Hacker News Reader

Comment by epistasis | Hacker News Reader

This article is more than three years old, but the story hasn't changed much. VC Summer was abandoned, leaving utility customers with massive bills. Vogtle soldiers on, but seems to be unlikely to finish.

Nucleae's problems are not the politics, the problem is that basic competency in design and construction logistics have been lost. Plans get delivered to construction that are "unconstructable," but construction soldiers on and wings it. Then it al has to go through design review again. And maybe redone. Delays delays delays. Incompetence abounds.

All the partners are planning from the beginning for a massive lawsuit at the end, and work harder to limit their liability (or create liability for others?) than to make the project work.

Executives lie about the progress, there are guilty pleas to fraud:

https://www.postandcourier.com/business/former-scana-executi...

20 years ago, I though nuclear was essential to fighting climate change. Today, I don't see how nuclear can ever help. We can't build it before it's too late, and by the time we build anything other technology has completely leapfrogged it.

We started these two AP-1000 reactors in 2008! A dozen years later we have nothing to show for it except for bankruptcy, plea deals, billions of dollars that would have been more effectively spent on solar and storage.

View on HN

Yup. For all the people boosting nuclear power, the reality on the ground seems to be that it just does not work anymore. It's way too expensive and never gets done, and this problem is global, not just in the US; see e.g. https://www.popularmechanics.com/science/energy/a33499619/fr...

At some point the results have to speak for themselves, and solar/wind plus battery storage is actually working and being built. That seems a more realistic path forward than nuclear power, which is demonstrably failing over and over.

Nuclear is superior in some ways, but battery and renewable is small, can be incrementally built, simple, safe, and is experiencing exponential cost reductions due to mass production scaling.

It reminds me of TSMC leapfrogging Intel because of mobile volume or PCs vs custom Unix workstations.

The incremental build is probably a big part of it. More volume and minor failures = more practice time and experience. How does someone "practice" building a nuclear power plant when failure can be so catastrophic?

Oh yeah... it's hard to get industrial scaling of expertise, production tech, and hands on experience when you're building a few monuments.

If we do make nuclear fission work it will be with modular reactors and standardized fuel form factors, etc.

Fusion won't fly even if we get over unity fusion unless we adopt a similar approach. Fusion reactors like to be big, but I can see them being composed of mass producible smaller modules and parts. Settling on a standard design that can be mass produced could make it work.

Thing is: as the nuclear people work to try to make anything above a reality, we are experiencing iterative rapid development in solar, wind, and batteries, with all of these getting geometrically cheaper until they approach the naked cost of the materials and the energy required to refine them... and the latter cost (energy) will drop as we get better at making these things.

In the long term I would not be surprised if Earth is almost 100% solar/wind/hydro/geothermal powered. You might see nuclear used in niche markets that combine high population density with low insolation and insufficient other resources and where geopolitical concerns make reliance on long distance power transport problematic. Earth is pretty close to a giant free fusion reactor that we don't need to worry about maintaining for at least two billion years. Nuclear power will be for space settlements further from the sun and deep space travel.

I could see Mars eventually relying on fusion in the far future. Mars is further from the sun, and most people who talk about settling there drastically underestimate the massive amounts of energy that will be needed to refine and purify literally every single material all the way down to the air and the water. The total energy footprint of your average Martian will be at least 10-20X your average Texan with a giant raised pickup and a 4000 square foot house.

Yes, but what will we do if/when wind and solar just "doesn't work anymore". In the US trains "don't work anymore"... but they do in other countries. This gradual erosion of technical capability is a generic problem the US needs to solve. Nuclear is a case study of what is wrong and where the problems are (political, business, education).

It's a good point, but I think we can fix trains more easily than nuclear power plants, because as you point out at least the trains are a solved problem in many other countries that would simply just emulate. And passenger trains have serviceable enough fallbacks (planes, Megabus, etc.), like solar/wind are proving to be fallbacks for nuclear.

This is a mischaracterization of the state of US rail. The US has prioritized freight rail over passenger rail. That’s not a terrible choice given the long distances and the lower sensitivity to slower, more efficient speeds.

The way the US chose to address accidents in rail was to armour plate their rail cars. In Europe they worked on better processes to ensure trains don’t run into each other in the first place. The erosion of technical capability in rail is driven by the legislature. You have many states that will need to work together to completely abandon the US regulations on rail, adopt European regulations, and replace every piece of rolling stock to comply with the new regulations. That’s quite the undertaking, but achievable since the required expertise still exists outside the USA.

The erosion of technical capability in nuclear power is likely unsolvable. There is no brains trust outside the USA from which to borrow the knowledge of metallurgy, concrete, site preparation, building design, or even design for manufacture. This is rebooting the entire industry from scratch to produce an unprofitable product.

Actually now the US has alternative compliance regulations that let operators use European stock without waivers: https://pedestrianobservations.com/2018/11/20/fra-reform-is-...

Now the problem is getting operators to actually buy such equipment, and then operators getting the slots and ability to run these at a usable frequency, speed, and reliability.

People have been saying for 30 years that solar/wind are going to take off and that we don't need nuclear.

We are now in a climate crisis, and solar/wind is still a fraction of the energy portfolio.

We als can't build nuclear anymore, because we have lost decades of experience.

I really hope you are correct this time.

The economics are changing though. An idealized stock market return is (used to be) 11.8%/year. This interest, compounded, doubles your position every 7 years. Once solar has a break even period of less than 7 years, the investment in solar beats a rosy, idealized stock market. When the payoff period drops to 4 years, you'd be an idiot _not_to do solar, because you're looking at an investment returning numbers in the credit card interest rate range.

I bought my current system for $45k four years ago. I only did it because the investment math worked in my case. (Yes, yes yes. It was for sure subsidized and the investment math _didn't_ work without the subsidies.) Now, however, that same system would cost me ~$20k. I checked when my coworker was looking into solar and wanted to run the numbers. The economics of solar are approaching beat-the-[idealized]-stock market investment. 4 years is a rule-of-thumb number for many businesses for investing in capital. That is, if a capital investment breaks even in 4 years, you pull the trigger. Once the economics hits 4 years for break even, you'll see a lot of CFO's start pulling the trigger, not to mention the encouragement from the marketing department for the optics.

Anyways, I think it is coming and for many of us, it is already here.

I love how people here are trying to say that technology simply does not progress forward for nuclear, as it does for every other technolgy.

It's a factual, descriptive argument. Nuclear takes so long to build and is so expensive that it's not really being built anymore, and thus the progress isn't happening.

It's expensive because it includes much more of the externalities in the price with much longer time-frames for ROI, as compared to other sources like coal or natural gas.

Nuclear as a power source is fantastic. One of the biggest operators is the US Navy which has been running reactors in military vessels for decades without incident. The problem is politics, ignorance and lack of will to actually see it through, not fundamental science or tech.

> One of the biggest operators is the US Navy which has been running reactors in military vessels for decades without incident.

AFAICT, none publicly acknowledged since a bunch in the 1970s, but that may mean less than it seems since the US government has a history of covering up nuclear incidents for decades (like a whole slew at the Hanford site from the 1940s-1980s that weren't acknowledged until the 1980s.)

Hanford was for nuclear production, not power generation, and was created during the early days of nuclear weapons research.

Modern reactors are in an entirely different league with an incredible safety record. [1] Considering how easy it is to detect nuclear failures and the consequences in a tightly-enclosed ship, it's unlikely that there are any serious incidents. It's clear that they're experience and technology is continuing to get better every year though.

1. https://www.forbes.com/sites/jamesconca/2019/12/23/americas-...

> Hanford was for nuclear production, not power generation, and was created during the early days of nuclear weapons research.

I'm not saying that Hanford has the same risk profile as Naval reactors, I’m saying that it demonstrates that the US government has a propensity for covering up nuclear incidents that it can get away with covering up (note that it also acknowledged other incidents at Hanford in the decades it was concealing many of them), so the fact that it hasn't acknowledged any related to naval power generation since a bunch in the 1970s is absence of evidence of such incidents, but not reliable evidence of absence of such incidents.

> the problem is that basic competency in design and construction logistics have been lost

I don't think this is just a nuclear problem. This is a civil engineering problem more broadly. IMO, the exact same forces which make nuclear impractical now are also responsible for the horrific budget explosions and overruns in big projects all over the US.

Yes, I think you are right, and it's a damn shame because there are so many infrastructure projects that we desperately need: high speed rail, commuter trains, subways, inexpensive housing builds with high du/acre and walkable designs...

My guess is that by the time we have regained any such capacity, renewable and storage will be cheaper than the thermodynamic process of harvesting heat and turning it into electricity; a process that has been hyper optimized for a century without much progress, and which has hit the wall of thermodynamics in efficiency. The only way to improve its cost would be far cheaper materials, a revolution which might happen with 3D printing, but which is super speculative today.

While I think we need efficient housing designs, I don't think we need high speed rail and commuter trains in the US. As much of that functionality as possible should be replaced with internet connectivity. There is no reason to be spending enormous sums per mile when we can find ways to destroy demand for moving atoms instead of bits. The cheapest mile of high speed or commuter rails it the mile you didn't build.

Efficient homes, renewables, ubiquitous internet connectivity, and electric vehicles are going to be your targets as people spend more time at their homes and the trips they make can be electrified.

(the above comment is predicated on developing countries aging rapidly causing workers to leave the workforce, the quick uptake of remote work due to the pandemic forcing employers to learn how to support it, and data showing people fleeing cities for the suburbs and lower density living arrangements in general)

I believe civil engineering part of nuclear reactor is it's easiest parts.

quod erat demonstrandum

Logistics isn't easy. Dirt isn't simple stuff, neither is waterway management, nor concrete. Hauling a giant steel pressure vessel from its point of manufacture to point of installation isn't easy, nor is its installation. Just because it looks simple to a layman doesn't make it simple.

It's a set of "normal" problems like welding pipes, made much harder by the need for zero defects.

I am not really familiar with the space, so please pardon the stupid question..

Why can't we use an affordable safe/competent reactor design from 30-40 years ago, and stamp that out at scale, rather than building a "new" model that results in failure?

Over the previous 30-40 years we found out what the safety problems with the old designs were. PWRs can fail badly if they run out of water or leak under pressure.

However, they're also the main type of reactor that actually got built and works.

If you look at the reactors that were completed 30 years ago, it looks like there are some affordable, safe, predictably-built reactors among them. But when you look at all the reactors that were under construction in the years prior to that, you will find that a lot of them were also late, over budget, and/or never completed. Even in the 1980s there was a large variation in cost and schedule between reactor projects. Some of the projects -- in hindsight -- were plagued by bad reactor designs. In other cases there's nothing obviously better/worse about the design-on-paper, but there was a significant difference in the quality of project management and execution.

You can't examine the differences in project management and execution while the project is still on paper. You have to commit years and billions of dollars to determine if a project is going to adhere to the original plan or blow out on time and/or money.

Right now South Korea has the best track record of rapidly, affordably building reactors. Some American nuclear proponents think that we should just have South Koreans build Korean reactor designs here, to bypass the apparently irremediable problems in the domestic nuclear industry.

But even the smooth progress of South Korean reactor builders is not assured when they are building outside of Korea.

The Barakah Nuclear Power Plant, built by Koreans with Korean reactors, just powered up its first reactor days ago.

https://en.wikipedia.org/wiki/Barakah_nuclear_power_plant

The timeline from construction start to operation was originally 5 years. It has actually taken 8 years. It's 60% over the time originally scheduled, even though it is using a standard reactor design built by the most competent reactor builders currently available.

(In addition to what pjc50 said.)

TIL Barakah NPP! That is interesting, considering the location, which tends to be very hot at times. So I'm wondering what they are doing about the things French and German NPPs do in hot summers. Shutting down, or going lower power because otherwise cooling water would make the rivers too warm, and fish would die(in masses). Evaporation? With what water? Desalinated(first)?

edit: Never mind. The system inherits elements from a design used in the Palo Verde NPP near Phoenix, AZ

Probably know how to handle the heat...

As someone who left the field of Civil Engineering out of frustration with the state of the industry, I largely agree, but for a different reason.

I have noticed that there is something like "regulatory capture" in the construction industry now, but not by the same methods as most other industries. With giant, multi-national corps handling most large construction projects now, the QA/QC engineers are either incorporated directly into the company and thus never stop terrible practices, or they are there as a sub-contractor to the prime contractor and thus will be replaced for 'other reasons' if they give push back. It's now a push to the bottom for quality and the industry is going to hell quickly as the Engineers who value quality and ethics are slowly pushed out or forced to submit. If a small engineering company tries to stand up to one of these behemoth construction corps, the smaller company will either be blackballed behind the scenes (if they're lucky enough to be temporarily successful in their challenge) or the behemoth will lean on the nearest municipal gov't and say something along the lines of "We would love to finish this critical project, but we're being held up by this silly requirement and this silly engineer. (implied: Since we're on a Cost Plus contract) It's costing you X,XXX,XXX per day while we're held up. You should probably give us a way to get past this issue."

It would be funny if it weren't so sad. Civil Engineers are getting crapped on by both sides for increasing costs. They are blamed for holding things up with quality checks and at the same time for not doing enough to stop the terrible practices that result in long-term failures which end up increasing costs further. Payouts are going down for QA/QC companies and the pressure is constantly increasing to let things slide. The differences between a well-regulated project and an 'unregulated' (for all intents and purposes) project are shocking - you can feel (and see) an attitude of IDGAF from everyone on projects where it is known that the regulations don't REALLY apply.

Its possible. I've managed to work in a couple of organizations with strong QA/QC. In both cases, the team responsible for QC had autonomy with respect to product delivery.

I kinda of feel like nuclear reactors are huge and complex and tax our current technological and organizational abilities to design and manufacture. There are other things like that but the downside risks are orders of magnitude lower. Example, friend of mine helped install a (I think) GE 7F turbine. Fired up and it threw a blade. And was down for a couple of months. Oh well. With nuke plants the stakes are higher.

And the root of this is essentially politics keeping competent people from getting interested in nuclear.

We can't do nuclear properly because society tells all the smart engineers and financiers that they should be doing other things.

A lot of "boring things" suffer from this problem.

I don't think that's true. We had the same construction problems in the 70s when we wouldn't have had such a shortage. There are tons of super smart people working on nuclear startups right now, with big funding.

It's all bad management, bad execution, and bad promises. And since few of any around the world have been able to keep construction promises recently, it may just be part of the natural variability of the construction process. And 2-3x variability is not a desirable trait for $10B projects, where most of the lifetime cost is in the upfront construction.

Yeah, you are right. The part that causes these projects to fail isn't the nuclear part. It's the construction part. You can't just take a construction crew that's used to building highways or skyscrapers and expect them to build a nuclear complex properly. This was the exact problem they had in the 70s, the contractors didn't understand why things had to be done a certain way, so they did it the way they thought it should be done. Then it turned out that the problems that many of the safety and operational issues that occurred later were due to not doing it properly in the first place. Nuclear is sexy, construction isn't.

There is a great book on the history of the US nuclear power industry that goes into this called Beyond Engineering: How Technology Shapes Society. The technical issues were only a small part of why nuclear power is not more widely used. Countries like France that standardized on reactor design and had enough projects that construction crews could gain experience with the different processes and requirements involved were much more successful in deploying nuclear power.

I live with in about 40 miles of Plant Vogtle and thats not really an accurate assessment. They have had to reshuffle some of the work due to COVID-19. but last I saw all of the major construction milestones on unit 3 are complete. its just down to ancillaries. Cold Hydro testing is slated to be conducted this fall and initial fuel loading by end of year. They are still on track for production to start Nov 2021 and 2022 for unit 4.

https://allongeorgia.com/georgia-business/georgia-power-anno...

I'm pro renewables but Asimov made an impact on how nuclear is the power of the future and it would be nice to have it in our portfolio.

The only likelihood of nuclear power succeeding would be to containerize the reactors to ensure build quality, keep costs down, and to address scaling.

A quick search hit this company that appears to be promoting this very thing: http://www.holosgen.com/

Nuclear is the future of power on Mars (containerized and launched on reliable person rated launch vehicles). It's clear we're not going to build it successfully on Earth, and that's okay. Renewables and batteries are faster and cheaper to build on Earth.

I think you are right, that the power of science fiction has clouded people's judgement of what the future will really be like.

Today, I don't think we can consider nuclear a good or modern energy system when compared to something like solar or wind. Nuclear is used as a heat source to boil water and power a turbine, and throws away half of its thermal energy as waste. disposing of that waste heat is becoming so difficult that reactors in France power down in hot days. This is not some sort of super futuristic power source!!

While small modular reactors are the only near term hope for nuclear, I think that fission is not a good fit for terrestrial power. And fusion is no better until we have direct conversion or some other massive tech leap.

Good point about heat waste disposal. That's why you almost always see nuclear reactors along lakes/rivers/oceans, which is bad for other reasons as we saw with Fukushima, and all of that hot water being released wreaks havoc with riparian/marine ecosystems. Nuclear has downsides beyond the most obvious one, that of the risk of fallout.

> I think that fission is not a good fit for terrestrial power.

Heh, it's even worse off-planet. If you think dumping the waste heat is hard when you have access to an atmosphere/large bodies of water, wait until you try it when you don't!! ;)

Palo Verde uses cooling ponds filled with grey water. Per Wikipedia, it’s the largest generation plant in the US. It’s not an impossible problem.

Around 2008-2010, ORNL published a paper describing how the cheap electricity and waste heat from a nuclear plant could be used to capture carbon from the air and create carbon neutral fuels for transportation. So again, if we were looking at a holistic picture, we could definitely improve the overall climate situation with nuclear.

How is there "waste heat" in a power plant that is designed to produce heat?

Thermodynamics!

> The efficiency of the Rankine cycle is limited on the cold side by the lower practical temperature of the working fluid.

https://en.m.wikipedia.org/wiki/Rankine_cycle

> The efficiency of even the best heat engines is low; usually below 50% and often far below. So the energy lost to the environment by heat engines is a major waste of energy resources. Since a large fraction of the fuels produced worldwide go to powering heat engines, perhaps up to half of the useful energy produced worldwide is wasted in engine inefficiency, although modern cogeneration, combined cycle and energy recycling schemes are beginning to use this heat for other purposes.

https://en.m.wikipedia.org/wiki/Thermal_efficiency

... the power plant is designed to produce electricity, not to produce heat. If it was designed to produce heat then it wouldn't need the turbines.

The heat is waste. For the same power generation, you want to minimize the amount of waste heat as much as possible.

Solar and wind are still not good at providing a baseline load, especially in winter where the power consumption is higher. France slows down some reactors during heat waves due to environmental regulations not allowing power plants to heat the river past a certain threshold, not because disposing excess heat is inherently difficult.

Why is baseline supply an issue? The only power supplies that are needed in a system are dispatchable aka peak. If your peaker plants can handle the entire load then you don't need any other type of power. You may want to have supplies that are cheaper and/or greener, but you don't need them.

So until Nuclear plants can provide dispatchable power, they're competing against the much cheaper solar/wind power, not against natural gas or batteries where the cost disadvantage isn't as bad. Nuclear theoretically could be dispatchable, but current designs aren't.

Rolls-Royce is working on small modular reactors.

There's a large amount of companies working on small modular reactors: https://en.wikipedia.org/wiki/List_of_small_modular_reactor_...

Wait until this guy finds out France is nearly all nuclear and did it at 1/10th the cost of solar and wind.

Are people not able to do an internet search for this stuff? All this data is right out there in the open.

France is going to retire end of life nuclear reactors and replace them with renewables, while their new generation reactors have had all the same cost and schedule overruns as elsewhere. Including BTW, China, which did all the construction expertise management for nuclear that the experts recommend - e.g. using the same crews the build multiple reactors, moving them from project to project. China too stopped building new reactors in favor of more renewables.

> while their new generation reactors have had all the same cost and schedule overruns as elsewhere

Reactor, singular. They've only tried to build one new reactor this century, Flamanville #3. Civaux reactors were finished in 1997. As in the U.S., it's clear the cost issue is primarily related to public sentiment, not fundamentals. France's existing nuclear energy infrastructure was built at a time when France was very pro-nuclear, and nuclear was a matter of national pride, thus the gall to sink the Rainbow Warrior. Times have changed, for better (less investment in nuclear weapons, appetite for murdering civilians) and worse (less support for nuclear energy).

> China too stopped building new reactors in favor of more renewables.

"China will build six to eight nuclear reactors a year between 2020 and 2025 and raise total capacity to 70 gigawatts (GW), up 43.5% compared to the end of May, the official China Daily said on Thursday, citing the country’s nuclear association." https://in.reuters.com/article/china-nuclearpower/china-to-b...

Since 2011, over 30 reactors have come online, 3 of which began construction after Fukushima: https://en.wikipedia.org/wiki/List_of_nuclear_reactors#China

It's been a while, so things might have changed, but this article had indicated that while having plants still on the schedule, China had drastically cut their actual nuclear project starts, electing to finish out started plants, with the exception of the one of their internal design.

https://www.technologyreview.com/2018/12/12/138271/chinas-lo...

We will see if those announcements actually go forward.

It's fair to say that China's enthusiasm for nuclear has diminished, but it seems energy demand in China is too great to completely sideline nuclear the way Western countries have.

FWIW, here'a a May 2020 video purporting to show a construction milestone, placement of the second steel containment ring, for the first Zhangzhou reactor: https://www.facebook.com/ChinaNationalNuclear/videos/2547931...

It's a coping mechanism. They want to believe solar and wind will solve all the problems because it is comforting. Of course, it also totally ignores how renewables are enabled by peaker plants burning natural gas. And that renewables generate less electricity as a percentage of total usage now than in the 1950s. But somehow batteries are going to solve that problem. Of course, they haven't calculated how large the battery would need to be for that scenario to work out either.

Renewables + storage is already cheaper than many workloads where natural gas peaker plants applied for short runs. The prices for nat gas plants is basically mature, while renewables and storage are on a manufacturing S-curve cost decline.

Basically this means there will be expanding workloads where it makes no financial sense to use natural gas at all, let alone nuclear. It's a matter of when, and we should be accelerating financial threshold to fight the climate crisis.

I understand what you are saying, but extrapolate out further. If we have 50% renewables how many batteries will be required? It's a lot of batteries. These batteries require lots of material inputs (copper, lithium, lead, cobalt, etc., depending on the chemistry). I did the calculation a while ago and it convinced me that it is not possible to consume as much energy as we currently do and meet the need with battery storage.

Edit: You have to be able to handle the worst case. Which means you've maxed what you can get from non-renewables and you have a period where renewables are generating less than normal (e.g. low-wind and or cloudy for a few days). Right now, renewables work because you just fire up your peaker plants when you have no sun or wind, so you also don't need much storage. What happens when you need more storage? You have to build batteries (or pumped-water or whatever).

Your calculations don't agree with the majority of what gets published in peer reviewed literature. Many different groups have come up with tunable models to match current energy consumption with renewable build from 80% to 100%, with tunable parameters to cost optimize. General consensus is that nuclear at a small percentage would help greatly. However these models tend to use cost curve estimates from 5 years ago, so they assume that nuclear is far cheaper than reality, and meanwhile wind solar and storage have been ourperforming their predicted cost curves.

I would be willing to bet a dollar that for 95% of the earth, a wind/solar/storage/hydro grid will be cheaper, use fewer toxic resources, and produce more excess energy than one that includes nuclear. Because once you don't plan for that baseload, you start overbuilding cheap solar and wind to give excess capacity. A renewables-based grid is one with excess unused capacity nearly all the time, and one that is right at capacity for ~week per year.

Can you provide a link to some of that research, I'm not familiar with it, but am interested in reading it.

Edit: I think solar/wind prices are artificially low because they exist in a system with peaker plants. When you remove the peaker plants the solar/wind utilities have to internalize the cost of storage, which is currently being externalized on natural gas peaker plants.

I'm having trouble finding the time for a proper literature cull (this is now area of research so I don't keep these articles in my bibliography...), but here's a few:

Very old article about using HVDC to drop prices (still assumes coal is viable, because it uses old tech prices) https://arstechnica.com/science/2016/01/making-a-single-us-e...

The more publicized papers are those from Clack and Jacobson in PNAS arguing about whether nuclear/storage would be needed or not; both situations assumed that renewables would be far above 50%. There was a really silly lawsuit even, which is pretty awful:

https://www.greentechmedia.com/articles/read/100-percent-ren...

So even in the case of bad modeling for 100% renewables, the critics are mostly people who think that 80% renewables is quite achievable with our current tech curves, and with building more nuclear.

Apologies again for not being more specific, but hopefully the papers in these press articles will help follow the citation chain to even better modeling. Both spatial and temporal modeling are important for both weather and electricity demand, so it's not an easy thing to get right. And I suspect that the market will find minimums that are quite different than anything these programs will find, due to political and other social factors nor modeled. But the models do show that a different future is possible, at least.

Thanks, this is causing me to re-evaluate my views a bit. I am still skeptical that renewables can replace existing electricity generation without peaker plants because of the large cost of storage in those scenarios, but it may be less of a problem than I thought.

If you come to different conclusions from reading some of these, I'd love to hear about it! These were the sort of papers that gave me a looooot of hope a few years ago. Since then I have found the non-tech challenges to be the biggest concern for an interchange to a low- or zero-carbon economy.

As I understand it, the only really hard part is that one week of really bad combination of weather for wind and solar in the winter. Because of that, we need to keep around a ton of backup fuel-based generation (which includes nuclear).

UK is already at about 40% renewables right this second with only a very small amount of storage.

Sorry, should have been clearer. I meant 50% solar/wind. It is because of peaker plants and because their renewables are often biomass, which just gets burned, so effectively functions as a peaker plant.

Edit: On reflection though, you may be right. I might have gone too low. The problem I'm trying to point out occurs when you don't have access to peaker plants. As long as there is enough peaker plant capacity you do not need much storage.

Oh I know about that. I also know about their attempts to do it again in the past decades.... not looking good!

Don't forget the Canada examples too! They were able to build effectively.

I don't think corporate culture in the US is capable of it. Ironically the one thing that may have saved Westinghouse is if there had been more regulation, and the NRC has to not only evaluate safety of the designs, but also "can we even build this thing." Something that an exec with long term thinking would ensure! But current executive culture is about cost cutting in the moment, and it has pervaded even those companies that need to be thinking on decade long builds of projects that need to last half a century or more.

Reply on news.ycombinator.com