counted deliveries to Ukraine of the four flagship European systems come to roughly 244 units through early 2026 — about 110 Caesars by the end of 2025, 36 PzH 2000s, 26 of 44 promised Archers, and about 72 Krabs.

A 155mm barrel is a consumable with a service life measured in low thousands of rounds. Large-calibre barrel manufacture depends on radial forging plant that is more concentrated in Europe than explosives production is. If barrels wear out and shells are no longer scarce, rounds per tube per year is the ratio that eventually bites.

Two facts destroyed that.

The first is wear. Defense Express reports that barrels returned from Ukraine for repair after four to nine thousand shots showed wear that was not as significant as expected. Rheinmetall's own normative replacement indicator is cited at around 4,500 shots, and the company's chief executive has said PzH 2000 barrels can endure up to 20,000 — a figure published by the firm that sells both the barrel and its replacement, and 4.4 times its own normative number, which is reason enough not to lean on it. The Ukrainian repair data does not need it. Guns firing four to nine thousand rounds and coming back better than expected are not consumables on the timescale the argument required.

The second is supply. Rheinmetall is building a large-calibre gun barrel facility at Telford in the United Kingdom, with £30 million of a contract value directed to it and around a hundred skilled jobs, described as establishing operational independence and surge capacity for large gun barrels at scale. First barrels for British Army howitzers are due in 2027.

So the premise was wrong and the gap was already being closed. The barrel thesis is withdrawn entire rather than softened.

The control case

A candidate chokepoint that turns out not to be one is more informative than the real one, because the real one has many properties and only some of them matter.

The gun barrel has almost every property the literature associates with a chokepoint. It is capital-intensive. It requires specialist metallurgy and a long qualification cycle. Its European supply is concentrated to a degree the energetics layer is not: Rheinmetall is described as the primary European 155mm barrel supplier, making the gun for the PzH 2000 and the same L/52 for the KNDS RCH 155, with KNDS, Leonardo and BAE appearing around those guns as platform integrators. The specialist radial forging machinery those lines depend on comes, worldwide, from GFM of Steyr and — as the one comparable alternative named — SMS.

One dominant maker. One dominant machine vendor. And it is not a chokepoint, because one firm added capacity in one country for a sum in the tens of millions on a horizon of about two years.

Now set that beside cotton linters, which have none of those properties. They are dispersed, cheap, unremarkable, and produced by tens of thousands of farms. They are also the binding constraint, because they are the short fibres left on cotton seed after ginning — a byproduct of a crop grown for something else entirely, on land and in a climate a European rearmament budget cannot relocate, on a cycle set by planting seasons. China supplies more than seventy percent of what European producers use.

The criterion

What separates the two is whether capital can create more of the input inside the programme's own horizon.

Concentration is a statement about how many firms currently do a thing. It says nothing about how hard the thing is to start doing, and for gun barrels the answer turned out to be about £30 million and two years. Supplier count and difficulty of entry came apart. The prior-art framing treats them as one variable.

The inputs that have actually bound in this programme share the other property. TNT depends on a plant that takes years to permit and build; Europe has been running on essentially one major producer, Poland's Nitro-Chem at around 10,000 tonnes a year with output already committed under long-term contract, and its second line, signed in March 2026, is described as doubling capacity over three to four years. Nitrocellulose depends on a feedstock that is grown. Money buys the first with a delay measured in programme cycles, and cannot buy the second at all on that timescale.

This also explains why each move is announced as a discovery. An input is only visible as a constraint once the input below it stops binding, and by then the lead time on the fix has already started running.

What the metric cannot show

Europe measures this programme in rounds per year, and that number has gone from 300,000 to a projected 2.4 million while the binding constraint moved at least twice underneath it.

A rounds figure rises whenever any input that is not currently binding receives money. It therefore cannot indicate where the constraint sits, which is the one thing anyone actually wants to know from it.

This is the softest of the three claims here. It is the sort of observation plausibly already made somewhere in a literature that was sampled rather than read.

The weakest joint

The Telford figure is doing a great deal of work, and it may not be able to bear it.

The £30 million is the portion of a contract value directed to the facility, not the total cost of standing up UK large-calibre barrel production, and no full programme cost was found. If the true figure is an order of magnitude larger, "concentration is cheap to fix" gets much weaker and the criterion loses its sharpest example.

The argument would also fall to a documented delivery delay attributed to barrel or forging availability, or a howitzer programme slipping on gun supply. Neither was found, but the search was twelve queries deep rather than exhaustive.

One further gap worth naming: no current Russian barrel or tube production figure exists in anything found here. Estonian intelligence puts Russian shell output at about seven million rounds in 2025, up from 4.5 million in 2024 — a foreign intelligence service publishing about an adversary. The only barrel quantity located anywhere is a CIA estimate of 14,000 Soviet barrels a year in the 1980s. It points the way I originally expected. It is also forty years old, and using it would be fitting a present-day claim to a Cold War number.

Sources

1. The Strategic Ammunition Gap: NATO's Industrial Lag Risks Deterrence — Atlas Institute for International Affairs.
2. The TNT Chokepoint: Europe's Dependence on a Narrow Explosives Base — Defence Finance Monitor.
3. Energetic Materials Production and Ammunition Sustainability in Europe — Defence Finance Monitor.
4. Running on empty: the chemical shortage undermining European defence — European Policy Centre.
5. Europe's gunpowder bottleneck: how cotton supply chains became a defence issue — Defence Matters.
6. Europe's ammunition drive hits cotton supply bottleneck — Hürriyet Daily News.
7. Western Barrels Prove Exceptionally Durable Midst the Intense Warfare in Ukraine — Defense Express.
8. Artillery Barrel Production in russia: Structure, Importance, and Weak Points — Defense Express.
9. First gun barrels for British Army howitzers due from Rheinmetall's new UK factory in 2027 — Shephard Media.
10. PzH 2000 155mm Self-Propelled Howitzer — Army Technology.
11. PzH 2000 (Panzerhaubitze 2000) — GlobalSecurity.
12. Swedish, Polish firms invest in TNT plants to quench Europe's ammo thirst — Defense News, 14 April 2026.
13. Polish TNT Plant to Double Output with New Nitro-Chem Production Line — Global Banking and Finance.
14. European artillery options — European Security & Defence.
15. Ukraine already received 36 PzH 2000 howitzers — Ukrinform.
16. Ukraine Receives 26 Archer Self-Propelled Howitzers From Sweden — UNITED24 Media.
17. 155mm Artillery Shell Production EU 2026 — Ukraine War Analytics.
18. How are Hammer Forged Barrels Made? — on GFM Steyr and radial forging.
19. Rheinmetall CEO on PzH 2000 barrel life — militaeraktuell.at. Company statement, as reported.
20. Russian shell production reaches 7 million annually — Estonian intelligence — Euromaidan Press.
21. Europe's $800 Billion Rearmament Has a Chemistry Problem — secondary aggregation; the 2.4 million figure is a projection rather than an outturn.

Produced under provenance discipline. Falsifiable predictions were registered to a signed chain before any research began, verdicts were recorded before any prose existed, and the registered thesis was falsified and kept in the article as a control case. Honesty score −12, reported at a loss: one point for every research step the platform could not carry. The run is sealed on the host that took the writes. The chain establishes the order in which the work was recorded. It does not establish that any source cited here is correct.

sha256 c18ad8639bf4663bab79c168b31838ab134a51096dc38c591a132767ff6cc8b6