The future is not one machine

Grids, storage, efficiency, materials and habits shape different futures.

Three futures use the same words

Replacing without changing

The first plausible scenario electrifies cars, heating and much of manufacturing industry, but leaves practically untouched the size of vehicles, how often people travel, and society's overall level of consumption of goods and services. In this scenario, fossil-fuel consumption falls steadily, but demand grows strongly, in almost the same proportion, for grid capacity, for minerals to manufacture batteries and equipment, and for new electricity generation to supply all that newly electrified additional demand. It is a scenario of pure technological substitution, with barely any change in the underlying behavior patterns.

Efficiency with rebound

The second scenario substantially improves each individual machine's technical efficiency — every vehicle, every appliance, every industrial process uses less energy per unit of service provided — but that same cheapening of the service encourages, as a direct consequence, an increase in the total number of machines in use and the total amount of services consumed. Society's total energy consumption, in this scenario, can end up not falling appreciably, or can even keep growing, even though each individual component of the system is, taken separately, considerably more efficient than before. Without absolute total-consumption targets accompanying the efficiency improvement, that same efficiency risks becoming, in practice, a social license to consume more, not a real reduction in aggregate consumption.

Sufficient service and a diverse system

The third scenario deliberately combines several different levers at once, instead of betting on a single one: buildings better designed and better insulated from the start, cities organized to structurally reduce the need for long trips, a widespread culture of repairing products instead of automatically replacing them, well-sized and well-occupied collective transport, electrification of the end uses that can benefit from it, renewable generation combined with dispatchable generation when the system needs it, storage at different timescales, electrical interconnection between different regions, and flexible-demand mechanisms that adapt to the real generation available at each moment. This third scenario reduces risks jointly, precisely because it does not depend on a single specific technology working perfectly: it instead requires sustained coordination among many different elements of the whole energy system.

No scenario is a prophecy about what will happen

None of these three scenarios claims to predict exactly what will actually happen in the real future. Their genuine value lies, instead, in clearly making visible the logical consequences of starting from different initial assumptions, so it becomes possible to reason rigorously about what additional conditions would have to be met for each of those specific scenarios to actually materialise in practice. Comparing several plausible scenarios against each other, instead of clinging to a single one treated as the definitive truth, is precisely what allows an energy-policy decision-maker to prepare robustly for more than one possible future at once, without betting everything on a single prediction that might not come true.

Signals worth tracking

Real in-service battery lifespan; effective mineral recovery at end of life; real permitting and construction time for new electricity grids; methane emissions actually measured, not just declared; data centers' electricity demand; the real pace of energy retrofitting for existing buildings; how energy poverty evolves; hydrogen electrolysers' real utilization factors; real costs and timelines for new nuclear projects; and fusion demonstrations at whole-system scale, not just isolated-experiment scale.

Good energy policy does not bet all its room for maneuver on getting a single specific prediction about the future right. It cuts avoidable demand starting now, diversifies its technology options, preserves the flexibility to change course if the data calls for it, learns continuously from real data rather than only from projections, and avoids building infrastructure that only works well if the future turns out exactly as anticipated in the most optimistic scenario.

The numbers, with their source

Critical-mineral demand could multiply up to sixfold

According to the International Energy Agency, achieving the global net-zero target would require tripling critical-mineral demand by 2030 and quadrupling it by 2040 relative to today's levels; a global net-zero-by-2050 scenario would require up to six times more minerals in 2040 than today. Lithium is by far the mineral with the fastest projected growth, with demand potentially multiplying by more than 40 times by 2040 in the most climate-ambitious scenarios, followed by graphite, cobalt and nickel, with multipliers of around 20 to 25 times. A typical electric vehicle requires about six times more minerals than an equivalent internal-combustion vehicle, and an onshore wind installation requires about nine times more mineral resources than a similarly sized gas plant.

Source: IEA — The Role of Critical Minerals in Clean Energy Transitions.

Why this figure is not a verdict, but a signal to watch

This projected growth in mineral demand does not automatically imply a supply crisis or a failure of the energy transition: it implies that the real pace of opening new mines, of improving recycling of already-manufactured batteries and equipment, and of technological substitution toward battery chemistries that use fewer critical minerals are, precisely, concrete signals worth actively monitoring to judge whether an accelerated-electrification scenario is realistic in practice or only on paper.

What this table is missing

No aggregated mineral-demand multiplier distinguishes between a scenario that invests heavily in mineral recycling and circular economy, developed in this same block's leaf on repairing, reusing and recycling, and one that relies exclusively on new mineral extraction — the difference between the two cases determines much of the real environmental and social cost of any future energy scenario.

Sources

Scenarios don't predict: they let us ask "if we do this, what else would have to be true?"