Join the discussion

Write your take first — we'll ask for email only when you're ready to publish.

  • Hacker News
  • The future will be solar. Especially in underdeveloped regions without a centralized power grid, solar makes the most sense. And even in developed countries having solar panels on your roof and a battery in your cellar is the most cost effective way of getting all your electricity needs.
  • Nuclear, while exorbitant, is necessary for security. See North Korea vs Iran/Venezuela. Nuclear scientists, supply chains, and a trained workforce takes decades to develop. Excess power production can be used to smelt aluminum. You can always have strategic excess supply of materials that can use flexible supply.
  • With solar and batteries getting cheaper and cheaper, is this still a topic? Or at least for the time being until that "cheaper" gets cheaper enough?
    by soco
  • It is if the nuclear industry pays for it to be. See the disclaimer at the bottom of the article.
  • Scaling is a problem.

    How many batteries do you need to power, say, entire Scandinavia during winter? That would be a lot of lithium, btw.

    "Dunkelflaute" periods when it is dark and no wind to run the wind turbines are common in northern winters, IIRC the longest one a few years ago was 12 days long.

    Which means that in order to have a reasonable buffer against it, you would need enough batteries to supply the entire region for three weeks. Not going to happen, unless we discover some much more efficient class of batteries.

  • > In March 2025, the US Department of Energy (DoE) announced US $900 million in grants to support the deployment of SMRs. One year later, the European Commission said it would invest up to €200 million (US $228 million) in the construction of SMRs.

    Yes, it does seem to still be a topic.

  • Solar + battery is very land inefficient, that's basically the main argument.

    Edit: Per ChatGPT's calculation, nuclear is significantly cheaper than solar + battery in my country (Czech Republic) if we're talking about adding new reactors to existing power plants.

  • The problem with 100% solar+batteries is that you need to massively overbuild for the system to meet demand. It's far more cost-effective to have some portion of generation come from a firm, weather-independent source. The main options are fossil fuels, hydro, geothermal, and nuclear. Since fossil fuels emit carbon and hydro and geothermal are geographically limited, nuclear is the best option in a lot of places.
  • Although generally safe these reactors aren't totally safe.

    I'm concerned lax physical security will allow miscreants to blow them up and spread nuclear materials over a wide area. The ensuing panic would totally destroy any goodwill nuclear power has gained over the last 20 years or so.

    I'm also concerned these things will mostly be used by Big Tech for their A.I. data-centers playing the Good Samaritan with their claims of "Carbon Neutral Environmentally Friendly A.I."

  • Regardless, better that they're the ones who spend the money for these reactors, and better that they use nuclear instead of coal or gas plants.

    One option is to also mandatorily lock the powerplants to the municipal grid, and provide rebates to data center operators via net metering instead of simply giving them cash back.

    But yes, the lax security and the laissez-faire don't give a shit attitude of the current admin far outweighs any of these benefits

  • > A.I. data-centers

    I wouldn't be so sure about that (in the US).

    Like usage in A.I. data-centers assumes

    that they get price competitive with low quality gas turbines run by state subventioniere natural gas...

    or the government actual enforces proper environmental protections so that they can't continue running gas turbines without proper filters (and this isn't even (mostly) about climate change, but air pollution harming people)

    and from how it currently looks both things seem unlikely :/

  • > I'm also concerned these things will mostly be used by Big Tech for their A.I. data-centers

    Why is that a concern?

  • > Although generally safe these reactors aren't totally safe.

    I have some seriously bad news about the safety of your totally-safe house against meteorites. Or for that matter, thermonuclear explosives detonated directly above its roof.

    Now tell me: Is Your Car Safe From Supermaneuverable Air-Defense Fighter Aircraft? https://www.youtube.com/watch?v=k7YVxLLIuGM

    As a bonus, here's a channel with hundreds of commercial aircraft accidents in which thousands of people have died: https://www.youtube.com/@MentourPilot. Now enjoy traveling using the objectively safest mode of transport — commercial aviation.

    by Mawr
  • "Miscreants" conjures in my mind an image of little British boys with dirty feet in Industrial Revolution era attire.
  • Terrorists are only one piece of the issue. Corporations and "Responsible Adults" have done more to harm nuclear power in this country than any terrorist act has.

    ConEd and Three Mile Island meltdown brought out a mass of normal people to chant "Hell No! We won't Glow!"

    Hanford in Washington State.

    San Onofre in So Cal.

    Santa Susana Meltdown.

    The reality is that everyone wants nuclear, but no one wants to live near a nuclear power plant.

    by bb88
  • Well, gas and oil plants are actively emitting pollutants that kill thousands. And nuclear still is the safest energy sort available.
  • There are no working SMR's ready to be deployed in the West. The company (NuScale) that was furthest along had its flagship project die after cost estimates ballooned. Its other project was a crypto scam.

    Whether any of these companies work out and make it to serial production is very speculative.

  • Are you forgetting Valar Atomics exists buddy? Or Aalo, or Antares or Radiant?
  • The SMR in Darlington, Canada is already in construction.

    https://www.world-nuclear-news.org/articles/darlington-smr-n...

  • "a supplement produced with financial support from Oklo Inc"

    "Oklo Corp. Logo Oklo is designing and deploying advanced fission power plants to provide clean, reliable, affordable energy"

  • A lot of people are betting serious money that they will cheaper, more flexible, faster to build, and safer. Including me - I'm working in the industry, and I think they might very well be future backbone of energy production.

    But there's a real possibility when all this shakes out that SMR's only advantage will be their flexibility. And that might be enough. A major issue with gigawatt scale nuclear is that it's frankly too big for most markets. Only very large electric markets can can easily digest a new always-on 1000 megawatts of electricity, and you need to be building multiple plants at a time for this all to be economical. That's why the nuclear power rollout of the 60s through 80s worked, and why China and to a lesser extent India's nuclear industry is thriving presently.

    With an SMRs smaller scale, there are just way more available projects where nuclear is feasible, and so a more consistent workload can keep everyone employed and subcontractor's backlogs filled. The flexibility in scaling lets you reclaim the benefits of having an experienced workforce and that knows how to build nuclear power plants, something we lost in the west when we stopped building them.

  • Big reactors also need a lot of cooling, and we've just had a very dry summer in Europe which forced some nuclear power stations to dial down or shut down entirely.
  • The problem is that you can’t assume ”always on” anymore.

    Why should consumers choose expensive nuclear powered electricity when cheap renewables, or stored renewables are available?

    They don’t and now capacity factors crater.

    Leading to what was once seen as ”baseload” plants being forced to become peakers. And running a nuclear plant with those fixed costs as a peaker/firming becomes stupidly expensive per MWh produced.

    See this Australian ”baseload” coal plant forced into a peaker role or be decommissioned.

    https://www.abc.net.au/news/2024-10-13/australian-coal-plant...

  • What % of global electricity production is in too small markets? Seems to be working well in EU at least, markets and grids being multinational.

    I'd guess many places in Africa could be potential markets that have underdeveloped grids but growing economies and populations. Hence the SMR industry in South Africa since the 90s I guess.

  • I wish the article had more details on the cost to build these reactors, the cost to operate, refuel, dispose of fuel and then decommission. Details about how long they’re expected to operate, and how much power they will generate in their lifetime. Then we could get a projected cost per kWh for the power they will generate.

    I ask because the last reactors the US brought online took so long to build and cost so much they caused the cost of power to go up.

    This seems unworkable when solar is already causing power to be free (Australia), and gets cheaper by the day.

  • There's two reasons they take so long: large plants need to be redesigned as conditions on the site develop (soil type unexpectedly changes as excavation occurs) or regulations change during the process, and the second reason is that there's basically no nuclear construction industry.

    The benefit of an SMR is that it can theoretically be built largely off-site without the same risk that larger plants have wrt needing to customize everything. That allows construction companies to centralize and reuse talent and suppliers rather than needing to drag them all over the country or subcontracting to people more accustomed to building offices.

    In areas like where I live, there's weeks or more of time in winter where every day is overcast, short, and has little wind. To go completely renewable, we would need to pay southern regions to overproduce by triple digit percentages of their needs and then send the energy to us by HVDC that don't exist yet either.

    A bad ice storm or tornado could leave millions without any power at all at the most dangerous time of the year to not have electricity.

    It's doable, but we are a long way off and if SMR miraculously hits the right numbers, it'd be a better fit for us. Until then, we'll keep burning gas supplemented with solar.

  • > but SMRs should be cheaper and easier to construct.

    The word should is doing some real heavy lifting there. Especially given the subject.

    I guess we'll know in 2030 when one should come online.

    by hvb2
  • Except Valar already did it earlier this year.
  • We should know by then, at least.
  • It just stupidly expensive if it is completed on time and budget.

    To make the figure at least palatable for public funding the project assumes enormous learning effects for subsequent reactors.