Europe's cotton-linter problem and the 155mm production target

The conversation about European artillery output has, for two years, been about shells. The shortage Ukraine is fighting is real; the EU's commitment to 1.7 million 155mm rounds per year by 2025–2026 is real; the new shell lines being built in Germany, France, Italy, Spain and the Czech Republic are real. None of this matters if the upstream raw material is not secured. And the upstream raw material is, today, neither European nor secure.

The bottleneck is cotton linters: the short fibres left on the cottonseed after the first cut of lint has been removed at the gin. Second-cut linters, processed through acid delinting, are the cellulose feedstock for the nitrocellulose that goes into modern artillery propellant. The chemistry has not materially changed in a century. The chemistry is also unforgiving: only second-cut linters of a defined purity and fibre length work for defence-grade nitrocellulose. Substitutes — wood pulp, recycled cellulose — have been promised for decades and have never reached the scale or specification that NATO propellant lines require.

Europe today imports roughly 70 per cent of its cotton-linter supply from China. Rheinmetall's chief executive has said this on the record; it has been published in the Financial Times. Then, in early 2024, Chinese deliveries to European nitrocellulose producers paused, with no formal explanation given. In the same window, Chinese nitrocellulose exports to Russia rose from zero in 2015–2021 to USD 3.25 million in 2022 and USD 6.45 million in 2023. The two facts are not necessarily connected. They do not have to be connected for the strategic reading to be the same: the input to European artillery propellant is sourced from a country that simultaneously supplies the adversary that the artillery is being produced against, and that has demonstrated the capacity to interrupt supply at short notice.

The downstream is in place; the upstream is not

The European nitrocellulose stack is not, structurally, the problem. Rheinmetall operates four plants across Germany, Switzerland, Spain and South Africa; Eurenco runs the MANUCO joint venture in Bergerac; Nitrochemie supplies EU and NATO from Switzerland. Estimated installed European nitrocellulose capacity is 4,500 to 10,000 tonnes per year. Total European nitrocellulose demand, at the 1.7 million-rounds 155mm target sustained through 2030, is in the range of 19,000 to 20,000 tonnes per year. The arithmetic is straightforward: even at full utilisation, European nitrocellulose producers cannot meet European propellant demand without imported linters.

How much linter? Roughly 13,000 to 15,000 tonnes per year, against current European secure supply of substantially less. Translated upward through the cotton chain — a conversion ratio of one tonne of second-cut linters to twenty-eight or forty-two tonnes of raw seed cotton, depending on variety — that requirement implies raw cotton in the range of 400,000 to 600,000 tonnes per year, and harvested area in the range of 90,000 to 180,000 hectares.

This is the missing number in European industrial-base discussions. It is large enough to matter. It is also not so large that it cannot be addressed.

Where the conversation usually stops

Most public discussion of the European defence industrial base treats raw materials as an afterthought of procurement. Buy more shells; pay more for them; ringfence the line items in the budget; trust the contractor to source upstream. That has been the operating assumption for the last decade and arguably the last three. It produced an industrial base that can absorb a sudden tripling of orders only by drawing more heavily on the supply chain that runs through China.

A bank financing a downstream shell-plant expansion will, in 2026, ask the contractor for evidence of secured raw-material supply through the term of the loan. The contractor will, in many cases, not have a defensible answer beyond multi-year linter purchase agreements with the same Chinese or Turkish suppliers. That answer is increasingly insufficient for tier-one European banks under their internal industrial-base risk frameworks; it will become more so as the European Defence Industrial Strategy and EDIRPA-style funding mechanisms require demonstrated supply-chain resilience as a condition of co-investment.

The financing community is already adjusting. The European Banking Federation's 2025 paper on the EU defence industrial base is unusually specific about the upstream conditions banks need to see before they will fund new capacity at scale. Bankability, in that document, is not a function of the shell margin. It is a function of whether the raw materials behind the propellant behind the shell are sourceable from jurisdictions the bank's risk committee can defend.

What a credible answer looks like

A credible answer to the cotton-linter question is not "Europe will substitute wood pulp." It is not "we will buy from Turkey instead." Turkey is a serious linter producer and is part of any sensible diversification, but it is not a sovereign solution; it is a different non-European dependency. The credible answer is partial European production of defence-grade linters within the EU's own jurisdiction, sourced from EU-cultivated cotton, processed in EU-owned ginning and delinting facilities, and contracted under multi-year offtake with European nitrocellulose producers.

Iberia is the only place in the EU where this is currently feasible at scale within a useful timeframe. Spain already cultivates 52,000 hectares of cotton, supported by eight operating gins and established irrigation districts. Portugal has the agronomy and a separate water basin. Targeting 10 to 20 per cent of European linter requirement — between 1,300 and 2,600 tonnes of defence-grade linters per year — implies roughly 8,000 to 31,000 hectares of cotton dedicated to the programme. That is a meaningful share of existing Iberian capacity but not a stretch beyond it, and it leaves an integrated facility with a clear path to commercial viability through commercial-grade linters, lint, and cottonseed by-products alongside the defence offtake.

It also leaves the politics in a useful place. A neutral upstream supplier serving every NATO-aligned nitrocellulose producer — Rheinmetall, Eurenco, Nitrochemie — avoids competition with the primes and gives finance ministries a single addressable project rather than a constellation of bilateral procurement contracts.

What this means for the next eighteen months

The 155mm scale-up is moving forward whether or not the linter question is resolved. Shell capacity will be built. Nitrocellulose lines will be expanded. At some point in the next twelve to twenty-four months, the difference between announced production capacity and actual deliverable rounds will widen, and somebody will be asked why. The answer will be that the raw material gap was visible from the beginning and that the EU chose, by default, to keep relying on the supplier that had paused deliveries once already.

The opportunity is to make the alternative answer available before the question is asked. That is industrial infrastructure work — agronomy, ginning, delinting, offtake, finance — not procurement work. It is the kind of mandate that requires sovereign sponsorship, ECA-backed financing, and disciplined project structuring. It is the work the European defence industrial base needs done.