This is an article about the argument over what the next generation of energy generation should look like.
All comments in BLACK are agreed by both sides to be true statements of fact (or uncontroversial opinions held by both parties)
All comments in ORANGE are statements made by the pro-nuclear side that are not agreed to be true by the anti-nuclear side
All comments in PURPLE are statements made by the pro-renewable side that are not agreed to be true by the anti-renewable side
All comments in BLUE are my own comments and analysis
The case for change
Both those who favour nuclear power and those who favour renewables agree that the status quo with respect to power generation is unsustainable. This is because current generation techniques use a lot of coal and gas, which are extremely polluting technologies. The current system therefor kills a lot people through the well-understood mechanism of air pollution, and has the potential to kill a catastrophic number of people through the less well-understood mechanism of global warming / climate change.
The two most common suggested alternatives to the status quo are nuclear and renewable. There are other technologies which could be advocated for – especially fusion and social technology that radically reduces our power dependence – but both sides agree that none of these solutions will plausibly be in place at the scale they are required to be by the time we must switch to a different type of grid, so they are not considered further here.
Nuclear generation is defined by the pro-nuclear side as:
- The ideal generation technology for a Western economy like the UK / US is mostly nuclear, with a mix of wind, geothermal, tide, etc. depending on what is most cost-effective (which depends on location, etc.)
Renewable generation is a bit harder to define as it consists of lots of discrete technologies which have a particular niche and will be more or less suitable in different countries (ie wind power is more suitable for windy countries, solar power more suitable for sunny countries etc). It is defined by the UK-based pro-renewable side as meaning, in the UK:
- The ideal generation technology for a Western economy like the UK is mostly wind, with solar on roofs for domestic and business use and a mix of solar, tidal, geothermal and wave etc depending on viability. Any existing nuclear capacity stays put.
(For the succinctness, I’ll call this ‘mixed wind’ when the distinction between the different possible renewable schema matters, and ‘renewable’ when it doesn’t.)
Although this is set up as being ‘renewable vs nuclear’, nobody really cares how we generate our power as long as it is clean and sustainable. So if it turned out the most cost-effective solution was 20% of one and 80% of another neither disputant would care that much, or care which technology was the 80%.
Comparing each technology
The most important factors in deciding generation sources were agreed to be externalities (especially the externalities of pollution and safety), followed by cost, followed by flexibility.
Externalities
Both nuclear and mixed-wind have some substantial advantages over coal / gas in that they don’t produce any CO2, their waste products are much better controlled and they are much safer.
Compared against each other, there is not much to choose between nuclear and mixed-wind in terms of externalities. It is agreed that nuclear has a track record of being safe vs renewables that we just expect to be safe by their nature, and it is also agreed that waste products are slightly easier to handle with renewables (in the sense that they don’t exist) than nuclear (in the sense that they are captured and contained after use).
Nuclear has a small chance of an accident causing a public safety issue, for example a meltdown. Both sides agree this chance is very small, and that safety concerns would be a poor reason not to pick nuclear. Insofar as irrational fear of catastrophe can be regarded as an externality, it might be something to take into account (either by weighting the decision towards renewables or educating the public on the actual risks).
To be cost-effective, mixed-wind (and renewables more generally) would require significant improvements in energy storage technology. Insofar as uncertainty about future generation capacity is an externality, this would weigh against renewables to a greater or lesser extent depending on beliefs about how close we are to achieving this technology. (This is sufficiently important that I’m placing discussion of it in its own section below.)
Cost
The most obvious way of comparing the cost of power sources is by looking at how much they cost per ‘megawatt hour’ (MWh), which is to say what the price is per 3.5 billion joules of energy produced (i.e. the price normalised to the equivalent output of producing a million watts with a particular method for an hour).
As a point of comparison, Wikipedia says the cost of gas-fired power is somewhere between $50 (£40) and $150 (£115) per MWh. This does not include the health costs associated with pollution, or various other costs associated with global warming.
Nuclear currently costs somewhere between $90 (£70) and $130 (£100) per MWh, with the different price estimates coming mostly from different ways of handling inflation / opportunity cost of the value of the large upfront cost of a new plant, as this is the largest component of the cost of generating nuclear.
- The pro-renewable side argues that it is misleading to straightforwardly compare this figure to the cost of wind. They point out that since nobody is talking about ripping down existing capacity it would be wrong to use the cost of generation taking place today as the benchmark for new plants in the future – legal challenges etc contribute to greatly rising costs
- The pro-nuclear side argues that the cost of nuclear as a technology could be brought as low as $40 (£30) per MWh based on the cost of new reactors in South Korea, which would essentially obviate the pro-renewable’s point above, since future capacity would be much cheaper than existing capacity.
- However the pro-renewable points out that the political context in South Korea is very different to the US / UK, so this price would probably not be attainable in practice – new capacity in the US looks likely to be closer to $140 / MWh, indicating that whatever the tradeoffs required to get down to $40 / MWh they are not straightforward or easy.
Wind currently costs somewhere between $80 (£60) to $130 (£100) per MWh, with the different price coming from the location of the wind source – onshore in an ideal location is the cheaper end, offshore is the dearer end.
- The pro-nuclear side argues that this figure is misleading. Although it is true that this is what a MWh generated by wind currently costs, it is false to imply that this is what every subsequent MWh would cost – given the lack of cost-effective storage technology – the price of generating a marginal MWh goes up exponentially the more of that power is being generated from capacity, to the extent that well before you get anywhere near 100% renewable you are paying an effectively infinite amount per MWh. This could change with new technology, but needs to be flagged in a discussion about its feasibility today.
Even given this disagreement about the precise price, both sides agree – even without the externalities – nuclear and mixed-wind are competitive with coal and gas, which are both mature technologies.
Flexibility
Flexibility is interesting because the preferred solution depends on the exact meaning of ‘flexibility’ – whether it is meant the flexibility to respond to changes in demand (which favours mixed-wind and renewables more generally) or the flexibility to deploy as a solution in all contexts (which favours nuclear)
Renewable energy can be considered ‘flexible’ in the sense that it is incredibly fast and cheap to add new capacity. This is true both in the sense that if we want 1 MWh more energy per year we can build 1MWh-worth of wind turbines, but also in the sense that if a shift in technology or climate makes solar (for example) relatively more cost-effective compared to wind we can begin to shift capacity over to solar very quickly. By contrast nuclear generates close to full capacity all the time, so if we think we will need more power we need to predict this at least a couple of decades in advance; it takes a long (and uncertain) amount of time to add new nuclear plants, since the track record of countries like the UK / US building nuclear power plants on time and to budget is not impressive.
Nuclear can be considered ‘flexible’ in the sense that it can be built anywhere, and still provide reliable and stable power. About the only constraint on nuclear in this regard is that it is cheaper to build them on the coast, so that they can use the water as coolant (otherwise you can use a newer design that recirculates water).
Nuclear power can also be reasonably insensitive to local conditions. Raw material costs are a very small fraction of the cost of generating nuclear power, so the cost is reliable even if there are market movements in the price of materials. Additionally, nuclear power costs don’t scale significantly with the number of nuclear power stations already on the grid – the next nuclear power station costs about as much to build and integrate into the network as the previous one.
I get the sense that coal / gas have the advantage here in either case – it is very easy to manage demand (burn less or more coal / gas) and they can be sited very flexibly, although there are attendant issues with the flexibility of coal / gas such as long term purchasing contracts with foreign governments which might fairly be said to affect their flexibility in both the ‘deploy in all conditions’ and the ‘respond to changes in demand’ sense. I didn’t ask the participants to talk about this, so I could be way off base.
Other
Whereas renewable power is (from the human race’s point of view) effectively a solution to our power needs forever, nuclear power still requires digging up a scarce resource from the ground and using that resource up. This resource will last maybe 100 – 200 years before we need to switch to a new form of energy.
Renewable power can be owned through a lot of different mechanisms. Most interestingly, individuals and communities could own their own power generation equipment. There are a lot of potentially interesting economic and social effects of this, which are a little outside the scope of this essay. Such a change in the ownership of power generation would not be possible with nuclear.
The storage problem
The biggest weakness of a renewable grid is that the power sources are intermittent, which means that we would have to store power generated during periods of peak generation (eg during the day with solar) and feed it back into the grid during periods of lesser generation (eg during the night with solar). This is a significant problem for renewables because technology to do this in a cost-effective way does not exist yet. This means that (given current technology) costs increase massively once you try generating power from storage. To be clear, the turbines themselves don’t appear to be a problem – we could build them today and use them happily for the next few decades. The problem is that we don’t know when the technology for capture will fall in price / improve in efficiency enough to make the generation & storage combination economically viable.
The speed at which we could switch to a mixed-wind system depends almost entirely on when this technology is invented and how it works – it could take a long time to deploy or it could be something that can be mass-manufactured and plugged in to the existing grid.
Consequently an important factor in the choice of renewables and nuclear is the speed at which our current capacity could be replaced by nuclear plants. Obviously the faster the speed of replacement, the more certain we would have to be that the storage technology was going to be invented soon in order to justify allowing the extra years of death by pollution and risk to the climate from continuing with coal / gas.
There is disagreement on this figure, which stems almost completely from the feasibility of building a very large number of nuclear plants at once. That is to say, both sides agree that it takes approximately 10-20 years to get a single nuclear plant from planning to full output, but both sides do not agree that therefore a sufficiently large number of nuclear plants can be built in 10-20 years. ‘A large number of plants’ is obviously context-dependent; the UK might need only 10 plants online at once, whereas the US might need 100.
- Predicated on the idea that we can build a very large number of nuclear plants at once, the pro-nuclear side argues that we could be mostly nuclear by 2050 (in a country with a significant nuclear baseline like France this could be even sooner).
- Predicated on the idea that we cannot build a very large number of nuclear plants at once, the pro-renewable thinks that in a country like the UK it would be closer to 2080 before most of the energy demand could be met by nuclear.
Whether or not we can build a large number of plants at once depends on various factors, including the availability of experts and the political will to push through unpopular decisions, which are both problems that are heightened by trying to build many plants at once.
The controversy
The pro-nuclear side describes opposition to nuclear as falling into three main groups:
- Fossil fuel advocates (the pro-nuclear side says that the fossil fuel industry benefits from spreading misinformation about nuclear because nuclear, rather than renewables, are their main competition)
- People who have the wrong facts (usually about safety or the current feasibility of storage)
- Holders of a specific philosophical belief that I’m really struggling to get consensus on, but which sort of corresponds to the stereotype of a ‘hippie’. The basic gist is that this group of people would prefer that we adopted a less consumerist / growth driven society, and running out of coal / gas presents a good opportunity to get their political demands met; that is to say that the apparently amazing panacea solution of nuclear removes the ability for virtuous self-sacrifice.
The pro-renewable side describes opposition to renewables as falling mostly into two camps:
- The main sort of opposition is people who do not want to make significant changes to their lifestyle, and would therefore be happy with an inferior solution that requires no changes in individual and community organisation to a superior solution that requires a change in perspective of how we view energy use.
- A lesser strain of thought, which is a group of people who are philosophically committed to the idea that humankind can triumph over nature using technology who therefore don’t want social change on principle, not because they are change-averse in general.
There are some groups which might benefit indirectly from selecting one method of generation over the other (eg the medical industry benefits from the ability to make medical isotopes in nuclear power plants on the one hand and utility companies may have a preference for more centralised power to lobby on the other) but to the extent that they matter at all these are not really major players in the debate.
Both sides seem very ready to change their mind on the issue if presented with evidence.
- The pro-nuclear side would immediately become pro mixed-wind if a change in technology made storage economically viable, or someone demonstrated a working grid based mostly on renewables that overcome the problem in a different way.
- The pro-renewable side would change their mind if a zero-subsidy nuclear plant was given planning permission in the UK. Indeed, they say that about 30-40 years ago they would have been in favour of nuclear (such as the path France went down) – they just believe that now society is close enough to viable storage technology that it would be foolish to invest in a technology like nuclear with such a long lag time.
Analysis
This is an issue on which I thought there would be absolutely no agreement on values, done as a bit of an experiment into how much agreement on facts there would be when the two sides completely hated each other. It is therefore striking to compare the consensus statements side-by-side and see that the whole issue hinges on one (maybe two) incredibly technical points:
- When will cost-effective energy storage technology be available?
If the answer to this is ‘before about 2040’ then building nuclear plants today is probably a mistake, especially if funding for these plants displaces funding for basic research into energy storage which would make an important discovery more likely. If the answer is ‘after about 2060’ then building the plants is probably not a mistake. The difficulty is that the twenty year gap in the middle is exactly the disputed territory, and it is really hard to predict when inventions will happen in the future (if you could do this you would just invent them yourself!).
For what it’s worth, the pro-renewable side thinks that there are some signs the breakthrough will happen in about five years. This potentially sets the stage for a compromise solution where the pro-nuclear side agree to support as much research into energy storage as possible until 2025 (or whatever), whereupon either the technology exists or the pro-renewable side concedes we are an unknown length of time away from the solution and agrees to support nuclear as insurance against uncertainty. However this might not be a sensible compromise from the point of view of the pro-renewable side, since there is a second outstanding question which is:
- What inefficiencies are generated trying to build multiple nuclear power plants at once?
This affects the work the word ‘about’ has to do in the first question above. If it is possible to build an unlimited number of plants at once, then if we started now we’d have all the nukes we could ever want in place by 2040 and no need to worry about our energy security. If there are significant losses induced by building multiple plants at once then it is possible that we couldn’t have a full grid up until – say – 2080 (enough time for the government to force through a lot of new nuclear engineer capacity plus each of those engineers to build a plant), in which case we are almost certainly over a barrel with respect to significant problems with our energy supply before then and effectively have no option but to go renewable and hope that storage technology is invented before the coal and gas runs out.
In some ways it is possible to view the dispute as purely boiling down to people’s preference for guaranteed energy security in a few decades with a good but not perfect system vs people’s willingness to take a risk on energy security (ie gamble on when cost-effective storage will be invented) for the chance of an effectively perfect generation system, potentially even faster than we would have got the nuclear solution in place. Without more facts I think it would be really hard to convince an advocate one way or the other, since people’s preference for risk and reward seems a really deep part of their personality and not likely to be swayed by emotional appeal.
Concluding Statements
I asked both sides if they wanted to add up to 100 words of ‘summary’ to clarify points of nuance they think I got wrong.
From the pro-nuke side:
- “There is not a single power plant in the United States that’s free of subsidies. Nuclear is actually the least subsidized relative to power produced. I’m not sure what the subsidy situation in the UK is like but I question that [requiring evidence of a subsidy-free nuclear plant] is a realistic request.”
The pro-renewable side wrote more than 100 words, so to be fair to the pro-nuke side I’ve picked out just the concluding 100 words:
- “[The article claims that the debate is between the ‘certainty’ of nuclear vs the ‘risk’ of renewables but] New Nuclear is also a gamble as the design for Hinkley C (for example) has never produced commercial energy anywhere in the world … batteries can be seen as less of a gamble since they exist now and just need economies of scale to make them cheaper and some more incremental technology advances which are being worked on by thousands of companies.”
I also asked if the participants would like to comment on the process, and received the following:
From the pro-nuke side:
- “I think the most interesting bit about this is that I didn’t end up getting the usual unfounded complaints, certainly no impression of bad faith. The process also kept things railroaded, no long digressions on specific topics”
From the pro-renewable side:
- “The process I think is good, in the sense that arguments on the internet often just descend into abuse often and this avoids that, maybe a bit about the evidence the two sides use, like the links they think are credible, often I have arguments with people and they just dismiss certain things because of a source they don’t like in the sense of a newspaper that they don’t like so everything in the newspaper is not worth reading.”
Thanks
Unlike the previous article about the Puppies I know next to nothing about nuclear power, recruited from outside a rationalist sub for the pro-renewable advocate and ‘know’ the issue to be deeply technical with bad-faith disagreement on both sides. Therefore I was expecting a real bunfight of an argument which i could use as a baseline for how bad a debate could get and still be worth publishing (possibly).
So to say I was overwhelmed with the quality of the disputants is a vast understatement – both sides offered strong arguments against their own position when asked, had very sophisticated models of the other side’s positions and counterarguments and a microscopic attention to detail (I especially enjoyed when both of them – very politely but completely independently – indicated that some of my first drafts of questions were total nonsense). I could listen to them talk all day, and you are sorely missing out by only reading my lay summary of their detailed arguments.
All participants requested to remain anonymous, but indicated they would consider making throwaway accounts when I post up to reddit. I hope they do!
