Off the Siberian coast, not far from Alaska, a Russian ship has been docked at port for four years. The Akademik Lomonosov, the world’s first floating nuclear power plant, sends energy to around 200,000 people on land using next-wave nuclear technology: small modular reactors.

This technology is also being used below sea level. Dozens of US submarines lurking in the depths of the world’s oceans are propelled by SMRs, as the compact reactors are known.

SMRs — which are smaller and less costly to build than traditional, large-scale reactors — are fast becoming the next great hope for a nuclear renaissance as the world scrambles to cut fossil fuels. And the US, Russia and China are battling for dominance to build and sell them.

  • @[email protected]
    link
    fedilink
    810 months ago

    If I’m not mistaken SMRs also handle power demand shifts better and don’t have to just do a base load. Something very useful with the growth of renewables and how they are not always supplying power.

    • KillingTimeItself
      link
      fedilink
      English
      210 months ago

      I doubt it. Unless they have power storage of some kind, like SSR designs where they use a thermal battery of some kind.

      The fundamental issue with nuclear power is that it produces a fixed output (which falls over time) which cannot be managed. Aside from just deleting what would otherwise be power (which is where the power storage comes into play)

      It’s not impossible though, but then again it’s not impossible for any nuclear plant to store energy.

      • ZooGuru
        link
        210 months ago

        The small reactors on submarines can maneuver very quickly without causing fuel damage. Less power per core = less heat generation. Large reactors are limited by flux rate because they can have such high localized heating during maneuvering which has the potential to damage fuel. In that sense, SMRs could raise and lower power to meet demand or even operate on full power/standby basis like what gas plants offer during peak load.

        I can’t speak to the strategy of an electric utility using SMRs, but to your point, I would think the idea would still be base load. Build a site with the potential for more SMRs to be built to meet demand in the future.

        • KillingTimeItself
          link
          fedilink
          English
          110 months ago

          ok so i get what you’re saying here.

          But there is a fundamental thing with nuclear power, where the “burn up” of nuclear fuel doesnt change. In a submarine it doesn’t matter because you’re backed by a military force and you use 70-80% or 90+% enrichment, where as on land we have 3-5% upwards of 20% for the higher enrichment stuff these days i believe.

          In the water its about safety and ensuring power production, on land it’s about ensuring reliable and efficient power production. The only beneficial way of doing this is electricity storage. If you’re nuclear reactor isn’t producing power and has fuel, you are quite literally burning money. Think about it like diverting gas/coal input into a gas/coal fired power plant when power demands lower, as opposed to just lessening the consumption.

          But yes it would be about 100% baseload first and foremost, everything else is a future concern, eliminate as much static load as you can and then deal with the rest in other manners.

          • ZooGuru
            link
            210 months ago

            Yeah I’m with you. I have a senior license at a US nuclear plant just for some background as I don’t know yours. What I’m saying is that I can see value of multiple, say 300MW SMRs at a single site, that can go from 0-100% very quickly compared to current 900-1100MW reactors. So the idea would be you could have a plant in Mode 3 Hot Standby ready to raise power for peak loading. Ideally you’d have at least one reactor online at all times that provides its own in house loads and the standby in house loads that would be quite low. That is the value I see.

            The issue at that point would be refueling and maintenance outages. It seems ideal that the design would need to support online refueling and enough loops/system availability to do the majority of plant maintenance online. In addition, the regulatory landscape has a lot of momentum to allowing plants to move to risk informed tech specs which allow for major equipment outages in modes of applicability. If the industry as a whole can agree on a handful of SMR designs with multiple capacity options, it really could be a stop gap to hopefully fingers crossed fusion power in, I don’t know, 50-100 years from now? My two cents.

    • @[email protected]
      link
      fedilink
      -210 months ago

      Renewables being unable to do base load is just a myth that has been debunked countless times.

      • KillingTimeItself
        link
        fedilink
        English
        210 months ago

        renewables can theoretically do baseload. The problem with renewables is that they don’t really have a good pairing with something that would make it SIGNIFICANTLY easier to do.

        Nuclear and solar power would make a great pairing for summer time midday peak draws for example. Wind is a good supplementary source. Hydro is a good stored energy source.

        You can definitely do full renewable but it will still inevitably be better complimented by some form of baseload plant (i.e. nuclear)

      • @IchNichtenLichten
        link
        English
        110 months ago

        I’ve love for just one of the people anonymously downvoting to chime in. What you wrote is completely accurate but every nuclear-themed post here and on Reddit is downvoted without anyone putting forward a counter-argument.

        • @[email protected]
          link
          fedilink
          English
          510 months ago

          here https://www.pnas.org/doi/full/10.1073/pnas.1610381114 we can talk about this, feel free to put forward counter arguments, the gist of the cited paper is that previous studies claiming 100% renewable baseload is possible requires sketchy manipulation of the expected demand as well as currently unavailable storage technology on an almost impossible scale. We’re working on all kinds of storage solutions but the reality is we’re not there yet. I’m rooting for molten salt storage or compressed gas storage rather than ramping up more lithium battery storage. Flow batteries are promising as well, but in any case we won’t have enough storage or transmission capability to have a 100% renewable baseload in the next couple of decades.

          • @IchNichtenLichten
            link
            English
            -410 months ago

            I don’t think it’s astroturfing, it’s just cognitive dissonance. Lots of people were raised thinking that nuclear power was the future and they can’t let go of that. That’s why they downvote without commenting - there’s no factual case for new nuclear and that goes double for SMRs.

            • KillingTimeItself
              link
              fedilink
              English
              610 months ago

              there’s no factual case for new nuclear and that goes double for SMRs.

              there absolutely is. It’s a good transitional source of power that we currently understand very well, and know how to manage, but simply cannot build. It would be a very prudent way of ensuring some “insurance” time before fusion starts being even remotely viable.

              Although i don’t think SMRs are the correct answer here.

              • @IchNichtenLichten
                link
                English
                010 months ago

                It’s a good transitional source of power

                Not with the design and build times new nuclear has. It can take 10-15 years to build a plant, and during that time costs will usually spiral and schedules will slip. At the same time, renewables and storage will have gotten even more competitive.

                • KillingTimeItself
                  link
                  fedilink
                  English
                  110 months ago

                  this is true, but nuclear plants are slated to run for 30-50 years. France has been running their existing fleet to 50 years with maintenance extensions.

                  There was a recent plant vogtle, i believe, that was finished. Although if im not mistaken i think they just stopped midway through that one, it is up and running right now though last i checked, maybe not generating power yet but definitely running.

                  I’m guessing you’re referring to the flammanvile reactor project in france? If so thats an EPR design, which are horrendously complicated, and the vast majority of the issues present in the construction are the inability to pour concrete correctly, and the inability to weld correctly. Which is something that happens after 30 years of not building any nuclear plants. We quite literally just have to build more if we want to be able to use it.

                  It’s true that renewables are more competitive, but solar requires significant power storage figures, which can be problematic at best. Or require other production methods to take up the slack. Wind is quite good, but has the significant problem of waste. Turbine blades are a huge mess. That’s mostly due to industry pressure to make it profitable, and the push for it to succeed, which nuclear hasn’t seen. Nuclear just needs the same thing.

                  • @IchNichtenLichten
                    link
                    English
                    310 months ago

                    No matter which angle I approach the topic from, it always comes back to this:

                    “findings suggest that the cost per kilowatt (KW) for utility-scale solar is less than $1,000, while the comparable cost per KW for nuclear power is between $6,500 and $12,250. At present estimates, the Vogtle nuclear plant will cost about $10,300 per KW, near the top of Lazard’s range. This means nuclear power is nearly 10 times more expensive to build than utility-scale solar on a cost per KW basis.”

                    https://www.energysage.com/about-clean-energy/nuclear-energy/solar-vs-nuclear/

                    I just don’t see how this makes any economic sense. Sure, we could go all in on new nuclear and it would work fine but I don’t want to pay for that, I want cheap renewable power.

                    Then there’s this:

                    https://www.reuters.com/business/energy/high-river-temperatures-limit-french-nuclear-power-production-2023-07-12/

                    Building a plant with a lifecycle of 30-50 years seems like a bad idea when our world is getting more and more unpredictable. We’ve got climate change, we’ve got Putin fucking around with 6 reactors in Ukraine, earthquakes, tsunamis, human error, etc.

                    If a wind turbine catches fire, it’s not that big of a deal.

                  • Sir_Osis_of_Liver
                    link
                    fedilink
                    210 months ago

                    Vogtle 3 & 4 are AP1000s. Construction started in 2013 (preliminary work had started before this, but a design change halted it). Unit 3 was originally supposed to complete commissioning in 2017, but only happened last year. Unit 4 should be online this year. The initial $12B budget went to $14B at the start of construction, but will end up somewhere over $30B.

                    V.C . Summer in South Carolina has a similar project with two AP1000s. The initial budget was $9B, but the project was cancelled while under construction when projections put the total cost over $23B.

                    There have been 6 EPRs built, Flamanville-3, Olkiluoto-3, Taishan-1 & 2, and Hinkley Point C (2 units).
                    All of them are/were massively over budget and behind schedule.

                    Olkiluoto started construction in 2005, was supposed to complete commissioning in 2010, but only came online last year. Costs went from €3B to somewhere over €11B, the contract ‘not-to-exceed’ amount.

                    Flamanville started construction in 2007, was supposed to complete commissioning in 2012, but is projected to complete commissioning late this year. Costs went from €3.3B to somewhere over €20B.

                    Hinkley Point C is still under construction. It’s difficult to put an actual start date because a pile of preliminary site prep work happened prior to real construction starting. Concrete was poured in 2016 though and it was supposed to be operational in 2023. They’re now estimating 2028 at the earliest. Costs have gone from £16B to and estimated £35B.

                    Taishan 1 & 2 started construction in 2009/10 and went online in 2018/19, roughly 5 years late. Unit 1 had to be taken offline for a year due to faulty fuel bundles. Both units have had reliability issues. Costs ended up at the equivalent of $7.5B, almost double the original estimate.

            • @MotoAsh
              link
              210 months ago

              No, it’s because it’s an off topic tangent. We’re talking about SMRs doing not-baseline. Not renewables doing baseline. The very fact they brought it up is indication of binary thought patterns like team sports thinking. “They are for this one thing I don’t like, therefore they must be against the thing I do like!” kind of thing. False dichotomy.

              Apparently it’s also false on top of that. Go figure.