Bioseq
Report · Edition 1 · October 2026

How big is biomining?

A first estimate of the sector, its potential from waste, and where biology could take rare earths.

18 pages · Free · Sources for every figure · By Martin Mantalvanos

Cover of How big is biomining? Edition 1

$2.6bn

a year: copper, nickel, zinc and cobalt in the 2020 benchmark, at reference prices.

$4.8bn

a year: an illustrative deployment scenario for tailings and battery waste, before costs.

~100,000t

of rare earths a year estimated to end up in phosphogypsum waste.

What's inside

From the 2020 benchmark to rare earths

  1. What changed since 2020: copper prices, cobalt export quotas, Nuton's first copper, Terrafame.
  2. What we know: the 2020 production benchmark and the latest site disclosures.
  3. The potential: $4.8bn under our working assumptions, by metal, and how far it could scale.
  4. The material is already out there: tailings, coal waste in Wales and Poland, EU mining waste, Appalachian mine drainage.
  5. What would change these numbers: the four assumptions that matter most.
  6. Rare earths: why now, where they sit in waste, what biology does well, how far along it is.
  7. What's holding it back: four open questions, each testable.
Key assumptions

Every assumption is stated, so it can be challenged

The workbook behind these numbers is available on request: martin@bioseq.io

Tailings operations
645 assumed, each reprocessing 750,000 t a year
Battery circuits
25 assumed biological or hybrid circuits, each treating 10,000 t of black mass a year
Reference prices
Copper $8,500/t, cobalt $40,000/t, nickel $16,000/t, zinc $2,800/t, lithium carbonate $12,000/t, manganese sulphate $600/t, mixed rare-earth oxide $15,000/t
Price sensitivity
~$6bn at copper $13,000/t and cobalt $55,000/t, other prices fixed
Range
$1.2bn–$12bn, changing site counts only (a quarter to 2.5×)
Value basis
Gross product value before costs. Some biological recovery could replace conventional recovery rather than add to it.
Help build the count

One dated site beats any assumption in this model

Send what you can: the site and date, feed type and annual tonnes, head grade, saleable recovery, and product and annual output.

Send a site

I'll share back what comes in, with the spreadsheet behind these numbers.