01 — MethodologyHow to read these numbersEvery figure on this page comes from public documents: published feasibility studies and technical report summaries, SEC filings, company investor updates, and cross-commodity cost research from S&P Global Market Intelligence and Wood Mackenzie. Capital intensity is calculated the way owners and lenders calculate it — initial development capex divided by nameplate annual production capacity — expressed as dollars per tonne of annual paid metal for copper, per tonne of lithium carbonate equivalent for lithium, and per annual ounce for gold.The copper benchmark anchors to S&P Global’s December 2025 capital cost study of 26 primary copper projects targeting startup by 2030, which reported development capex from $172M to $9.14B and a weighted-average capital intensity of $22,359 per tonne of annual paid copper. Project-level anchors come from company disclosures: Lithium Americas’ Thacker Pass project updates and 2026 capex guidance, Ioneer’s October 2025 Rhyolite Ridge DFS update, Hudbay’s Copper World studies, Perpetua Resources’ Stibnite filings, and Rio Tinto’s Resolution Copper disclosures.Comparability caveats matter. Estimates carry different contingency levels (10% at Rhyolite Ridge, 20% on Copper World directs), different vintages of escalation, and different treatment of owner’s costs and first fills. Co-product revenue — boric acid at Rhyolite Ridge, antimony at Stibnite — changes the effective intensity per unit of primary metal. Treat every point estimate here as the midpoint of a spread, not a bid.What these figures are — and are notThese are published, owner-reported figures — feasibility-study estimates, board-approved budgets, and filed guidance — assembled for benchmarking. They are not iRecruit placement data, not contractor bids, and not a substitute for a project-specific estimate. Capital intensity compares initial development capex against nameplate capacity; it excludes sustaining capital, which on long-life mines can eventually rival the initial build, and it excludes operating costs entirely (all-in sustaining cost is a different metric). Estimates also move: Thacker Pass Phase 1 was published at $2.27B and now stands near $2.93B in company guidance, a reminder that the vintage of an estimate is as important as the number itself. Use the spreads, note the contingency and co-product assumptions, and read every figure against its source and date.02 — At a glanceThe benchmarks at a glanceThe whole guide in one screen. Each row jumps to the full section.01CopperS&P weighted average $22,359 per annual tonne across 26 projects; US-Canada projects average about $28,900$22.4k/t02LithiumThacker Pass Phase 1: $2.93B for 40,000 tpa LCE (~$73,000/t); Rhyolite Ridge: $1.68B for 24,500 tpa (~$69,000/t)$73k/t LCE03GoldStibnite: $2.58B for ~463,000 oz/yr in years 1-4, roughly $5,600 per annual ounce$5.6k/oz04Rare earthsMP Materials: $400M DoD equity plus $150M loan; 10X plant lifts US magnet capacity toward 10,000 tDoD-backed05Capex structureProcess plant 30-40% of total capex; site infrastructure 20-30% on remote builds; power alone 10-15%30-40%The anchor figures for 2026: about $22,400 per tonne of annual paid copper on S&P Global’s global weighted average — roughly $29,000 across its US and Canadian projects — about $73,000 per tonne of annual lithium carbonate capacity at Thacker Pass on current guidance, about $69,000 at co-product-supported Rhyolite Ridge, and near $5,600 per annual ounce at Stibnite, with the processing plant absorbing 30-40% of whatever the total turns out to be.
03 — Benchmarks2026 capital-intensity benchmarks by commoditySix benchmarks below, each traceable to a named public source. Copper is quoted per tonne of annual paid metal, lithium per tonne of annual lithium carbonate equivalent, gold per annual ounce of early-years production. Ranges reflect real project spreads, not modeling uncertainty.$22,359Copper (open pit, benchmark)S&P Global’s December 2025 study of 26 copper projects due by 2030: weighted-average capital intensity of $22,359 per tonne of annual paid copper, with US and Canadian projects the most capital-intensive region at roughly $28,900 per tonne, against about $19,000 in Latin America.$1.3BCopper (Copper World, AZ)Hudbay’s September 2023 enhanced pre-feasibility study puts Phase I initial capital at $1.3B, including a 20% contingency on direct costs, with a definitive feasibility study and sanction decision due late 2026. S&P names Copper World among the projects pushing the US-Canada region’s capital intensity above the global average.$73kLithium (Thacker Pass, NV)Lithium Americas’ Phase 1 now carries a budget near $2.93B for 40,000 tpa of battery-grade lithium carbonate, about $73,000 per annual tonne, up from the earlier $2.27B estimate; DOE lent $2.26B and mechanical completion targets late 2027.$69kLithium (Rhyolite Ridge, NV)Ioneer’s October 2025 update pegs capex at $1,683M (AACE Class 2, 10% contingency) for roughly 24,500 tpa LCE plus 135,500 tpa of boric acid, near $69,000 per annual LCE tonne before co-product credits, supported by a $996M DOE loan closed in January 2025.$5,600Gold-antimony (Stibnite, ID)Perpetua Resources broke ground in October 2025; its March 2026 technical report summary raised initial capital to $2.58B for ~463,000 oz/yr of gold in the first four years — roughly $5,600 per annual ounce — plus the only US antimony reserve, backed by a $2.9B EXIM loan approved in May 2026.$2B+Copper (Resolution, pre-build)Rio Tinto and BHP report over $2B invested at Resolution since 2004 on shafts, studies and permitting — before a construction decision. Deep block caves front-load a decade-plus of underground development ahead of the first paid tonne.Sources: S&P Global Market Intelligence copper capital cost study (Dec 2025); Lithium Americas Thacker Pass project updates and 2026 capex guidance; Ioneer Rhyolite Ridge update (Oct 2025); Hudbay Copper World pre-feasibility study (Sep 2023); Perpetua Resources Stibnite technical report summary (Mar 2026) and EXIM loan approval (May 2026); Rio Tinto Resolution Copper disclosures; MP Materials–DoD announcements (Jul 2025); Wood Mackenzie mine cost outlooks (retrieved Aug 2026).How to read a capital-intensity numberNameplate capacity is a design number; realized throughput in year one rarely matches it. Contingency varies from 10% to 20%+ across the studies cited here, and the estimate vintage matters as much as the figure — a 2020 DFS and a 2026 budget describe different projects. Co-products cut effective intensity (boric acid at Rhyolite Ridge, antimony at Stibnite), and none of these figures include sustaining capital, which accrues over decades and is budgeted separately from the initial build. 04 — Cost driversWhat moves the numberTwo projects with identical nameplate capacity can land billions apart. Six drivers explain most of the spread between a $22,000-per-tonne copper build and a $73,000-per-tonne lithium one.InverseGrade and strip ratioLower grade means more tonnes moved and a bigger plant per unit of metal — capex scales with mass flow, not metal output. It also scales the earthworks and pre-strip package that dominates early construction.Top rankUnderground depthBlock caves like Resolution stack shaft-sinking and years of development ahead of revenue; S&P’s December 2025 study separately ranks the US-Canada region the most capital-intensive globally, citing remote locations, regulatory complexity and advanced processing.+30-50%Remote infrastructurePublished capex analyses put remote-site process plants 30-50% above accessible-region equivalents; self-generated power alone can absorb 10-15% of total capex, before roads, water and camp.20+ yrsPermitting carryResolution’s owners spent over $2B across two decades before any construction decision; Stibnite carried nearly a decade of permitting cost before its October 2025 groundbreaking.+29%Estimate escalationThacker Pass moved from a $2.27B published estimate to about $2.93B in current guidance on engineering growth and labor agreements — a live example of feasibility-to-build drift.-$/tCo-product creditsBoric acid at Rhyolite Ridge and antimony at Stibnite offset the effective cost per unit of primary product — one reason single-number benchmarks need the footnotes attached.The pattern across drivers: capital intensity is mostly a function of what must be built before the first saleable tonne — mass moved, plant size, infrastructure distance, and years of carrying cost. Commodity price barely appears in the equation; it decides whether the number gets funded, not what the number is. 05 — VariationThe project-by-project pictureThe clearest read on 2026 US mine construction is the projects actually breaking ground or guiding budgets this cycle — public market activity, from company filings and government announcements.Thacker Pass Phase 1$2.93B40,000 tpa lithium carbonate; $2.26B DOE loan and GM backing; mechanical completion targeted late 2027 (Lithium Americas, 2026 guidance).Copper World Phase 1$1.3BHudbay’s Arizona build, 20% contingency on directs (2023 PFS); DFS and sanction decision due late 2026; flagged by S&P among the most capital-intense US projects.Stibnite Gold$2.58BBroke ground October 2025; ~463,000 oz/yr gold in years 1-4 plus the only US antimony reserve; $2.9B EXIM loan approved May 2026 (Perpetua TRS, March 2026).Rhyolite Ridge$1.68B24,500 tpa LCE plus 135,500 tpa boric acid; $996M DOE loan closed January 2025 (Ioneer, October 2025 update).Resolution Copper$2B+Invested since 2004 on shafts, studies and permitting; land exchange completed 2026; the main construction decision is still ahead.MP Materials 10X$550M+DoD took $400M in equity and lent $150M; the 10X magnet plant targets ~10,000 t of US capacity by 2028 — processing-side market activity, not a mine-build benchmark.Read together, the slate spans $1.3B to nearly $3B per project — and that range excludes Resolution, where more than $2B has been spent without a build decision. Federal capital sits behind most of it: DOE loans at Thacker Pass and Rhyolite Ridge, EXIM at Stibnite, and direct DoD equity at MP Materials. 06 — TrendThe escalation problem: why 2026 estimates age fastThe single most instructive escalation datapoint in US mining is Thacker Pass: a published $2.27B Phase 1 estimate revised to roughly $2.93B — about 29% — on engineering growth, a union labor agreement and related factors, per Lithium Americas’ disclosures. That drift happened on a fully financed, actively constructed flagship, not a paper project.The sector-wide data points the same way. S&P Global’s December 2025 copper study recorded development capex of $172M to $9.14B across its 26-project cohort, a substantial step up from the prior study’s $53M to $5.45B range, and its 2026 mine cost outlook flags continued inflation reshaping project economics. Tariff exposure on imported steel and consumables adds a further US-specific layer, per Wood Mackenzie’s cost commentary.Notably, the squeeze is on the build side more than the run side: Wood Mackenzie’s 2025 copper outlook showed operating costs plus sustaining capex easing about 13% year on year even as development capex kept climbing. Owners budgeting 2026-28 construction starts should assume the feasibility number is a floor, and that contingency percentages set in 2023-24 studies are the first line to stress-test. 07 — WorkforceThe workforce behind the capex lineEvery one of these budgets is mostly a labor plan. A $1B-plus mine build needs EPCM project managers and construction managers who have delivered process facilities, structural-mechanical-piping and E&I leadership for a concentrator or chemical plant that alone absorbs 30-40% of capex, earthworks superintendents who can run pre-strip and bulk excavation at a scale most heavy-civil careers never touch, and commissioning and metallurgical staff to bridge construction into operations. Remote sites — northern Nevada, central Idaho, the Arizona copper belt — add rotation schedules, camp logistics and documented remote-site pay premiums to every one of those hires.iRecruit is building candidate networks across these mine-construction disciplines ahead of demand — mapping EPCM, process-plant and heavy-civil talent as the US project slate moves from permitting into construction. The projects named on this page are cited as public market activity that defines where that talent will be needed; owners and contractors staffing against these budgets can draw on a network assembled before the hiring crunch, not after it.For the hiring side of this market, see the Mining & Critical Minerals practice. 08 — FAQFrequently asked questions What does capex per ton of annual capacity actually measure?+It divides a project’s initial development capital by its nameplate annual production — so a $2.93B lithium plant designed for 40,000 tpa runs about $73,000 per annual tonne. It measures the cost of creating capacity, not the cost of operating it; all-in sustaining cost is a separate metric.Why is lithium so much more capital-intense per tonne than copper?+The units differ: lithium intensity is quoted per tonne of finished battery-grade chemical from what is effectively a mine plus a chemical plant, while copper is quoted per tonne of paid metal from a concentrator. Smaller annual tonnages and heavier processing push lithium’s per-tonne figure to $69,000–$73,000 against copper’s roughly $22,000–$29,000.How much of a mine build is the processing plant versus mine development?+Published capex breakdowns put the process plant at 30-40% of total initial capital — typically the single largest line — with site infrastructure another 20-30% on remote builds and self-generated power alone up to 10-15%. Mine development’s share swings widely: modest on shallow open pits, dominant on deep underground projects like Resolution.Why do feasibility estimates escalate before construction finishes?+Engineering definition matures, labor agreements land, and input prices move between study and build — Thacker Pass rose about 29% from its $2.27B published estimate to roughly $2.93B in current guidance. S&P Global’s 2025 copper study likewise recorded a step-change in project capex ranges versus its prior edition.What construction talent do these projects compete for?+EPCM and owner’s-team project managers, process-plant construction and commissioning leadership, structural-mechanical-piping and E&I supervision, and earthworks management at bulk-excavation scale — usually on remote sites carrying rotation and premium pay structures. iRecruit is assembling candidate networks across these disciplines ahead of the US mining build-out. Mine construction talentPlanning a mine build? The talent market is tightening faster than the capex line. iRecruit is assembling candidate networks across EPCM, process-plant and heavy-civil disciplines ahead of the US mining build-out. Start the conversation early.Scope a search →Candidates — register with us →
