From Reserve Size to Processing Capability
ABSTRACT
Network power in supply chains concentrates where substitution is hardest, which in rare earths is separation rather than extraction. This paper applies that logic to Türkiye’s Beylikova project, specifies a four-step sequence from resource certification to downstream integration, and identifies financing partner selection as the principal risk to the strategy’s purpose.
1. Where power sits in a supply chain
Farrell and Newman’s account of weaponised interdependence establishes that coercive capacity in economic networks concentrates at nodes through which flows must pass and for which substitution is costly. The insight transfers directly to material supply chains, with one specification: the chokepoint is not the stage with the largest physical volume but the stage with the fewest viable alternatives.
In rare earths that stage is separation. China accounts for approximately 70 percent of global mine production, which is a large share but not a decisive one, since rare earth ores are geologically common and deposits exist across many jurisdictions. It accounts for 85 to 90 percent of separation and refining capacity, and controls an estimated 80 to 90 percent of the solvent extraction process technology required to separate individual oxides from one another. That is where substitution is genuinely difficult, because the capability is embedded in process knowledge, operational experience and engineering workforce rather than in equipment that can be purchased.
A strategy organised around reserve size therefore targets the wrong variable. Ore in the ground without separation capability produces a mixed concentrate that must be sold to entities holding the constrained capability, which reproduces the dependency at a lower margin.
2. Türkiye’s position
The Beylikova deposit in Eskişehir is reported by the Ministry of Energy and Natural Resources and Eti Maden at 694 million tonnes of ore containing ten of the seventeen rare earth elements together with fluorite, barite and thorium. The government describes it as the second largest known reserve after Bayan Obo. The figure is state-sourced and has not been certified to an internationally recognised mineral resource reporting standard.
Operational reality is at a different scale. A pilot facility has run since 2023 processing 1,200 tonnes of ore annually. The announced industrial complex carries a 600 million dollar budget targeting 570,000 tonnes of ore per year and roughly 10,000 tonnes of rare earth oxides, with design in 2026 and construction from 2027 — a throughput increase of approximately 475 times from the pilot.
Between the reserve and the announced output sits the process technology. A 2024 memorandum of understanding with China stalled over China’s refusal to transfer separation technology, after which Turkish engagement extended toward the United States, Canada and Switzerland. The energy minister has acknowledged the gap publicly while stating that the country is progressing toward industrial-scale processing including separation.
3. Step one: certify the resource
The first step is publication of a resource statement prepared to an internationally recognised reporting standard by a competent person.
The reason is that certification is the entry condition for every subsequent step. Project financiers require it because it is the basis on which reserves can be treated as collateral or as revenue security. Offtake counterparties require it because volume commitments depend on it. Technology partners require it because their return depends on plant utilisation over decades. A state-sourced tonnage figure, however accurate, does not perform these functions, because the standard exists precisely to remove the issuer’s discretion from the estimate.
Certification also produces information the state itself needs. Ore tonnage is a weak indicator of value: grade, the light-to-heavy rare earth distribution, and the metallurgical complexity of separating this particular ore body determine whether the project is commercially viable at prevailing prices. Those parameters emerge from the certification process.
4. Step two: acquire separation capability
The second step is securing solvent extraction capability through licensing or partnership with a party that does not reintroduce the dependency the project exists to reduce.
Three routes exist and each has a characteristic failure mode. Licensing from a Western or allied operator is the preferred route and is constrained by the small number of operators outside China and by their own commercial interest in not creating competitors. Indigenous development through domestic research is slower and more expensive, and the roughly 60 researchers currently at the site indicate a serious but small programme relative to the task. Partnership with the dominant incumbent resolves the technology question and defeats the strategy’s purpose, which is the risk addressed in section seven.
The Turkish–German Energy and Minerals Partnership signed on 19 June 2026 and the mining cooperation agreement with Uzbekistan of 29 January 2026 establish diplomatic architecture without naming a separation partner. The architecture is a precondition rather than a solution.
5. Step three: bind demand before committing capital
The third step is securing offtake agreements before final investment decision.
Rare earth oxide markets are thin and price-volatile, and the dominant supplier has both the capacity and the demonstrated willingness to influence prices. A new producer entering without contracted demand is exposed to a price path that a competitor can move, which is the mechanism by which several non-Chinese rare earth ventures have failed commercially despite technical success.
Türkiye’s position on this step is stronger than it may appear, because two domestic demand centres exist. The defence and aerospace sector exported 10.05 billion dollars in 2025 and signed 17.8 billion in new contracts, with magnet-dependent product lines — guided munitions, unmanned systems, electronic warfare, sonar — among its fastest-growing. The automotive sector is transitioning toward electric and hybrid production, with electric vehicles at 17 percent of new car sales in 2025 and a fleet projected to exceed 5 million by 2035. Domestic demand does not fill a 10,000 tonne plant, but it anchors a portion of it, and an anchor materially improves the terms available for the remainder.
6. Step four: integrate downstream
The fourth step is integration into permanent magnet production rather than stopping at oxide.
The value distribution in the chain runs from mining through separation and metal and alloy production to magnet manufacture, with margin concentrated toward the downstream end and, more importantly, with strategic leverage located where the buyer’s substitution options are fewest. A magnet producer supplying qualified components into defence and automotive programmes occupies a position that is costly to replace because qualification is expensive; an oxide producer does not.
This step also determines whether the project contributes to the conversion problem identified across Turkish industrial strategy generally. Oxide exports are a commodity revenue stream. Magnet production is an industrial capability that raises the technology content of the manufacturing base and feeds directly into two sectors already identified as strategic.
7. The principal risk
Financing partner selection is where the strategy is most likely to fail, and the failure would be invisible at the moment it occurred.
Scaling from a 1,200 tonne pilot to a 570,000 tonne industrial complex requires capital and technology that Türkiye does not currently hold. The party best positioned to supply both is the dominant incumbent. An arrangement in which Chinese capital and Chinese process technology build a Turkish separation plant would resolve every technical obstacle and convert a diversification project into an extension of the network it was designed to diversify away from — with the additional feature that the resulting facility would be located inside a NATO member and would therefore complicate rather than assist allied supply chain de-risking.
Public controversy in 2025 over claims that the project would be transferred to United States control, which were denied, illustrates the political sensitivity in the other direction. The project’s partner selection is a strategic decision rather than a commercial one, and it should be evaluated against the strategy’s purpose rather than against the cost of capital.
8. Conclusion
Reserve size is the variable most discussed in relation to Beylikova and the least determinative of outcome. Power in the rare earth chain sits at separation, where China holds 85 to 90 percent of capacity and a comparable share of process technology. The sequence that follows is determinate: certify the resource, because financing and offtake begin there; acquire separation capability from a party that does not reintroduce the dependency; bind demand before committing capital, because the market is thin and the incumbent can move prices; and integrate into magnet production, where the leverage and the industrial return both sit. The strategy’s principal risk is not technical failure but a financing arrangement that resolves the technical problem by surrendering the strategic one.
SOURCES
Farrell, H. & Newman, A. (2019) ‘Weaponized Interdependence: How Global Economic Networks Shape State Coercion’, International Security 44(1): 42–79.
USGS, Mineral Commodity Summaries 2025, rare earths chapter; USGS 2025 List of Critical Minerals, final list of 60 commodities published 7 November 2025.
International Energy Agency, Global Critical Minerals Outlook 2025, 21 May 2025.
Ministry of Energy and Natural Resources statements, OECD Critical Minerals Forum, Istanbul, 28 April 2026, and subsequent remarks on the Beylikova industrial facility.
Türkiye–Germany Energy and Minerals Partnership, 19 June 2026; Türkiye–Uzbekistan mining cooperation agreement, 29 January 2026.
Presidency of Defence Industries, 2025 defence and aerospace export and contract data.
Energy Market Regulatory Authority (EPDK), electric vehicle fleet projection to 2035.
Furkan Mahmut Tameroglu