Years remaining, linear scale
Bar length is proportional to years (metals and fuels stretched on the left; flows still span the full axis). Colour groups by category. R/P figures are what USGS, IAEA, and the Energy Institute publish; the metals R/P ratio is a genuinely misleading number and the caveats below spell out why.
Scroll horizontally, or use < / > to jump to the start or the long-lived end. Tiny bars get a pin at the true year mark. Faint purple ticks are bonus galactic content.
Data period: —. Fossil fuels and uranium are R/P ratios (reserves ÷ current annual production). Metals are also R/P; the column that actually matters is the recycling rate in the next card. Renewables have no R/P because they are not stockpiles.
Three categories, three arguments
—
—
The oil R/P ratio has stayed flat for decades
“We have 50 years of oil left” is the most widely cited depletion statistic. We have had roughly 50 years of oil left for 40 years. New reserves get discovered or reclassified about as fast as we pump. That does not mean the barrel is infinite; it means R/P is a bad proxy for when the last drop gets burned.
—
Metals are a stock, not a flow
—
Sources: USGS Mineral Commodity Summaries 2025 (reserves); UN International Resource Panel in-use stock estimates. Lead deserves a callout: in-use stock already exceeds reserves because the battery recycling loop recirculates the same lead over and over.
The “we will run out of lithium” talking point
—
- USGS 2025 reserves
- —
- USGS 2025 identified resources
- —
- 2024 world production
- —
- Reserves / production (years)
- —
- Resources / production (years)
- —
—
Lithium reserves have grown faster than consumption every year this decade. First-mover battery recyclers (Redwood, Li-Cycle, Umicore) are already pulling cathode-grade material out of end-of-life EV packs; the industrial-scale loop is a 2030s problem, not a physics problem.
Caveats
- R/P is not a time-to-empty. It is today’s known economic reserves divided by today’s production rate. Both numerator and denominator move. Higher prices pull lower-grade deposits into “reserves”; cheaper substitutes reduce demand. The ratio usually drifts sideways for decades as those two terms compete.
- Reserves vs resources. “Reserves” are the USGS-audited, currently-economic slice. “Resources” are the broader set of known deposits that may become economic. For lithium the two figures are 30 Mt vs 115 Mt; they are different scales of the same question.
- Fossil fuels probably will not run out. Current IEA modelling has oil demand peaking this decade and coal already past its peak in several regions. We will stop using them for economic and climate reasons before geology forces our hand. The real question is how much we leave in the ground voluntarily.
- Breeders and thorium are speculative at commercial scale. The 10,000-year figures for U-238 breeders and thorium assume fuel cycles that do not yet exist at commercial scale. They are physically achievable; they are not an engineering fait accompli.
- The sun comparison is deliberately absurd. Fossil fuels live on civilisational timescales; the sun lives on stellar timescales. That they are not substitutes is exactly the point the log chart makes.