This group brings together the best thinkers on energy and climate. Join us for smart, insightful posts and conversations about where the energy industry is and where it is going.


US can run on 100% renewable power – but should it?

Seb Kennedy's picture
Founding Editor, Energy Flux newsletter

I am professional energy journalist, writer and editor who has been chronicling the renewables and fossil fuel energy sectors since 2008.  I am passionate about the energy transition, so much so...

  • Member since 2020
  • 103 items added with 54,499 views
  • Jul 30, 2021

Get 30 day free trial

It could be cheaper than you think to shift the US power mix to 100% renewables by 2050, but driving out the most efficient fossil fuel generators will almost certainly prove to be economically impractical – particularly if the transition is accelerated. That’s the headline conclusion from a detailed new study that seeks to quantify the challenge of reaching a 100% renewable energy power system for the contiguous US.

Subscribe now

The academic paper (see full reference below) simulated pathways for achieving up to 100% RE power systems for the Lower 48 states. It found that a 100% RE mix could be achieved at an average CO2 abatement cost of just $61 per tonne.

Moreover, the overall system cost – that is, the cumulative cost of building and operating the bulk power system assets over the model horizon to 2050 – rises only marginally compared to preserving the current power mix at today’s ~20% RE penetration.

Replacing like-for-like gas, coal and gas-fired capacity retirements out to 2050 implies a cumulative system cost of $3 trillion, according to the modelling. For the same cost, the power system could achieve 90% RE penetration, while 100% RE implies only a marginal average cost increase.

This is due to the huge fuel savings achieved by phasing out fossil fuel thermal generators, which is offset against higher capital costs and transmission upgrades required to accommodate variable-output RE sources and utility-grade battery storage.

The 57% scenario is the baseline, since it would be achieved by implementing existing policy measures. A key finding is that shifting from the current mix of ~20% RE to 57% by 2050 could be achieved at a negative system cost.

This means that 57% RE by mid-century is the optimum mix from a system cost perspective, and any incremental increase of RE penetration beyond that implies a cost burden compared to achieving 57% by 2050.

That’s because more efficient or lower cost generators, some of which would not have yet paid off their initial investment, would be retired early (i.e. stranded assets).

However, when compared to today’s 20% RE mix, achieving any penetration up to 90% would offer system-wide cost savings.

The incremental cost of abatement increases sharply the deeper RE penetration goes. To get from 99% to 100% RE, the incremental CO2 cost more than doubles to $930/tonne.

This cost applies to only a tiny fraction of the eliminated carbon, which is why the average (i.e. levelised) cost of abatement at 100% RE is much lower ($61/tonne).

Still, this very high cost of abatement means it will almost certainly be most cost-effective to reduce carbon in other non-power sectors of the energy economy, for example heavy industry, heating or transport.

Subscribe now

The crux of the challenge is providing adequate year-round resource capacity as reliance on wind and solar rises. Even in a 97% RE scenario there would be a requirement for some coal-fired capacity, and gas-fired generation is still present at 99% RE.

Displacing these generators, which would by necessity be the most efficient and cheapest in class, would require widespread deployment of combustion turbines running on renewable fuels: thermal generators running on green hydrogen, biofuel or carbon-neutral biogas. These are referred to in the study as RE-CT.

The greater capital expenditure implicit in RE-CT deployment outpaces reductions in fuel costs and other power system operational costs, the study states:

“[W]hen reaching a 100% system, the costs are significantly lower if there is a cost-effective source of firm capacity that can qualify for the 100% definition. The last few percent cannot cost effectively be satisfied using only wind, solar, and diurnal storage or load flexibility, and thus other resources that can bridge this gap become particularly important.”


Known unknowns

The paper identifies several key factors that exert great influence on the cost of shifting to a 100% RE power grid.

Changing the definition of 100% RE can have as large an impact on system cost as other sensitivities such as battery costs or natural gas prices.

If the 100% requirement is defined as a fraction of final electricity sales rather than generation, the cost and emissions of meeting that requirement are similar to those of scenarios with lower RE penetration.

That’s because transmission, storage and curtailment losses can be as high as 10% of generation.

Similarly, including non-renewable power sources such as nuclear and thermal capacity equipped with carbon capture can sway overall system costs in both directions.

And, the quicker the transition to 100% RE the more expensive it becomes – but the greater the CO2 savings.

Hitting 100% RE by 2030 implies a levelised abatement cost of $95/tonne (compared to $61/tonne if the deadline is 2050), but cumulative power sector emissions would be 38% lower by mid-century.

These conclusions should be of particular interest to the Biden administration, which has pledged to achieve a carbon-free power sector by 2035. It is highly likely that this will include nuclear and CCS, rather than 100% RE.

There are other factors that could prove equally decisive, but were not included in the study.

These include the potentially higher cost and constraints of scaling up supply chains to reach 100% sooner than 2050; the social or environmental issues related to siting new assets; and the evolution of system costs after a penetration target has been achieved and stranded assets have been resolved, among other factors.

Leave a comment

Study reference: Cole et al., Quantifying the challenge of reaching a 100% renewable energy power system for the United States, Joule (2021),

Article updated on 1st July 2021 to clarify paragraphs 5-8.

Matt Chester's picture
Matt Chester on Jul 30, 2021

This means that 57% RE by mid-century is the optimum mix from a system cost perspective, and any incremental increase of RE penetration beyond that implies a cost burden compared to achieving 57% by 2050.

This is an interesting result-- of course because of many externalities it's worth looking at more than just the cost perspective, but so many commitments we're seeing by utilities, governments, corporations are shooting for well beyond these levels. But they'll obviously take us through 57% first, and probably sooner than 2050, so I'll be curious how the numbers shift once we've already achieved that milestone

Seb Kennedy's picture
Seb Kennedy on Jul 30, 2021

Let's see how 57% pans out first Matt. Gaining experience operating grids with high VRE penetration will be crucial to underpinning confidence at conservative utilities. They need to be reassured that there is merit in going beyond the ideal system cost-benefit threshold.

Patrick McGarry's picture
Patrick McGarry on Jul 30, 2021


Thanks very much for sharing this insightful post. I also read the white paper.

I had heard much discussion about the nonlinear costs for the last few percentages towards 100% ROE, but I had never seen it quantified in this manner.

The study made me think of another question outside the scope of this paper, but not totally detached from the topic. Which are costs.

What does system reliability look like in a 100% RE world. I would imagine the percentage of likelihood of loss of load would rise, but not sure how much. 

If you took the new percentage of loss of load and compared it to today's world as a reference, what would be the cost?

If you knew the cost, would it be appropriate to add it to the costs outlined in this paper?


Patrick McGarry

Seb Kennedy's picture
Seb Kennedy on Jul 30, 2021

Thanks Patrick. You're right, reliability is beyond the scope of this report, albeit their modelling does account for a sufficient volume of firm capacity in all scenarios to keep the lights on. I would imagine that any increase in likelihood of loss of load occurring would be reflected in a higher overall system cost, which is borne out via either 1) loss of load events happening, or 2) investment in dispatchable RE generation that has a very high LCOE and upfront capital cost in order to mitigate the risk of those events occurring.

Bob Meinetz's picture
Bob Meinetz on Jul 30, 2021

Seb, why do you continue to publish NREL studies that repeatedly cite debunked work from Stanford's gas-funded Precourt Institute of Energy?

Attractive though the idea may be, credible evidence supporting the idea the "US can run on 100% renewable power" does not exist - period.

  • Of 191 in the world, two countries come close: Iceland and Norway.
  • Both have an abundant natural resource: geothermal and hydropower, respectively.
  • Both are completely lacking in quantity and quality in the U.S.
  • Neither are scalable, i.e., can reliably grow to meet increased demand.
  • The renewable resources in these countries serve less than 1% of the world's population - and among the wealthiest.

The U.S. would be 100% reliant on solar and wind to achieve that goal. Given not a single country, state, town, or community worldwide is powered by solar, wind, and storage, wouldn't you agree a first step would be to see if it's possible - even on a smaller scale?

To continue support for this outlandish, evidence-free hypothesis, IMO, is the height of irresponsibility.

Seb Kennedy's picture
Thank Seb for the Post!
Energy Central contributors share their experience and insights for the benefit of other Members (like you). Please show them your appreciation by leaving a comment, 'liking' this post, or following this Member.
More posts from this member

Get Published - Build a Following

The Energy Central Power Industry Network is based on one core idea - power industry professionals helping each other and advancing the industry by sharing and learning from each other.

If you have an experience or insight to share or have learned something from a conference or seminar, your peers and colleagues on Energy Central want to hear about it. It's also easy to share a link to an article you've liked or an industry resource that you think would be helpful.

                 Learn more about posting on Energy Central »