Hacker News Reader

Comment by mch82 | Hacker News Reader

TLDR: Extrapolating current trends in renewable power generation, by the time we’re able to complete construction of new nuclear plants those plants won’t be necessary.

There are six significant flaws in most recent pro-nuclear arguments I’ve read. To me, these flaws undermine the pro-nuclear case.

1. Authors ignore how long it takes to bring a new nuclear plant online (>7 years).

2. Authors focus on the percentage of energy production contributed by clean sources (solar, wind, etc.) TODAY, but ignore the growth trend (which matters since reactors take so long to bring online).

3. Authors ignore the waste storage problem, which is a debt future generations will be saddled with for hundreds or thousands of years.

4. Authors ignore the power storage problem, which causes lots of energy to be wasted.

5. Authors ignore energy efficiency trends (electronics get more efficient over time).

6. Authors ignore the safety data and that safety incidents are likely under-reported.

[1]: https://www.quora.com/How-long-does-it-take-to-build-a-nucle..., “Quora”

[6]: https://youtu.be/rUeHPCYtWYQ, “Chernobyl Podcast”

Edit: added TLDR. I’ll update with references for points 2-5 later. Slow internet connection now.

View on HN

> 1. Authors ignore how long it takes to bring a new nuclear plant online (>7 years).

Most of the wait time is due to lack of demand, e.g. it takes multiple years to build a pressure vessel but this production could be ramped up 10-100x if need be. The Quora source you cite actually says 48 months for modern reactors so I'm not sure where >7 years came from.

> 3. Authors ignore the waste storage problem, which is a debt future generations will be saddled with for hundreds or thousands of years.

There is no waste storage problem. This is a myth that should have died a decade ago. Vitrified nuclear waste is inert and harmless. Modern reactors are also able to consume most of current waste. We already tolerate far worse waste products from e.g. the coal industry where it's a non-newsworthy occurrence for toxic tailings to completely destroy river systems.

> 4. Authors ignore the power storage problem, which causes lots of energy to be wasted.

The energy (not power) storage problem is unique to unreliable peaky power sources like solar and wind. Nuclear power plants provide a consistent and diallable power output. Energy storage is also a big impediment to large scale rollout of renewables because energy storage costs are superlinear with respect to demand.

> 5. Authors ignore energy efficiency trends (electronics get more efficient over time).

Increase in efficiency does not lead to decrease in consumption. It does the opposite. Look up induced demand.

> 6. Authors ignore the safety data and that safety incidents are likely under-reported.

Nuclear power is orders of magnitude safer and cleaner than anything else that exists including renewables.

https://www.forbes.com/sites/jamesconca/2012/06/10/energys-d...

This comment is entirely misleading and the author is seriously misinformed about nuclear power. This should not be top comment on any discussion of nuclear power. This is just perpetuating FUD and myths that have all been deconstructed years ago.

Your comment is entirely misleading, based on outdated data, and whitewashes the hell out of nuclear power. The fact you're trying to claim the parent post is bad is not cool.

1. Please look at actual data for power plant construction times. [1] provides an overview. Estimates are useless - if we look at how long things actually take to build... "The mean construction time of 441 reactors in use today was 7.5 years."

3. Are you actually serious right now? Radioactive waste sitting in temporary pools that will be an environmental catastrophe for hundreds of years if they leak isn't a problem? No, just because we tolerate other ecological disasters doesn't make this one ok.

4. This is no longer true, and will definitely not be true in another 5 years. Battery costs are decreasing at a reliable rate year over year. Power plants are already being built making use of them. Here's one going up now with 300MWh of storage [2].

5. No, you look up per capita demand. Here: [3]. It’s been dropping for a decade.

6. Your link does not report solar farm deaths because as far as I can tell there haven't ever been any. Unless you count birds. Rooftop solar is "dangerous" because installers fall off roofs.

[1] http://euanmearns.com/how-long-does-it-take-to-build-a-nucle...

[2] https://pv-magazine-usa.com/2019/10/08/the-sun-shines-only-a...

[3] https://www.eia.gov/todayinenergy/detail.php?id=32212

> Radioactive waste sitting in temporary pools

Spent fuel is sitting in cooling pools because it emits enough residual heat to melt if not in water. Spent fuel will typically sit in these pools for around a decade. At which point the residual heat is low enough to process it via vitrification (turning it into chemically inert glass) and then moving it to a permanent storage site. Vitrified spent fuel is radiologically and chemically safe. The only way it can become a danger is if somehow liquefied or aerosolised. The volume needed to store a century worth of all humanity's spent nuclear fuel is less than a single large landfill site, of which we have hundreds of thousands.

> that will be an environmental catastrophe for hundreds of years if they leak

If what leaks? Do you actually know what's in those pools and why? Spent fuel pools are harmless. You can go for a swim in them and receive a lower radiation dose than background. There's nothing to leak. The only failure mode is if water is not supplied to them and eventually evaporates.

> This is no longer true, and will definitely not be true in another 5 years

Cite this. Show that at any point in the foreseeable future the cost of energy storage with regard to total capacity is not superlinear. China would love to hear from you since they're stuck unable to use renewables at large scale because of this issue.

> No, you look up per capita demand. Here: [3]. It’s been dropping for a decade.

Per capita is not a relevant number, The number that is being discussed is total demand. The number that matters, since power sources need to supply a total demand not a per capita demand, whatever that would even mean. The number that's rapidly increasing.[1]

Actually take the time to read up on induced demand, then you will understand that it results in a per-capita decrease in consumption (since higher efficiency) and a paradoxical higher total consumption, since now there are more consumers who were previously priced out.

GP post is bad and your post is bad. Sorry. If you don't understand the very eli5 basics of nuclear power like the purpose of cooling ponds (and the fact that most modern reactor designs don't need them), then be aware that you're perpetuating a bunch of myths every time you talk about it. Energy science is not a topic where all opinions are equal.

[1] https://en.wikipedia.org/wiki/World_energy_consumption#Trend...

Here we go again, downplaying risks and ignoring facts that aren't supportive of your argument.

What do I think is in the pools? Radioactive waste that will have a mini meltdown if the pools leak enough water to expose the waste to air. You know, the kind of leak that can result from a natural disaster, terrorist attack, or just plain old maintenance negligence [1]. But that's ok, we'll just pretend Fukushima never happened, I guess.

I already provided a link to a renewable power plant that's going up, backed by battery storage for non-generating hours. Perhaps you didn't like it? Here's a report on how costs of batteries are dropping fast enough to make them cost competitive with traditional sources [2]. You probably won't like that, either, since you're hung up on the whole "linear with demand" thing, which no one actually cares about. What they do care about is total cost per reliable kWh, taking into account all negative externalities. But I guess we'll just pretend nuclear accidents don't occur, especially small ones [3], and that the whole long-term storage problem doesn't exist because that whole Yucca Mountain thing is just needless worrying.

I like how you focus on total demand globally because if you look at total demand in the West, it's been flat for years. You conveniently don't point out that the places demand is increasing are in the developing world, which is the last place we should build nuclear plants. Do you want more accidents due to poor maintenance or operating procedures? Because that's how you'll get them.

And I understand induced demand perfectly - my whole point of showing per capita consumption is to point out that it's bullshit in this case. Electricity has been plentiful in the US and EU for decades, but per capita use peaked 9 years ago. Induced demand is only relevant in developing markets where there is latent demand, which again is back to the developing world where I will never agree we should build nuclear.

You like nuclear, I get it. It's a good power source. It's just not the best one anymore and it's time to recognize that. But it's clear you have already made up your mind on the issue, so I won't be engaging in any more discussion with you on this.

[1] https://thebulletin.org/2017/08/pushing-the-storage-horse-wi...

[2] https://www.greentechmedia.com/articles/read/report-levelize...

[3] https://www.theguardian.com/us-news/2017/jan/09/indian-point...

> Induced demand is only relevant in developing markets where there is latent demand

...Yeah. And since the topic of this entire post is climate change, the developing nations and induced demand are all that matters. No one gives a fuck about per capita energy consumption in the US, the question is about the total energy demand of the planet and how it's supplied.

What the developed world does from here on out is irrelevant in terms of first order effects on climate change. It's already done the bulk of the damage it was going to do. Switch all of the developed world to fairy-powered zero emission fusion overnight, or anything else you want, it won't make a difference.

The deciding factor for climate change is how China and India will power themselves, and if the answer isn't nuclear, then we're all screwed. Just like TFA states.

The world is an oil based economy. If people really want to move away from oil you need to transform the economy from being based on it to on based on electricity.

Which means you need a lot more power. Solar and wind still factor in but you still want that massive amount that a power plant provides, because all current storage technology won’t be able to keep up.

Plus if all reactors were the same design such as in France it lowers costs dramatically as well.

There are 5 Quora answers: 3.3-5 years, 9 years, 7 years, 10-12 years (with 35 year outliers), and 4.8 years. 7 years is near the average of all those answers.

> “ Most of the wait time is due to lack of demand”

None of the Quora answers cited this as a reason. And the Quora answers are the low end of the scale (40 months) only included construction time in their answer, not site finding, licensing, or startup.

There is a HUGE if, IF they get built. We can't actually build these anymore. There isn't enough funding/willingness to get these built. 16 billion pound project can't get off the ground: https://www.theguardian.com/environment/2019/jan/11/hitachi-...

I think building modern reactors just stretches our industrial capability too far. Doesn't help that in the US due to de-industrialization there isn't enough skilled managers, engineers or skilled workers either.

I would also argue that #2 was already being said 7 years ago. Further, growth is not always constant growth, let alone exponential growth. It may plateau, and then grow again further down the track after other problems (eg storage) are solved.

As far as #6, it is disingenuous to count indirect coal deaths due to respiratory issues and completely ignore the indirect nuclear deaths.

For example, the article you link does not seem to account for the 2000+ deaths that occurred due to the Fukushima evacuation, nor those that survived but later starved due to their homes and livelihood being destroyed in the Chernobyl accident.

A proper model needs to take into account the economic implications of a nuclear incident.

Unluckily, you can't just extrapolate from current trends for solar and wind, because storage becomes more and more of a problem as the percentage of power coming from uncontrollable sources becomes higher and higher. When going from 0 to 10%, you don't need storage at all. But, if you want to go from eg 50% to 100%, then you need MASSIVE storage to guarantee that you have power when you need it.

By the way, your point 4 for nuclear power doesn't hold ground. If you have lots of nuclear power plants, then you're going to not use (waste) a percentage of the power you produce at night by burning fissile material. But if you don't have any nuclear power plants, then 100% of the potential energy from fissile material that is available on Earth is going to be "wasted" (ie, not used).

> if you want to go from eg 50% to 100%, then you need MASSIVE storage to guarantee that you have power when you need it.

You really dont, you just need to burn fuel to fill in for gaps caused by weather, that become increasingly less frequent the more wind and solar is oversupplied. And oversupply is not a big problem - it can be easily curtailed. The optimal amount of renewables supply to build is over the level where they regularly exceed demand - and that is the stage where big batteries become very economical to build.

Carbon neutral fuel-wise, biofuels and electrically generated fuels are very viable future options.

Storage doesn't become more of a problem with more wind and solar - it becomes more of an opportunity to further reduce fuel consumption.

Storage is not a problem now - economic solutions exist already to build it where it is valuable. Its just not needed much yet.

> You really don't, you just need to burn fuel to fill in for gaps caused by weather

That suggests that we'd need to keep most of the existing infrastructure for power generation and also replicate its capacity with wind and solar plants.

That sounds both expensive and inefficient.

> Storage is not a problem now - economic solutions exist already to build it where it is valuable. Its just not needed much yet.

Nuclear has actually proven that it can work at the scale we need it (~20% of actual power generation from memory). Renewables and storage have not. So this is an argument of hypotheticals vs actuals. The burden of evidence on the hypothetical solution is a bit higher than an assertion that it'll work.

The only grid-level storage solutions that actually store energy (not just modulate grid supply short-term) is big hydroelectric dams which are not one size fits all solutions.

The UK is already at over 20% renewable energy over last year, with peaks of over 40% . https://www.theguardian.com/business/2019/oct/14/renewable-e...

Meanwhile the nuclear power station we started building ten years ago is contributing 0%.

Aren't the UK windfarms sans-storage though? I thought they were reliant on offshore wind being a fairly consistent source.

They're not fairly consistent, its a genuine misunderstanding of the system that they have to be. UK has a bit of nuclear, a lot of gas and biomass plant, some very useful pumped hydro and sizable interconnects to trade electricity with Europe. Everything is adjusted to match generation to demand. Generation is dynamic - Demand is dynamic. There is no panic when demand or generation changes, its handled.

There's a small amount of pre-existing pumped hydro storage in the system, but mostly it relies on the UK having already built a lot of CCGT plants which can cover the gaps.

> That suggests that we'd need to keep most of the existing infrastructure for power generation and also replicate its capacity with wind and solar plants. That sounds both expensive and inefficient.

Its efficient to use existing plant for as long as it is useful. The main expense of running most of it is fuel - and the cost of that in legacy terms and in CO2 is slashed by 'replicating' its capacity with wind/solar.

Currently when we add wind and solar capacity almost their entire output replaces fuel burn. When in a handful of years wind and solar have replaced 30-40% of demand supply, about 10-20% of their output may possibly be getting rejected. That makes wind/solar maximally 10-20% more expensive than what they would be with full storage and use. Hydro, batteries and long distance transmission is already economically viable in certain situations. Hydrogen synthesis will become increasingly economically viable, with carbon pricing and development. There are technologies existing which are ready to reduce waste supply from wind/solar as it arises. But even without them a potential overhead of 10-20% cost wont stop more wind/solar getting built, because its already currently cheaper than fuel burning -without carbon pricing and has been getting more efficient year on year, with no end in sight.

This actual economic performance contrasts with new nuclears promises which are in truth in their final throws, completely outperformed already by the alternatives which have been developed. Investment in Nuclear is no real threat to wind and solar anymore, the arguments seem to persist only to delay overall action on policy transition. New nuclear plants may or may not get built, wind and solar IS getting built and its variable nature will make nuclear supply uneconomic - there's no sense in financing plants on the basis of supplying constant power, when a huge competitor supplies much much cheaper power 75% of the time, and encourages hydro, biofuels and maybe cheap storage to fill in its gaps. Nuclear is not going to be economical for that, and that is the opportunity which is being created by wind and solar manufacturers and installers, as we hash over this outdated debate.

Burning fuel to fill the gaps is technically possible; I'd like to see an analysis of costs though. Building and maintaining those plans is expensive even when you don't use them. Who is going to pay for those expenses?

>Authors ignore the power storage problem, which causes lots of energy to be wasted.

This "problem" is why we need nuclear. You can't store power generated by solar and wind, so when the sun goes down or the wind dies down, the power goes out unless you rely on some other source. Nuclear can work every hour of the day. It's the only power generation method which can supply power at every hour of the day and can be deployed almost everywhere (hydroelectric is neat but you're screwed if you don't have a good river). All these other "problems" have been written about extensively by specialists, just not by journalists.

As for the rest of your problems with nuclear specifically, I don't think you're comparing the negatives of nuclear with the negatives of literally any other power source.

The real reason we won't go nuclear is that Democrats and progressives don't like it because it's scary and makes them quake at night. We could use nuclear as one tool in fighting climate change, but we aren't willing to do that. We're dying by our own hands.

> so when the sun goes down or the wind dies down, the power goes out unless you rely on some other source.

Power can be transmitted reasonably efficiently over very long distances. Long enough that this need not be a problem.

Solar is more predictable than wind in this respect but the combination is more reliable than either alone.

Transmission is actually one of the biggest problems identified by US ISOs. Not easy to get the power where it is needed bc big transmission lines are costly.

There is a whole revolution happening in grid transmission right now. High voltage DC lines that were considered impossible not all that long ago now span 1000's of KM between unsynchronized grids and with reduced transmission losses to boot.

China has spent the last 10 years building out a $90B UHV transmission system. They're nearly done. We're still thinking.

https://en.wikipedia.org/wiki/Ultra-high-voltage_electricity...

If the transmission and storage isn't a problem, why hasn't Germany's carbon emissions from energy stayed roughly the same? Since 2000 they've invested a lot into solar and wind, and yet they still produce roughly the same amount of energy from coal and gas.

Increased consumption, for the most part.

Are you sure about that? I've been keeping an eye on the German numbers because it is the fairest experiment and it looks to me like they've failed to have any sort of environmentally positive outcome. Their energy use is flat, and they've swapped nuclear for renewables, with a little reduction in hard-coal [0].

If the goal of Germany's Energiewende was cleaner energy, France is thrashing them without even trying. That is really a stunning endorsement of just how technically excellent nuclear power is.

[0] https://www.cleanenergywire.org/factsheets/germanys-energy-c...

It's cheating to require a perfect solution before admitting a better one.

So, concretely, which source of universally available baseload power meets your criteria? Which would YOU advocate, given the urgent need to reduce emissions?

I think it makes sense to keep operating reactors that are currently active. We’re already on the hook for storing that waste and managing those decommissioned sites.

Similarly, I’m okay with a phased decommissioning of other dirty fuels. I just think it’s insane to set a long-term goal of growing dirty power jobs/industry.

I’d advocate for an immediate stop to any anti-renewable efforts. We should be doing everything we can to help achieve renewables.

Last, if we really think it’s a problem, we need to reduce energy consumption.

I feel like you might be missing some key knowledge here, including nuclear in "dirty power".

Unless you live near to a geothermal or hydro source, there is no renewable option available today that can provide baseload power, the "day in, day out" load that must always be available on the grid. The options are fossils and nuclear.

Coal waste is massive: 68 tons of solid and 77 tons of CO2 per person-lifetime of electricity. The majority of that waste goes into the air, where we can't control it and don't know what it's up to (apart from fucking our climate of course). That's not counting fly ash and flue gases, the world's largest source of released radioactivity. That stuff is packed with heavy metals like lead, arsenic, and mercury. The air pollution from coal causes 35,000 deaths from lung disease per year in the US alone - when it's working perfectly. The same figure for China is estimated to be 350,000 deaths per year.

In contrast, one person-lifetime of nuclear waste is about 12oz of solid, dry matter, which we can watch (and apparently re-use in modern reactor models). It produces zero CO2, and zero gaseous emissions. The total lifetime death toll from nuclear power is - at it's most generous - under 1,000 people.

Renewables are a part of the solution, and long term, they will be the complete solution. But even if we could solve all the power generation, storage, and transmission problems within 50 years (which would be amazing)... we need a way to provide base load in the meantime. And we have an option that is safe, zero-carbon, and produces a fraction of the waste of the alternatives.

Additionally, "we need to reduce energy consumption" is only an option if you're cool with 2/3 of the world continuing to live in abject poverty. By the end of this century the global population is projected to level off at 11 billion or so, with almost all of that growth coming outside the wealthy West. The 90% of the world which is currently using a fraction of global power generation wants washing machines, electric lighting, and computers, too. If the US cut average power consumption in half to match Europe, and then we all cut power consumption by a further 50%... we would still need to build multiples of our current baseload power generation capability.

TL;DR: We need way more baseload power than we've got. It's not helpful to lump all our options into one category and rule them out completely... especially when one option comes with zero carbon.

> 3. Authors ignore the waste storage problem, which is a debt future generations will be saddled with for hundreds or thousands of years.

Taking on debt is profitable if the interest rate is less than the return on investment. I'd rather take on the nuclear waste storage 'debt' (does no real damage if we do it properly, can be paid back later with much more powerful technological resources) than the climate change 'debt' (does real damage now, resulting in permanent loss of some amount of economic growth).

I don't understand why point 5 about efficiency trends is relevant. Power consumption trends show us power consumption is continuing to climb irrespective of efficiency. It's clear that civilisation's need for power continues to expand even if some of our devices get more efficient.

except that US per capita electricity consumption has been falling for decades as efficiency has increased.

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

> 1. Authors ignore how long it takes to bring a new nuclear plant online

As per the article:

> An argument I haven’t seen before is that no other clean energy source can scale up as quickly as nuclear—IF the reactors are standardized.

True, development costs are big, but these can be spread over multiple reactors. And once you have trained the workforce constructing one, the others can come much quicker. See historical values for France [1]. If we were to pipeline new reactors buildup everywhere, we'd be quite close to netting 0% emission in the energy sector. Granted, calculations and finding the optimal isn't that simple.

> 3. Authors ignore the waste storage problem, which is a debt future generations will be saddled with for hundreds or thousands of years.

The consequences of our fossil fuel reliance are staging _right now_ what the ecosystem will look like millions of years in the future. Even if we were to produce 1000 times the amount of nuclear waste we currently have, we wouldn't come close to that. At least, it is easier to keep the stuff in one place. This is an emergency. Yes, nuclear is far from optimal, but is a whole lot better than the status quo.

The other arguments aren't really valid in my opinion.

2. Builds on 1. The point is that nuclear can ramp up pretty quickly, as has been shown in the past.

> 4. Authors ignore the power storage problem, which causes lots of energy to be wasted.

This isn't really a problem with nuclear. Provide the baseload. You can tune up/down the amount of power, depending on your renewable output. Exchange energy with the neighbors depending on the needs. Pumped hydro can store quite a bit of energy as well.

> 5.

I fail to understand your point. Are we using less energy? I don't think so. [2] And a lot more stuff is going to need electricity in a clean energy transition (heaters, cars, industrial ovens...). If anything, cleaner/cheaper energy is likely going to increase consumption.

> 6. Authors ignore the safety data and that safety incidents are likely under-reported.

What safety data? The points they talk about in the article and use to say that reactors seem safe? Other energy generators aren't that safe, and also likely under-report.

[1]: https://en.wikipedia.org/wiki/File:Electricity_production_by...

[2]: https://en.wikipedia.org/wiki/World_energy_consumption

1. Authors ignore how long it takes to bring a new nuclear plant online (>7 years).

Nuclear has been shown to be able to build fast and with massive scale. Faster then renewable by far.

The waste problem is really no an issue once you understand it.

Modern nuclear power plants and future nuclear plants can actually load follow quite well.

Efficiency does not lead to less consumtion, but rather more consumtion. That is a basic and well known fact.

Civilian Nuclear has a steller safty record and nuclear has the strictest observation of any industry.

I don't think your TLDR accurately covers the entire article, but fine. As for the rest, the big thing missing is that the per-capita energy consumption has been on the rise for as long as we have electricity. It's at 2.5 MWh/y or thereabouts. As developing countries move up on the ladder their consumption will increase, drastically so.

This is going to cause substantial challenges.

Personally, I'd love to see a Manhattan style project to get us to fusion. We know that it can be done, we have a shining example for proof of that every day.

And in the meantime throw every $, Euro or Yen aimed at more nuclear power based on fission either at the fusion project or at renewables. You can build an awful lot of windmills for the price of one nuclear plant, and you can do it a lot quicker too.

And some of the increase is in our interest to happen. Moving cars from petrol to electric makes sense since we have ways to generate electricity is reasonable ways.

#5 may be unlikely to benefit us, given Jevons’ Paradox: https://en.wikipedia.org/wiki/Jevons_paradox

Reply on news.ycombinator.com