The groundbreaking at Osceola on 30 October 2026 was the ceremonial milestone; the substantive decision had already been taken months earlier. U.S. Steel announced the $1.9 billion investment on 29 April 2026, and Arkansas regulators had received the wastewater application for the DRI unit in November 2025. Trade press coverage following the ceremony (SteelOrbis, 16 November) captured the strategy narrative but omitted the two numbers that matter most: 2.5 million tonnes a year of capacity, and Midrex as the technology supplier.
What is actually being built
| Parameter | Detail | Source type |
|---|---|---|
| Capex | $1.9 billion | Company, repeated by Midrex and Arkansas AEDC |
| Capacity | 2.5 Mt/yr HDRI/HBI, single shaft furnace | Midrex / Kobe Steel releases, June 2026 |
| Technology | MIDREX Flex, natural-gas based, designed to accept hydrogen addition | Midrex |
| Feedstock | DR-grade pellets from Keetac, Minnesota | Company |
| Product route | Hot DRI charged directly to adjacent Big River 2 EAFs; HBI for internal use or merchant sale | Company (DRI direct); Midrex (explicit merchant/export intent) |
| First production | 2029 (AIST reports first-half 2029; company has not been confirmed on the month) | Midrex; AIST |
| Employment | ~200 permanent, plus ~35 embedded contractor roles; up to 2,000 construction jobs at peak | Company (April and October releases) |
| Existing campus | Four EAFs, >6 Mt/yr steel | Company, SteelOrbis |
The distinction between "first of its kind in the United States" and "first of its scale" is worth keeping straight. The U.S. already has DRI/HBI capacity — Nucor's Louisiana and Trinidad plants, Cleveland-Cliffs' Toledo HBI joint venture — but nothing of this size tied hot, without reheating or briquetting, into an adjacent EAF shop. The design choice is the point: hot charging bypasses the cooling, briquetting and remelting steps that erode the metallics value of conventional HBI.
Why it matters: the metallics arithmetic
At 2.5 Mt/yr and a typical 90–94% metallization, the plant yields roughly 2.25–2.35 Mt of contained iron units a year. Against a 6+ Mt/yr EAF campus, that is a meaningful but partial substitution — in the order of a third or more of Big River's metallic charge, depending on product mix and scrap quality requirements.
The strategic logic is less about displacing all scrap than about diluting the part of the scrap basket that constrains grade capability. Big River's pitch to automotive, appliance and solar customers rests on surface quality and residual control. Prime scrap has become the scarce, volatile and increasingly expensive input in that equation; DRI is the lever that lets an EAF producer hold residuals down without paying the prime-scrap premium through the cycle.
The integration with Keetac is the part that differentiates this from a merchant DRI purchase strategy. U.S. Steel spent roughly $150 million (completed May 2024) adding DR-grade pellet capability in Minnesota, explicitly framed at the time as feedstock for its own potential DRI/HBI plant — or for sale to third parties. The Arkansas plant converts that optionality into captive demand. That is a genuine mine-to-mill chain, and it is what Nippon Steel's ~$11 billion U.S. investment commitment is partly buying.
The uncomfortable questions
Gas and power economics. A 2.5 Mt/yr Midrex Flex plant is a large, continuous industrial gas consumer. The "hydrogen-ready" label is real engineering — the Flex configuration can accept rising H2 shares — but in Arkansas in 2029 the operating reality will be natural gas. The project's carbon intensity will be set by the gas supply and by whether any low-carbon hydrogen is available at commercial scale and price by the time the plant starts. The company's framing — "supporting lower-emission steel production" — should be read as relative to blast-furnace virgin iron, not as near-zero.
Pellet balance. U.S. Steel's Minnesota operations have DR-grade capability in the region of 4 Mt/yr, so the Arkansas plant's needs are broadly covered. But that assumes Keetac runs at rate and that no competing internal or external claim on those pellets emerges. Any interruption in Minnesota — logistics, mine plan, winter disruption on the Great Lakes and rail — propagates directly into Arkansas EAF costs.
HBI: flexibility or hedge? The company describes HBI as capable of being used at Big River "or other destinations"; Midrex states plainly that the HBI is for "merchant sales and export to steelmaking facilities worldwide." That divergence in emphasis is not accidental. HBI capability gives U.S. Steel a way to run the DRI plant at full rate even when EAF demand for hot DRI is soft, and a potential export revenue line. It also means a new 2.5 Mt-scale HBI supplier could enter the Atlantic and Gulf merchant metallics market — a development that would interest every EAF producer currently buying imported HBI, and every scrap processor watching prime grades get squeezed.
Timing risk. Groundbreaking in late 2026 with first production in 2029 implies roughly 30 months of construction. That is plausible for a single-shaft Midrex module but tight. The permitting record is notably more advanced than the construction schedule — air permit 2445-AOP-R5 was issued in April 2026 as a significant modification to the existing Title V permit, and the renewal (R6) was finalised in late October, still listed as in review — consistent with a project that is shovel-ready but not yet in heavy construction.
Competitive read-across
The project is a template as much as a plant. It demonstrates that the U.S. can assemble the full chain — DR-grade pellets, gas-based shaft furnace, hot charging into EAFs, merchant HBI optionality — without imported metallics. For EAF producers without captive iron ore, the implication is the opposite: the scrap-quality premium is likely to persist where they cannot replicate this integration. For scrap processors, prime-grade demand from the integrated EAF producers may soften over time as hot DRI displaces the top of the basket, even as total scrap consumption in the U.S. continues to grow.
The 2029 start date will be the test. Until then, the most informative signals are not press releases but gas contracts, Keetac pellet output and whether any HBI offtake agreements emerge before the plant is commissioned.
References
- SteelOrbis, "US Steel breaks ground on $1.9 billion DRI plant in Arkansas," 16 November 2026.
- U.S. Steel, "U. S. Steel Breaks Ground on $1.9 Billion Arkansas Direct Reduced Iron Production Plant," 30 October 2026.
- U.S. Steel, "U. S. Steel Announces First-of-its-Kind in the United States DRI Facility at Big River Steel Works," 29 April 2026.
- Midrex Technologies, "Midrex Selected for U. S. Steel's First-of-its-Kind DRI Facility at Big River Steel Works," June 2026.
- Kobe Steel, release confirming 2.5 Mt/yr MIDREX Flex plant and 2029 production start.
- Arkansas Economic Development Commission, groundbreaking announcement, 30 October 2026.
- Arkansas DEQ, Air Permit 2445-AOP-R5 (AFIN 47-01073) and renewal R6; NPDES Construction Application AR0053384C1 (November–December 2025).
- U.S. Steel, 2024 Sustainability Report and Keetac DR-grade pellet announcements.
Source & methodology +
MetalMate separates original-source evidence from summary and analysis. Verify the original source before procurement, legal or engineering decisions.
- Article status
- Published · Version 1
- Source date
- 16 November 2026
- MetalMate publication
- 16 November 2026
- Verification
- Automated evidence checks; not human reviewed
US Steel breaks ground on $1.9 billion DRI plant in Arkansas — SteelOrbis

