Loading…
Loading…
One screen answering: “if the dominant supplier halts, who can fill the gap, how fast, at what regulatory risk?” Replaces manual cross-referencing of the production table with the IPTM register.
If China refining (65% of global) halts: 6 substitutes cover 100% of non-CN capacity · fastest ramp 0-6mo (CL) · highest policy-risk: US (score 12.9, 19 filings).
Stage: Lithium chemicals refining (carbonate + hydroxide) · Year: 2025 · Source: USGS Mineral Commodity Summaries 2026 (Feb 2026 release; 2025 data).
Post-removal HHI = 2,266 (moderate). Lead-time is a heuristic from share-of-stage (≥10% → 0-6mo brownfield · ≥2% → 6-12mo ramp · >0 → 12-24mo new line · 0 + supplier-directory present → 24-36mo greenfield). Policy-risk = Σ over last-24m lithium filings issued by that country (severity × polarity-sign × 5y-linear recency). This measures a country's own export-restriction behaviour only — it is a separate axis from Western sanctions exposure (flagged below in red where it applies). A country with zero restrictive filings can still be sanctioned by the US/EU/UK and therefore not a real sourcing option.
Mining is diversified (Australia largest); refining is the binding chokepoint. **Source divergence to note**: USGS MCS 2026 puts China lithium-chemicals refining share at ~65%; IEA Global Critical Minerals Outlook 2025 and SWP Aug 2025 (Müller et al.) put it at ~70%. The gap reflects methodology: USGS counts installed nameplate capacity (some Chinese plants ran below utilisation in 2023-2024 on a weak lithium-carbonate price); IEA counts effective-output share with better visibility into Chinese domestic consumption. Both sources confirm the direction: Chinese refining share has steadily climbed (~50% 2014 → ~60% 2019 → ~65-70% 2025) while Australian conversion capacity (Kwinana, Kemerton) has ramped more slowly than forecast. The policy-relevant point: even if Australian spodumene mining share is ~32%, Australia refines only ~8-11% — the rest of Australian-mined spodumene is shipped to Chinese converters. Diversifying the mining origin does NOT diversify the refining dependency.