Germany's September Steel Slump Was a Blast-Furnace Story, Not a Demand Signal — But the 35.4 Mt Run-Rate Is Fragile

German crude steel production dropped 13.0% year on year to 2.627 Mt in September, interrupting what had been a modest recovery. WV Stahl used the release to cut its annualised production level for 2026 to 35.4 Mt, down from the 37.3 Mt implied by first-half data — a reduction of 1.9 Mt, or 5.1%.

Read as a headline, that looks like a demand downgrade. The underlying route data say something narrower and more technical.

The decline is a single-route event

The federation's own table separates the two production routes, and they moved in opposite directions:

Source: WV Stahl, 22 October 2026.

The arithmetic is unambiguous. Backing out last year's base (2,627 / 0.870 ≈ 3,020 kt), Germany lost roughly 390 kt of monthly crude steel year on year. Oxygen steel accounts for about 430 kt of that loss; electric steel added roughly 38 kt. Essentially all of the September decline came from the converter route.

The pig iron number confirms the mechanism. Pig iron intensity fell to 59.3% of crude steel in September (1,559/2,627) against roughly 66.7% a year earlier — a fall of about seven percentage points. Steelmaking did not shrink; hot metal supply did.

Schwelgern 2 explains most of it

The proximate cause is documented. thyssenkrupp Steel took blast furnace Schwelgern 2 in Duisburg — Europe's largest, with roughly 12,000 t/day of hot metal capacity and a 14.9 m hearth diameter — out of service from 30 August 2026 for a planned partial relining of about six weeks, covering cooling systems, refractory, gas cleaning and slag granulation. A furnace of that scale, offline for a full month, removes on the order of 350–370 kt of hot metal — closely matching the observed pig iron shortfall.

That framing matters for how the number should be read. September is a maintenance artefact, not evidence of a fresh collapse in order intake. The distinction is not academic: the same headline, if attributed to demand, would imply a much darker Q4 than the route split supports.

What the 35.4 Mt figure actually is

WV Stahl's "forecast cut" is a run-rate annualisation, not a demand forecast. The number is mechanical: nine months at 26.529 Mt, extrapolated. What is worth tracking is the direction of travel of that extrapolation:

Two consecutive downgrades. Against a 2025 base of 34.1 Mt — the lowest since 2009 — the new figure still implies growth of about 3.8%, so the recovery narrative survives, but marginally.

The figure also embeds an assumption worth making explicit. Hitting 35.4 Mt requires Q4 output of 8.87 Mt, or roughly 2.96 Mt per month, against 2.63 Mt in September. That is achievable once Schwelgern 2 is back and the seasonal maintenance window closes, but it is not a floor. Should September's rate persist through the quarter (2.63 × 3 = 7.89 Mt), the year would land near 34.4 Mt — barely above 2025. This is my calculation, not WV Stahl's.

One further cross-check: hot rolled output (2,322 kt) fell less than crude steel (−10.6% vs −13.0%), and the HR/crude ratio held at 88.4% for the month and 87.9% for the nine months. Rolling operations did not follow the blast furnaces down, implying either a drawdown of semi-finished inventory or a shift in product mix. Either way, downstream capacity was not the constraint in September.

Better than Europe, on a weak base

The comparison with the wider EU is instructive. worldsteel independently confirms Germany at 2.6 Mt (−13.0%) for September, while EU27 output was 10.1 Mt (−5.0%) and 94.5 Mt for January–September (−0.8%). Germany's nine-month growth of 4.0% is therefore well ahead of the bloc — but from a deeper trough. Germany's 2025 collapse was sharper than the EU average, so the base effect flatters the comparison. In absolute terms the country is still running far below its pre-2022 norm.

A note on the data

One circulating version of the September figures — a summary published by the recycling association BVSE, credited to WV Stahl — reports an 11.3% total decline and an 18.5% fall in oxygen steel. Those figures conflict with WV Stahl's own table and with worldsteel's independent −13.0%. No amendment to the WV Stahl release has been issued. The −13.0% and −20.2% figures should be treated as the reliable ones; the BVSE numbers appear to be a transcription error. This is my assessment, not a documented correction.

What to watch

The question for Q4 is not whether German steel demand has deteriorated — September gives little evidence of that — but whether the BOF route returns to full availability and whether electric steel's steady 4%-plus growth is enough to offset structural attrition in the integrated route. The scheduled end of thyssenkrupp's hot metal supply to ArcelorMittal Duisburg in September 2027, with the closure of that site's steelmaking and billet rolling operations, points to further permanent subtraction from the oxygen-steel column next year. Germany's recovery, such as it is, is increasingly an electric-route story.