Critical Minerals
Power and resources are converging.
AI infrastructure requires energy and the minerals to build it. PowerLink is positioned at both: stranded energy, critical minerals, and sovereign capability.
Lithium demand growth to 2050 under transition scenarios (LCE basis); ranges by scenario.[7]
Copper intensity of AI/hyperscale data centers vs. conventional facilities (per facility; per-MW differential is lower).[8]
Share of global mined cobalt from the DRC — supply-concentration / geopolitical risk.[6]
China's share of global rare-earth processing and magnet production (2024).[5]
Supply chain security is infrastructure.
Why an energy company is looking at minerals
A data center is not built out of electricity. It is built out of copper, and it is backed by batteries, and it is cooled and controlled by systems that depend on rare-earth magnets. The same buildout that makes power scarce also makes a specific short list of materials scarce, and the two constraints tighten at the same time and for the same reason. PowerLink's position is that an infrastructure company exposed to one of these constraints and not the other is only half-hedged.
The materials are upstream of the megawatt
Transformers, switchgear, cabling, and generation equipment all consume the same commodities. A shortage in the material shows up first as a lead-time problem in the equipment — standard large power transformers averaged roughly 128 weeks as of Q2 2025, against about 50 weeks in 2021 [2].
Where PowerLink already operates
Several of the geographies where PowerLink is developing energy projects sit on or near meaningful mineral endowments. Development activity in the DRC and Nigeria is at pre-development and prospective stage respectively, and reflects the overlap between where stranded power can be built and where these materials are found.
Power is the input minerals need
Extraction and, more acutely, processing are electricity-intensive. In jurisdictions where grid supply is the binding constraint on adding processing capacity, an operator that can bring its own generation is solving the prerequisite rather than queuing for it.
Sovereign capability is the buyer's concern
Governments and industrial buyers are increasingly evaluating supply on where it is refined, not only where it is mined. That reframes a processing position as a strategic asset, which is the lens PowerLink applies to its pipeline.
The four exposures
PowerLink's stated interest is in lithium, copper, cobalt, and rare earths. Each is exposed to the AI and electrification buildout differently, and each carries a different concentration profile.
Lithium — demand growth
Lithium demand rises several-fold by 2050 across transition scenarios; the IEA's Net-Zero pathway implies roughly a tenfold increase and Wood Mackenzie models a range of 5.6 to 13.2 Mt LCE. The +500% figure PowerLink cites sits in the conservative-to-middle part of that range [7].
Copper — intensity per facility
AI and hyperscale facilities can run roughly three to four times the copper intensity of conventional data centers on a per-facility basis, though the per-megawatt differential is lower [8]. Copper is the material where the AI buildout translates most directly into incremental tonnes.
Cobalt — supply concentration
The Democratic Republic of the Congo accounts for approximately 70–73% of global mined cobalt supply [6]. That is a concentration risk before it is a volume risk: the question for a buyer is not whether the material exists but whether a single jurisdiction sits between them and it.
Rare earths — processing concentration
China accounts for roughly 91% of refined rare-earth output and about 94% of sintered permanent magnets, against roughly 60% of mining [5]. The bottleneck is the refinery and the magnet plant, not the mine — which is why processing capability, not resource ownership, is the position that matters.
Concentration risk is the actual thesis
Read together, those four figures describe one condition rather than four. The materials that the AI buildout consumes are produced and, more sharply, refined in a small number of places. A buyer's exposure is therefore not primarily to price — it is to the possibility that a single jurisdiction's policy decision removes access entirely. That is a different risk with a different hedge: the response is not to buy forward, it is to develop capability somewhere else. PowerLink's positioning statement, that supply chain security is infrastructure, is a claim about which of those two responses the market is moving toward.
What a position looks like at this stage
Mineral development has a long ladder between an idea and a shipment, and it is worth being precise about which rung PowerLink's positions sit on. None of them is producing. The activity described below is early-stage evaluation and relationship-building, not extraction.
Prospective
A geography and a counterparty have been identified and are being assessed. There is no committed capital, no defined resource, and no development schedule. Nigeria sits at this stage, as do several of the energy-led sites shown on the map.
Pre-development
Work has moved past assessment into structuring — technical review, local partnership, and the question of what infrastructure a site would require to be viable. The DRC positions at Manono and Kinshasa sit here. Pre-development is not a commitment to develop.
Development
Active project work with a defined counterparty and a schedule. Among the sites shown, only Eswatini carries this label, and it is an energy project rather than a minerals one — an accurate picture of where the two halves of the platform currently stand relative to each other.
Why the labels matter
Early-stage mineral positions are routinely described in language that implies more certainty than exists. Publishing the stage alongside the site is how a reader can tell the difference between a signed development agreement and a conversation, and it is why the map carries a disclaimer rather than a projection.
How energy and minerals reinforce each other
The two halves of the platform are intended to be mutually enabling rather than merely adjacent.
Generation unlocks processing
Processing capacity in an underserved grid is gated on reliable power. On-site generation is designed to remove that gate, making a processing position developable in locations where grid supply would otherwise rule it out.
Processing anchors the load
Running in the other direction, a processing facility is a firm, long-duration industrial load — the kind of offtaker that makes a generation asset financeable. Each side improves the other's case.
Shared development footprint
Site access, local partnerships, permitting relationships, and logistics are largely common to both activities. Developing them once and applying them twice is the practical argument for holding both exposures in one company.
Stage discipline
Site statuses shown on this page — prospective, pre-development, development — are internal stage assessments as of the date shown. No site is operational, and there is no assurance any site will be developed, financed, or reach operation.
Global Development Pipeline. Strategic Positioning.
- 01Bozeman, MTProspective
Tire Waste + BTC / AI Compute
- 02Wilkinson, MSProspective
Stranded Gas + Energy Infrastructure
- 03Clearfield, PAProspective
Stranded Gas + Energy Infrastructure
- 04EswatiniDevelopment
Sovereign Micro-Hydro Partnership
- 05NigeriaProspective
Critical Minerals + Energy Infrastructure
- 06DRC / ManonoPre-Development
Critical Minerals + Energy Infrastructure
- 07DRC / KinshasaPre-Development
Tire & Plastic Waste
PowerLink does not follow the grid. PowerLink builds where the grid cannot.
Site statuses (prospective / pre-development / development) reflect internal stage assessments as of the date shown. No site is operational; there is no assurance any site will be developed, financed, or reach operation.