Copperbelt Data Explorer — Magarini Mining

Who Magarini should actually spend days on — and the sanity-check column. score = product form × √(output kt) × status factor × access factor. Product form ranks what an intermediary can genuinely resell: LME-grade cathode (refineries, 1.3) over blister/anode (smelters, 1.0) over concentrate (mines, 0.8). Access marks down global majors that market through in-house trading arms (Glencore, CMOC/IXM, Zijin, FQM… ×0.4) and SOE channels (×0.7); state-linked and mid-tier producers score full access — and anything Gécamines-linked is flagged, because its JV offtake share is sold via tender, the classic legitimate source of DRC spot metal. Realistic spot MT/mo assumes ~95% of output is committed under offtake: it is the number to hold against any offer or demand. An offer of 50,000 MT concentrate "origin Kenya" or 10,000 MT/month of "99.99% cathode" fails this column instantly — which is precisely why it exists. Weights live in prospects.js (MAGARINI block) and are meant to be tuned.

How the prospect score works. score = type weight × √(Cu output kt) × status factor. Type weight reflects how bearing-intensive the plant is (smelter 1.3, integrated mine 1.2, underground 1.15, open pit 1.0, refinery 0.9, exploration 0.3). Output uses the reported 2024 figure; where unreported, a conservative type default is assumed and the row is marked est. Status scales it down for limited operations (×0.7), care & maintenance (×0.45 — flagged restart watch: restarts trigger large one-time rebuild orders) and suspension (×0.3). Tier A ≥ 14, Tier B ≥ 7, else C. These are territory-planning heuristics, not SKF data — the weights live in prospects.js and are meant to be tuned.