ArcelorMittal's Liberia bet nears $4 billion: what a 20 Mt export target really tests

ArcelorMittal Liberia's cumulative spending in the country is "approaching $4 billion," CEO Michiel van der Merwe told the Liberia International Mining and Energy Conference in late October, with exports projected to exceed 20 Mt in 2027 against under 5 Mt in 2024. The headline number is striking, but the strategically important claim is quieter: Phase II includes Liberia's first large-scale iron ore concentrator, producing premium concentrate for steelmakers, including newer low-carbon plants.

That is a change in what Liberia sells, not just how much. It also puts the country's single 243 km-class rail corridor to Buchanan under a load it has never carried — at the same moment the government is designing a multi-user access regime and handing operations to an independent company from 2030.

Caveat on sourcing: the figures below come from a single trade-press report of a conference speech. They are not corroborated here by an ArcelorMittal release, the amended Mineral Development Agreement (MDA) text or Liberian government documents, and should be read as reported rather than verified.

The investment arithmetic

Two derived points are worth flagging. First, the roughly $0.3–0.5bn added between January and October 2026 implies peak construction spend on the concentrator and associated works — consistent with a project moving from build to commissioning (analysis). Second, rail and port account for roughly 20–23% of cumulative investment. That is high for a mining project, and it reflects the fact that ArcelorMittal rebuilt government-owned infrastructure rather than owning it.

The export trajectory is the demanding part. Moving from under 5 Mt to over 20 Mt in three years implies annual growth of roughly 60% on an illustrative 4.5 Mt to 20.5 Mt basis — a step change rather than an incremental ramp. Note also that 20 Mt is an export figure while 30 Mt is mine production; the two metrics are not directly comparable and the 30 Mt has no stated date. Treat it as a later-stage ambition, not a 2027-era number.

Why the concentrator matters more than the tonnage

Liberia's traditional export product has been direct-shipping ore — lump and fines shipped largely as mined. A concentrator changes the product's chemistry and its customer base. Beneficiation raises iron grade and, critically, strips out silica and alumina. Lower alumina is what allows a burden to be used in direct-reduction grades of pellet feed and in blast furnaces running at lower slag volumes and lower carbon intensity per tonne of hot metal.

The CEO's framing — supplying existing mills and newer plants built to cut emissions — is the commercial logic. Atlantic-basin DR-grade feedstock is structurally short as European DRI projects move from announcement to construction, and a West African concentrate has a short sea haul to European and Mediterranean customers relative to Brazilian or Australian supply.

The gap in the public record is material: no Fe grade, alumina or silica specification, no concentrator capacity, no capex line item and no commissioning date have been disclosed in this reporting. Without those, the claim of "premium concentrate" cannot be tested against the pellet-feed benchmarks that actually determine its price realisations.

The corridor is the bottleneck — and the policy question

The Tokadeh–Buchanan railway is the constraint. Moving more than 20 Mt/y down one corridor is a throughput problem before it is a mining problem, and the ~$800m already spent is the enabling investment. Three things now determine whether the 2027 target holds:

1. Ramp-up of rail capacity. The reported plan rests on rehabilitation and expansion of a government-owned line; rolling stock, signalling and maintenance regimens are the usual failure points in West African iron ore logistics.
2. The 2030 operator transition. An independent company is expected to run the railway from 2030 under operating principles still being formalised. Handing a corridor over mid-ramp introduces execution risk precisely when volumes peak.
3. Third-party access. Van der Merwe reaffirmed support for multiple users, with the Liberian government deciding access and other miners required to finance any incremental capacity they need, while spare capacity can be used meanwhile.

That last point is the most consequential for the region. A credible multi-user regime on Tokadeh–Buchanan would make rail access a tradeable asset and could unlock deposits in the Nimba range and neighbouring Guinea that are otherwise stranded. The risk is equally well known in West African corridors: capacity allocation disputes, and the temptation for the incumbent to define "spare capacity" in ways that leave little for others. The distinction between "supporting multi-user access" and "financing only your own capacity" is where the argument will be had.

What is not being said

The public record leaves several gaps that matter to anyone modelling this supply:

• Reserves and mine life. The amended MDA extends operations to 2050 with a 25-year renewal right — a long horizon that presupposes orebody support not quantified here.
• Product mix and offtake. Whether concentrate is sold as fines, pellet feed or converted to pellets, and to whom, determines realised premiums and freight economics.
• The 30 Mt case. The demand assumption behind a further expansion study is unstated, in a seaborne market where the marginal tonne is increasingly contested.

The wider read

For steelmakers, ArcelorMittal Liberia's value is optionality: a non-Australian, non-Brazilian Atlantic supply of higher-grade feedstock arriving as carbon-cost differentials sharpen in Europe. For Liberia, the value is a concentrator that keeps more of the beneficiation margin in-country and a rail asset that, if genuinely opened, could serve more than one balance sheet. The $4 billion figure is a useful measure of commitment. The 20 Mt target and the 2030 rail handover are the tests that will show whether the commitment converts into a reliable, contestable supply chain.