The future of energy: from scarcity to abundance?

RethinkX's big idea is that cheap solar, wind and batteries do more than replace power stations: they could make clean electricity so plentiful that we find new uses for it. The evidence supports part of this. Solar and wind are now competitive generators and batteries are growing very fast. The evidence does not establish the reports' full claims about costs everywhere, the speed at which conventional assets collapse, or inevitable abundance.

Our position is qualified confidence in the direction, with open questions about costs, timing, delivery and who benefits. We label each claim below as observed, a forecast, a scenario or unresolved.

The idea: build more than you need

RethinkX asks what becomes possible when a clean-electricity system produces substantial surplus output. If you build enough solar and wind to cover the hardest days, and add storage, what happens on the many other days when you produce more than you need?

The 2020 report calls that extra output "super power". It models the trade-off between building more generation and building more storage: extra generation can cut how much battery you need, and the surplus can serve flexible new demand (RethinkX, Rethinking Energy 2020–2030, pp. 7–13). That is a useful way to think about a clean system, whatever one makes of the specific numbers.

What the 2020 report actually models

It is a scenario, not a forecast. The 2020 report models whether solar, wind and batteries could supply existing electricity demand by 2030 in three US regions: California, Texas and New England. It uses hourly data for two years, 1 July 2017 to 30 June 2019 (pp. 24–27). The report itself says its analysis is "not a forecast, but rather an illustrative 'limit scenario'" (p. 11). It excludes electricity imports, distributed resources such as rooftop solar, electric vehicles, demand response and conventional backup. It also deliberately excludes networks (pp. 28–29).

The "under 3 cents" claim is conditional. The report's headline says the continental US could reach a 100% solar, wind and battery system by 2030 for under US$2 trillion, with electricity under 3 US cents per kilowatt-hour "if 50% or more of the system's super power is utilized" (p. 11). The report's own summary table (Table 1, p. 13) shows how much depends on the region and on how much surplus is used. These are 2030 scenario outputs in US cents per kilowatt-hour, for generation and storage only.

Source: RethinkX, 2020, Table 1, p. 13 (PDF p. 15).

Three cautions follow. New England stays above 3 cents unless extra investment is made and all the surplus is used. Using the surplus needs equipment, factories or appliances that can take it, and changing the denominator does not make those free. These generation-and-storage figures exclude networks and are not Australian retail electricity-price forecasts.

An arithmetic correction. The 2020 report says that 20% more investment in California raises annual surplus "by over 190%", from 309 to 592 terawatt-hours (p. 12). Those figures are an increase of 91.6%, to 191.6% of the original amount: (592 − 309) ÷ 309 = 0.916. We use the endpoints, not the "190%" wording. This is our own arithmetic check.

What the 2025 report adds

Understanding Stellar Energy (Dorr, Arbib and Seba, May 2025) extends the idea to much larger energy use. It asks what a clean system could supply if people used as much energy per person as today's wealthiest societies, using ten years of weather data (1 July 2010 to 30 June 2020 in the northern hemisphere, calendar 2010 to 2019 in the southern; p. 66).

Its main dated claim is that "superabundance" could be reached by 2035 in leading regions and before 2045 for most of the world (p. 49). Terms such as "leading regions" and "abundance" would need operational definitions before that claim could be marked right or wrong. It is a conditional future claim, not an observed result.

The scenarios have important boundaries (pp. 63–67):

• Energy per person is an input, not a forecast. The "prosperity" scenario sets about 6,150 kWh per person per year for residential, commercial and industrial electricity, plus 4,200 kWh per person annually for road transport, plus additional climate-dependent electricity for heating and cooling. Heating and cooling are separate from that 4,200 kWh allowance. The superabundance version triples only the first category, to about 18,400 kWh (pp. 64–65). These standards are chosen, not predicted.
• Existing hydro and nuclear stay. "100% coverage" does not mean every region's generation is newly built solar, wind and batteries.
• Aviation and shipping fossil fuels are excluded.
• Networks and land costs are excluded. Interconnection, transmission and distribution are outside the model, and it uses US 2024 cost benchmarks in US dollars worldwide.
• A small conventional reserve could cut capital needs by up to 25% in some regions (pp. 66–67), which is more accommodating to legacy assets than the 2020 framing.

The 2025 report's regional case studies do not include an Australian national or National Electricity Market case (pp. 63, 68–111).

What independent evidence shows

The following comes from an editorial evidence assessment dated 7 November 2026, which compiled it from published sources. We have not independently re-checked each figure against its original publication, so verify them before relying on them. Each is a different measure and they cannot simply be combined.

Observed: renewable generation is competitive. IRENA reported that in 2024 the global weighted-average cost of new solar PV was US$0.043 per kWh and onshore wind US$0.034, and that 91% of newly commissioned utility-scale renewable capacity was cheaper than the cheapest new fossil option. Costs rose 0.6% and 3% from 2023, so this is competitiveness, not a smooth annual decline (IRENA, Renewable Power Generation Costs in 2024, July 2025). IRENA's 2026 report for 2025 outcomes gives US$44/MWh for solar and US$33/MWh for onshore wind (IRENA, 2025 costs, July 2026). These are generation costs, not 24-hour delivered prices.

Observed: batteries are growing very fast. The IEA reports 108 GW of battery storage added globally in 2025, up 40%, with installed power capacity eleven times its 2021 level. Most projects still have about two hours of duration, and power capacity is not stored energy (IEA, Global Energy Review 2026). Separately, the IEA reports that average battery prices fell 8% in 2025, with regional differences (IEA, Global EV Outlook 2026, Electric vehicle batteries). Battery-pack prices are not the same as the cost of a complete stationary storage system.

Observed: the global mix is changing, but incumbents have not collapsed. The IEA estimates that in 2025 solar and wind supplied 17% of global electricity, all renewables 34%, and renewables plus nuclear 43%. Coal generation fell about 0.5%, gas rose about 0.5% and nuclear reached a record (IEA, Global Energy Review 2026, electricity supply). The IEA forecasts renewables plus nuclear at 50% of generation in 2030, with coal at 27% and gas still growing (IEA, Electricity 2026). That is a forecast that differs from RethinkX's, not proof that either is right.

Modelled: firm renewable power is plausible but not free. IRENA's May 2026 study models firm renewable power at US$54–82/MWh at favourable sites using 2025 inputs. Firm power here means renewable generation combined with storage to meet a specified delivery profile, modelled at 95% reliability. That is not the same as whole-grid reliability, and a 95% target does not establish uninterrupted availability. It supports the mechanism of overbuilding plus storage, but its costs for 2030 and 2035 are scenarios and it does not replicate RethinkX's 100% claim (IRENA, 24/7 Renewables, May 2026).

What this means for Australia

Observed in the National Electricity Market. AEMO's Quarterly Energy Dynamics for July to September 2026 reports renewables supplied 44.3% of NEM generation, with a half-hour record of 80.4%, and batteries supplied 8.0% of evening-peak generation against gas at 5.1%. Negative or zero prices occurred in 14.4% of dispatch intervals, down 4.5 percentage points year on year as charging absorbed more output (AEMO, QED Q3 2026). Quarterly, half-hourly and annual shares are different measures, and NEM figures are not all-Australia figures. The fall in negative prices is itself a lesson: surplus opportunities shrink when storage and new demand respond.

Modelled: what the least-cost system looks like. AEMO's final 2026 Integrated System Plan identifies renewables connected by networks, firmed with storage and backed by gas as the least-cost, policy-consistent pathway (AEMO, 2026 ISP). CSIRO's GenCost 2025–26 continues to find renewables with storage the lowest-cost route to net zero, and its FAQ says replacing ageing assets raises generation costs, so cheap new equipment does not guarantee falling whole-system costs (CSIRO, GenCost 2025–26, 15 July 2026). These are planning models, not observed futures. They support a renewable-led system with networks, storage and backup, not a solar-wind-battery-only design.

Does abundance leave materials out of the picture?

No. Lower fuel extraction does not mean an end to equipment replacement, mining, land use or recycling. An IEA PVPS study of silicon solar materials found that end-of-life modules could potentially supply 30–45% of cumulative PV-sector silver demand over 2025–2050. That is a scenario potential, not an achieved recycling rate (IEA PVPS Task 12, 2026). It supports circular design. It does not support claims that material inputs become unnecessary. See our article on what happens to solar panels at the end of their life.

The questions that decide whether abundance happens

Abundance is a possibility to design for, with measurable conditions for success, not an automatic result of a falling cost curve. Questions we will keep asking, in editorial inference not drawn from a single source:

• Who can access surplus energy, and what equipment do they need to use it?
• What does it cost to deliver, including networks, with equivalent reliability?
• Do heat storage, charging and flexible industrial loads earn enough from surplus to pay for themselves?
• Do materials stay in productive circulation?
• Are the benefits shared, for example with renters and people without rooftops?

For how the numbers behind RethinkX's forecasts hold up so far, see Predictions vs Reality. For the person behind the reports, see Tony Seba and RethinkX. For the practical home side, see Energy & Storage.

About this article

We read the three RethinkX reports directly and checked the figures quoted from them against the PDFs. The independent evidence is taken from an editorial evidence assessment dated 7 November 2026 and we have not re-checked each figure against the original publications. The assessment did not rerun RethinkX's models or the full GenCost model, and it did not inspect the IRENA 2025-cost PDF (only its official summary). This is an early version, published so that it can be corrected and improved.