Comment by Manuel_D | Hacker News Reader

Comment by Manuel_D | Hacker News Reader

Parent Full thread

Each plant does not require custom designs for each specific site when the same design is built in series. This is big reason why nuclear plants built in the last few decades have been so expensive: they're usually the first one of their kind build in that country. That first build is when a lot of the design quirks are worked out, and construction teams gain experience. Furthermore, sub-components like steam-generators[1] and heat exchangers can't benefit from economies of scale.

By comparison, when plants of the same design are built in series they're much cheaper. Nuclear plants that started construction in the late 1960s and early 1970s were some of the most effective clean energy investments, delivered at a cost of $1-2 billion per gigawatt. Check out the costs in the historical list and there's a large stretch of plants that were built at this price point [2].

1. https://en.wikipedia.org/wiki/Steam_generator_(nuclear_power...

2. https://en.wikipedia.org/wiki/Nuclear_power_in_the_United_St...

Nukes are expensive just because they are expensive. Always have been, always will be.

They probably could be made more cheaply than they have been. Making them cheap enough to be economically competitive will be very hard and and very uncertain of success. And, trying will take time we cannot afford, because you get no power out until it is all finished, and you burn coal until then. Renewables start producing and displacing coal almost instantly.

1-2 billion per gigawatt is very economically competitive. And that's what nuclear has historically cost when it was built at scale in the US. A fair comparison against intermittent sources like wind and solar needs to include the costs of energy storage for the latter.

Why would you include integration costs for renewables but not nuclear? They face very similar challenges with very similar solutions.

The demand curve is lumpy on a daily, weekly and yearly cycle. Nuclear is flat and predictable, renewables are (like demand) predictable to a reasonable degree but variable (weather and energy demand are linked too).

Nuclear (and coals) difficulty meeting peaky demand is a big reason for pumped hydro to exist for decades. It is still the biggest source of energy storage we have. Why should renewables pay for storage built for nuclear?

No, they do not. The demand curve is mostly flat. Peaky demand is not nearly as large as problem as people make it out to be [1]. Furthermore, nuclear poet's output can be modulated by more aggressively cooling the reactor.

1. https://www.eia.gov/todayinenergy/detail.php?id=42915

I can clearly see daily, weekly and seasonal lumps in your graph. Why do you think that contradicts what I said?

And note that's averaged across 4 years, across the whole US across multiple time zones and doesn't include energy for heat currently not sourced from electricity.

Cold places use more energy in winter and in cold years, the opposite in hot places that need a lot of cooling.

But, let's put that all to one side, on that graph what amount of energy are you intending to generate with nuclear and what are you intending to do with the excess (if you aim high) or with the peaks (if you aim low). Because whatever you do, is almost certainly going to be the same things renewables would do when demand and supply don't meet exactly. Like pumped hydro or off peak pricing which already exists and is therefore included in your "flat" graph.

Weekly and seasonal changes are not a concern for nuclear power - reactors can scale up or down thermal generation on that larger time frame. Daily power fluctuations are much smaller than you seem to think, 20% to 30%.

Excesses of energy are easily solved: use it to desalinate water, create hydrogen fuel, or some other energy intensive things. And as stated previously, nuclear power plants can reduce their output if needed by more aggressively cooling the reactor. Nuclear plants want to operate at 100% load as much as they can, but they don't need to.

Renewables have the opposite problem: they don't produce energy when we need it. No amount of solar panels will produce energy when the sun is down. Peak pricing is not going to help us here. This is why renewable advocates assume that there's going to be some magic storage solution that offers effectively free and unlimited storage.

> Like pumped hydro or off peak pricing which already exists and is therefore included in your "flat" graph.

The USA has 20 GWh of pumped hydro storage. That's 2.4 minutes of electricity supply at an average hourly consumption of 500 GW.

Okay, so your answer is that you're going to aim for the high end, and your going to waste the extra energy as heat when required on the short term and turn entire plants off during seasonal lulls.

This is called "overprovisioning" and it's exactly what renewables intend to do. In both cases it's more expensive than having magical demand matching production.

Or your going to use it to desalinate when power is abundant. This is "demand response" and again, it's exactly what renewables grids are going to do.

I see you also want to create Hydrogen, another great idea that renewables also plan to do, except nuclear generated is called Pink Hydrogen and renewable powered is called Green Hydrogen for some reason. Same basic idea though.

So, I don't understand why you're so angry about renewables when you already know all the solutions a fossil free grid needs.

> No amount of solar panels will produce energy when the sun is down.

Did you happen to notice where the bottom of the lumps in your graph were? Nighttime. So we'll need some wind power and/or storage and/or HVDC lines.

But if you have to pretend you've never heard of wind power in order to justify nuclear power, maybe that's a sign that it's not as good an idea as you think.

And finally, the reason we don't have much storage on the grid today, is because we burn fossil fuels for a great deal of power and turning them off actually saves money (and lives) so that's what we do rather than store energy.

I'm fairly chill about storage because cheap solar that lets you turn off gas during the day and cheap wind that lets you turn gas off another half of the time is the quickest and cheapest way to stop burning fossil fuels. We only need to come up with crazy solutions once we try to get rid of the last few bits of fossil fuels. And you already listed them for a fully nuclear grid so we don't need to go over that again.

Some people try to act all appalled about that, as if burning 80% less fossil fuels easily at an overall cost saving in a short period of time is some big conspiracy by fossil fuel companies to undermine nuclear. It doesn't really pencil out though. Not even if you pretend wind power doesn't exist.

Wind still needs to be over-provisioned by a factor of 3 or 4 [1] due to seasonal variation. By comparison, nuclear only needs to be over-provisioned by a factor 1.2 or 1.3.

> And finally, the reason we don't have much storage on the grid today, is because we burn fossil fuels for a great deal of power and turning them off actually saves money (and lives) so that's what we do rather than store energy.

No, the reason why we don't have storage on the grid is because we have no feasible way of storing energy at grid scale. Hawaii, California, and parts of Germany all produce excesses of electricity during peak solar and wind generation. We were promised that people would take this energy off the grid, store it, and release it later when energy is in greater demand. But these promises have not panned out, because storage remains infeasible. So instead, we keep burning fossil fuels. And we'll continue to keep burning fossil fuels until we either adopt nuclear power or some massive breakthrough in storage is found.

It'll still take significant amounts of storage even to drive fossil fuel use down to 20%. Estimates say 12 hours of storage [2]. That's an incredible amount of storage, hundreds of times greater than what we have now.

1. https://www.eia.gov/todayinenergy/detail.php?id=20112

2. https://pv-magazine-usa.com/2018/03/01/12-hours-energy-stora...

Your first link doesn't say what you think it says. Did you link to the wrong thing? Or just make a wild assumption based on capacity factor, which is what that page talks about?

The second link also doesnt say what you claim. It says 12 hours of storage or HVDC connections would be required. 12 hours storage being half an EV per American household, which isnt particulalry massive and is explicitly an upper bound, ignoring hydro and nuclear and biomass and demand response etc.

Their solar only model suggests 1.5x overprovision. How much cheaper does solar need to be for a purely solar grid to beat nuclear using your number? 1.2/1.5 suggest 20% cheaper is all they need in this extreme solar only model. But in reality solar (and wind) are 20% of the cost of new nuclear (80% cheaper) and predicted to fall further.

So even your entirely unsubstantiated claim about needing 3-4x wind overprovision means it would be cheaper than nuclear.

Look at the chart, the seasonal variation in wind generation can be as large as 4x: https://www.eia.gov/todayinenergy/images/2015.02.25/main.png

Overproduction without storage results in diminishing returns, as most of the excess energy is wasted.

"New nuclear" is 4-8x more expensive than nuclear power built at scale. When multiple copies of the same plant design built in series, costs in the range of $1-2 billion per GW is typical. And remember, that's a GW of generation with >90% capacitor factor, as compared to 25-35% for most renewables.

"Wasted" solar and wind power was free. Overprovisioned renewable capacity is cheap. Charging storage, later, from overprovisioned renewable capacity, or connecting to the power grid and selling on the open market, converts overproduction to income.

Wasted nuke overcapacity is very, very expensive.

To see what later copies of existing nukes cost, we need look no further than Vogtle.

> Charging storage, later, from overprovisioned renewable capacity, or connecting to the power grid and selling on the open market, converts overproduction to income.

Except we have no way of building grid scale storage. Again, we have a few minutes of storage. Until then, wind and solar is really wind, solar, and fossil fuels. We have no plan to get rid of the fossil fuels until a storage breakthrough occurs.

> To see what later copies of existing nukes cost, we need look no further than Vogtle.

Like I said: building one power plant of a given design is way more expensive than building repeated copies of the same design. There were dozens of nuclear projects that had a cost to energy production ratio of less than a fifth of Vogtle.

We have many ways of building grid scale storage. All we lack is certainty about which kind will end up cheapest.

It just isn't built out yet. Building out storage will be premature until after there is enough extra renewable generating capacity to charge it from.

If lots of projects would have been cheaper than Vogtle, we may reasonably conclude that they didn't want one cheaper than Vogtle: it would have left too little scope for corruption and graft. Corruption is the ever-present companion of all nuke projects. Any nuke project without a ready conduit for graft will not be funded.

Real research on this looks at hour by hour demand and weather records in specific locations.

If you use more energy in the winter, and wind power works better in the winter, you've got a positive correlation.

You don't just assume you need an average output equal to the max demand.

And the price for solar is based on generation, not capacity.

> The LCOE "represents the average revenue per unit of electricity generated that would be required to recover the costs of building and operating a generating plant during an assumed financial life and duty cycle", and is calculated as the ratio between all the discounted costs over the lifetime of an electricity generating plant divided by a discounted sum of the actual energy amounts delivered.

Real research examine both hour by hour and seasonal variations. Both must be offset either by storage and/or overproduction.

Unfortunately for the US, energy use is highest in the summer when wind is producing the least amount of energy. And then there's the hour-by hour variations on top of that.

Wind production: https://www.eia.gov/todayinenergy/detail.php?id=20112

Energy use: https://www.eia.gov/todayinenergy/detail.php?id=42915

> Nukes are expensive just because they are expensive. Always have been, always will be.

You are replying to a post that provides historical evidence and geuine argument, and your only counter argument is a Tautology and outright dismissal?

There is a great deal of history in nuke implementation. We don't need to guess how things play out, because we have seen how they play out.

Back in the '50s, the promise was that power from nukes would be "too cheap to meter". They knew they were lying when they promised that, as we now know from internal memos. Cost estimates published have never matched actual costs. Delivery schedules published have never matched actual delivery, when anything is delivered; and when not, none of the money is ever returned.

So, the first and most important lesson about nukes is that what promoters estimate bears no resemblance to what might be delivered, and relying on such estimates puts you in a bottomless moneypit.

How can it be called "clean" when it produces toxic waste that requires its own infrastructure and expertise to be handled.

I think you meant "co2-emission-free-while-operating"

Nothing humans do is "clean", so you always have to read the word in context. In the sixties the alternative would've been coal which in comparison is less clean.

It’s still euphemistic and you know it.

I don't think so. What would be "clean" energy for you? Renewables also take considerable resources to build. I assume you weigh nuclear waste very heavily, but I don't believe it to be such a serious problem. There is not that much highly radioactive waste being produced. I prefer renewables over nuclear because I think that in the end they'll be cheaper while being okay for the environment. But I would be very easily convinced that nuclear energy has a smaller environmental footprint.

The sum total of the waste produced by nuclear power generation in the US occupies a volume the footprint of a football field and 10 yards high [1]. The "infrastructure" it requires is concrete cylinders that are perfectly safe to stand next to [2]. The amount of waste it produces is miniscule relative to the amount of power it produces. And most importantly, this waste is contained rather than released out into the atmosphere like greenhouse gases.

1. https://www.energy.gov/ne/articles/5-fast-facts-about-spent-...

2. https://en.wikipedia.org/wiki/Dry_cask_storage#/media/File:N...

Even manure from a pig farm requires infrastructure and expertise to be handled, thats how you get ocean dead zones where everything except jellyfish suffocates

> How can it be called "clean" when it produces toxic waste that requires its own infrastructure and expertise to be handled.

That's always been a paradox with nuclear power. On one hand, of all electric power generation technologies, nuclear power has the most toxic waste; on the other hand, that toxic waste is completely contained under normal operation, so that looking from the outside, it's one of the cleanest. That's why in these discussions you always have people talking past each other: depending on how you look at it, nuclear power can be very clean or very dirty.

Reply on news.ycombinator.com