How solar, wind and hydro could power the world, at lower cost

Print Friendly, PDF & Email

Stanford University professor Mark Jacobson and colleagues at the University of California at Berkeley and Aalborg University in Denmark have updated and expanded their analysis on how the world – well, at lest 139 countries – could be powered entirely by solar, wind and hydro resources.

The study, whose earlier version caused controversy and a strident critique by rival academics, now includes further modelling and a range of scenarios that include hydrogen storage, heat pumps and battery storage, the newly arrived technology du jour which is changing many assumptions about future energy markets.

Where this study – published this week in Renewable Energygoes further than others is that it lays out three different methods of not just providing 100 per cent renewables for electricity, but also for heating and cooling, for transportation, and even agriculture, forestry and fishing.

In short, the Jacobson team is proposing to electrify the world, and doing it through three main renewable energy sources: solar, wind and hydro.

It says it can do this at a comparable cost, even slightly cheaper, than business as usual (fossil fuels), and at just one-quarter of the cost if you dial in savings from avoided fossil fuel damage to the environment and health.

“Based on these results, I can more confidently state that there is no technical or economic barrier to transitioning the entire world to 100 percent clean renewable energy with a stable electric grid at low cost,” Jacobson says.

“This solution would go a long way toward eliminating global warming and the 4-7 million air pollution-related deaths that occur worldwide each year, while also providing energy security.”

The modelling assumes a phenomenal amount of wind, solar and other technologies to be built over the next few decades – some 18,000GW of wind, 10,000GW of rooftop PV, 16,000GW of utility-scale PV, and 2,850GW of solar thermal.

In Australia, for instance, it assumes around 110GW of wind, 140GW of rooftop PV, 106GW of utility-scale PV and 47GW of solar thermal.

If this sounds immense, then it is because Stanford assumes that everything becomes electrified. The total load becomes 108GW, with around one-third of that considered to be “inflexible” and two-thirds “flexible”, which includes demand response, stored heat, and hydrogen storage.

How much would all this cost? In Australia’s case, $1.4 trillion in 2013 dollars, or a delivered energy cost of less than 9c/kWh (electricity only, in the low scenario), which is about the price paid over the last year or two.

And if that sounds improbable, it is worth noting that companies like Nectar Farms are already planning to switch from gas to electricity for their immense greenhouses, and run on 100 per cent renewable energy for electricity.

Not only that, electric vehicles are capturing the imagination of consumers, electric trucks are being produced, electric ferries are now in use, and even electric planes are being considered.

The latest paper, which builds on a previous 2015 study by Jacobson and colleagues that examined the ability of the grid to stay stable in the US, addresses some of the criticisms raised last year.

Most of those were centred on the single scenario, and accusations that it relied too much on adding turbines to existing hydroelectric dams.

Hence the different scenarios in the latest paper, including ones with no added hydropower turbines and no storage in water, ice, or rocks.

“Our main result is that there are multiple solutions to the problem,” Jacobson says. “This is important because the greatest barrier to the large-scale implementation of clean renewable energy is people’s perception that it’s too hard to keep the lights on with random wind and solar output.”

The new study matches supply and demand in 30-second increments for five years (2050-2054) to account for the variability in wind and solar power as well as the variability in demand over hours and seasons.

The modelling includes programs that predict global weather patterns from 2050 to 2054, adapts this to predict the amount of energy that could be produced from wind and solar, and overlays it with power from hydro, storage, and other “firm” sources like geothermal, tidal and wave devices.

The average contribution from wind and solar totals more than 90 per cent in the main scenarios, with storage in hydro, batteries, hydrogen, and solar thermal, and considerable amount of demand flexibility in transport (charging and discharging) and industrial uses.

“The fact that no blackouts occurred under three different scenarios suggests that many possible solutions to grid stability with 100 percent wind, water and solar power are possible,” the report says.

That, it notes, is a conclusion that contradicts previous claims that the grid cannot stay stable with such high penetrations of just renewables.

Interestingly, apart from the cost savings from health and climate impacts of fossil fuels, the study also estimates that reduced water vapour from the estimated 3.5 to 5 million wind turbines would offset about 3 per cent of global warming to date.

Jacobson and his colleagues say a remaining challenge of implementing their roadmaps is coordination required across political boundaries.

“Ideally, you’d have cooperation in deciding where you’re going to put the wind farms, where you’re going to put the solar panels, where you’re going to put the battery storage,”

“The whole system is most efficient when it is planned ahead of time as opposed to done one piece at a time.”

It says only water, wind and solar technologies were used in that study, as they provide greater air pollution, health and climate benefits than bioenergy or fossil fuels with carbon capture and sequestration (CCS).

They also use less land than crop-based bioenergy; and result in less catastrophic risk, weapons proliferation risk, waste, and delays than nuclear power.


  • Farmer Dave

    “Reduced water vapour from … wind turbines ..” ? Presumably, this is about not releasing water from the combustion process, and is therefore a benefit of phasing out fossil fuels, regardless of how that is achieved. Putting that interesting side issue aside, this is an interesting study, and thank you for reporting on it. I heard about the criticism of their earlier paper, and it sounds like they have taken many of the criticisms on board.

    • JamesWimberley

      Jacobson et al had already shifted, between the initial US study and the first multicountry one, away from burst hydro and in favour of CSP with hot salt storage. After Copiapó this was a no-brainer: 6.3c/kwh for 24/365 supply. The one issue with CSP is that it’s only workable in very dry and sunny deserts or near-deserts. The US Southwest qualifies, but not the rest of the country. It’s not like PV where clouds just lower the yield, with CSP they stop it dead. Of coursem you can stil use the molten salt technology as a straight megabattery. Maybe this is what Jacobson et al are using for higher latitudes>

      BTW, a big black mark to J for publishing work of the very highest importance to public debate in a paywalled Elsevier journal. Commercial publishers of science journals are pure vampire parasites and should be driven to extinction. There is no reason why all research should not be open access. The real work – the research, writing it up, proofreading, internal and external review – is all done by scientists as part of their daily jobs. The housekeeping of a journal – mailing list, subscriptions at cost, archives, paper copies for the odd paper reader – is a trivial administrative chore that any university library can handle. Google the Cambridge mathematician Tim Gowers and the Lingua/Glossa rebellion in linguistics.

  • JamesWimberley

    Jacobson’s plea for international coordination by central planning is politically unwise, as it will rouse fierce ideological opposition, and doubtfully justified. There is no central planning body requiring interconnectors between Norway (with plentiful hydro) and the UK (with plentiful wind), and they are still being built, because both parties have done their sums. The same goes for the expanded trans-Mediterranean links being studied. The market will determine whether these are worthwhile, in the light of relative prices of North African CSP, Italian PV solar plus storage, and the undersea cables. India and Pakistan could probably save money by cooperation, but they won’t. The best we can hope for is a widespread push by influential bodies like the World Bank for more interconnectors and the effective spot wholesale markets they enable.

    Another disadvantage of central planning by technocrats is that is historically it has been biased towards megaprojects, like Brazil’s huge remote dams or the East Asian supergrid being considered, and against superficially chaotic distributed generation close to the point of consumption. Hayek would be all for microgrids.

  • Matt

    “The modelling includes programs that predict global weather patterns from 2050 to 2054”.

    You’ve got to be kidding! The IPCC can’t accurately predict climate over the next 5 years, let alone a window more than 30 years into the future. How can anyone take this seriously??

  • Tony Yen

    Were I an energy transition skeptic, I would now begin to find some trivial errors (but in fact it is just a scenario assumption you consider to be implausible) in this study and write a paper for that “finding”. I could then claim that I had proved that 100% RE is impossible, regardless of the logic flaw behind this way of thinking.

    I think this was the debate on Jacobson’s 2015 paper. It’s kind of boring because a scenario is not a prediction, so I doubt anyone would think that Jacobson’s proposal is the only way to go to 100%. The point is that we need more renewables and we need to study more about how to integrate them as much as possible onto the energy system. Jacobson’s scenario will probably not become the exact reality by 2050, but what harm does that make? We now know that there are several ways to get nearly 100% RE, and we now know modeling a 100% RE system is possible. In the next three decades we will come up with more accurate scenarios of the year 2050, and our modeling will have more practical use.