Hacker News Reader

Hacker News

Comment by hntrader | Hacker News Reader

Parent Full thread

hntrader

"four times as expensive as wind"

This isn't a valid comparison, since wind isn't capable of base load supply.

You would need to factor in storage costs to make the comparison valid.

View on HN

Here's a project that's supplying energy at $40/MWh *including storage costs*

https://www.energy-storage.news/news/developer-8minute-says-...

That project is a great step forward to cutting the dependency on using fossil fuels as "reserve energy" for when demand exceeds what the grid of renewables can create. There are however some details to that design. The linked project storage is 75% capacity for 4 hours, and it usage is for solar and storage which has a daily charge and discharge cycle (customers will be daily buying the stored energy and thus repaying the storage investment on a predictable and daily basis).

In order for Germany to abandon the dependency of using fossil fuels for reserve energy the storage would need to have the capacity of the full grid, it would need to be able to last weeks, and the capacity of wind would need to be able to fill both the storage and grid demands at the same time.

If and people are willing to invest into wind power plants which overcapacity won't have a buyer but will simply be there in case the batteries need to be recharged, and people also investments into batteries which won't be discharged and generating revenue unless the weather makes wind production go below demand, then there exist a chance we can replace the fossil fueled based reserve energy with wind + storage.

Currently there does not seem to exist a single wind + storage project out there. I suspect the above economics is the reason for that.

Batteries are also getting cheaper and improving in capacity at immense rates too no? It seems like a double whammy against nuclear, unless there’s something I’m missing?

Triple whammy if you include the larger and larger capacity factor of ever growing wind turbines (and offshore). Even if you're overprovisioning 2x and fill in with battery capacity for some of the shrinking holes inbetween, you'll end up cheaper than nuclear.

Which, on the other hand also still does not factor in the externalities of the still-not-solved issues of final deposition of nuclear waste (some quote 38 billion still outstanding for just getting rid of the current ones that haven't been factored into historic electricity prices), defending against proliferation and antitrust issues due to the massive capital investment and minimum size of nuclear power plants.

I used to be pro-nuclear at some point, but just looking at the current wholesale prices and trends, it's basically game over for nuclear.

Average lifetime of those batteries are about 7 years before a full replacement.

Everyone also seems to clearly ignore how detrimental they are to the environment. There's huge swathes of toxic wastelands created from mining the materials needed.

So cost, lifecycle and externalised environmental damage is what I would say is missing here.

https://www.technologyreview.com/2018/07/27/141282/the-25-tr...

Late edit: Batteries certainly have their place in smoothing out the grid inputs, I don't want to come across as saying they are not useful when they certainly are. In South Australia the newly installed grid scale batteries have paid themselves off within a few years.

The article linked makes clear the engineering tradeoffs involved, and as a society we should always base these decisions on the numbers over emotion. Think there is a bright future for small modular reactors in remote and rural areas if communities are willing to accept them.

> Average lifetime of those batteries are about 7 years before a full replacement.

Based on what?

> Everyone also seems to clearly ignore how detrimental they are to the environment. There's huge swathes of toxic wastelands created from mining the materials needed.

Ah, you're one of those people.

Never mind, carry on.

I happened to read an article a few weeks back regarding Aluminum-Air batteries, and a potential use in EVs, yet while they might not be ideal for EVs, a potential ability for use in the grid might be convenient.

> Average lifetime of those batteries are about 7 years before a full replacement.

The significance of this is a function of our ability to recycle them.

We just spent 2 decades saying this.

I suppose we’re going to say it for another.

that’s 30 years of emissions that could have been avoided.

We've also spent 30+ years trying to build nuclear plants on time and on budget and we've been unable to do that.

Even with that, 20% of US electricity is still from nuclear.

We managed to build these before the industry became expensive.

Now Germany and the United States are rushing to close the operating plants.

I’m not sure why we can’t figure out how to build a standard design and replicate it

Because we can't build anything larger than an office building on time and on budget.

Perhaps because that standard design didn't work out?

Even Switzerland is shutting down nuclear plants and they managed to build the longest railroad tunnel of the world without budget overruns and half a year before due date.

In Switzerland it's a political decision same as Germany it has nothing todo with cost or the ability to run the plants.

That's not so clear cut. Mühleberg near Berne has said that 2019 they didn't make profit, however the operator company BKW never published gains with that plant.

https://www.bernerzeitung.ch/region/kanton-bern/das-akw-mueh... (paywalled)

And we've failed doing that with software for longer, yet we still keep building software.

That's fine, but show your work. Nuclear gives us 24/7 reliable power. What is the cost comparison for renewables for the same product?

It's obvious that energy sources based on local meteo conditions don't have 24/7 reliability. The question is: how much of that reliability do we need? What would be possible if we did forgo some? Most of our industrial processes are optimized for pure efficiency (optimizing asymptotic behavior with no regard to constants), which is biased towards high-energy, constantly running chains and immense infrastructure. These need reliability. If we optimize for more smaller workloads, which can start/stop depending on local conditions (and power source), then we loose some efficiency in the process but gain a lot in the simplicity of the energy supply chain (which may in the end translate in lower/comparable overall energy consumption).

> how much of that reliability do we need?

There are two guiding clues for that. One is to look at https://www.electricitymap.org/ranking and see how Germany is doing. One can derive the weather based on how brown the color is.

The second clue is energy prices swings more heavier as a country replaces nuclear with renewable. Heavy energy using industries can't double production when prices shrink by 50%, but they will stop production when costs exceed profits. This is already happening in Sweden when prices jumped by +70% this winter. It will be interesting to see what happens when the steel industry is becoming even more dependent on energy prices and will have to stop for weeks during the winter, and the effect that will have on employment contracts.

Electricity production in the U.S. is only about 2% of GDP, but basically every economic activity relies on electricity to function. Having very much of the electric supply be truly intermittent would create enormous costs elsewhere in the economy. If you're going to power the U.S. on renewables, you're going to need a lot of storage to get the needed reliability, so if you're comparing costs, you need to compare an equivalent end product.

There’s no prospect of batteries acting as base load supply anytime in the foreseeable future. Batteries aren’t even great for peak supply yet. Those big grid-scale battery farms installed by Tesla (et al) are used primarily for grid stabilisation and a limited amount of useful arbitrage.

That is if, not only can we keep making them, we produce them at massive scales without inciting coups around the world for Lithium [1]. Because let us be honest with ourselves, the CIA has staged coups around the world many-a-times [2], and iirc but can't locate that some of said instances were done to support private entities.

[1] https://www.indepthnews.net/index.php/opinion/3735-we-will-c...

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

Lithium for energy storage is not the only option. Pumped water already has been a thing for decades. Additionally other avenues are being explored like [1] pumped air and [2] lifting concrete weights.

[1] https://www.planete-energies.com/en/medias/close/what-compre...

[2] https://profittrends.com/energy-investing/renewable-energy/s...

Interesting solutions, especially the latter, absolutely brilliant, despite the reduction in efficiency, it probably trumps batteries wrt to first materials used.

> improving in capacity at immense rates There is almost no increase in capacity at least not without making the batteries bigger.

Just curious where we are we at in terms of the environmental aspects of mass scale heavy duty battery production?

> wind isn't capable of base load supply.

I'm sorry, but really: Citation required on that claim.

There's plenty of counter positions[1][2][3] which say that if you assume 'base load' is a genuine thing, that renewables, including wind, are capable of meeting all our power generation needs - through a mix of energy sources, and also energy storage.

[1] https://www.abc.net.au/news/science/2017-10-12/renewable-ene...

[2] https://theconversation.com/renewable-energy-can-provide-bas...

[3] https://www.energymatters.com.au/renewable-news/baseload-ene...

Base load generation is unnecessary, the only type of power that's needed is dispatchable or peak power. If you can meet demand at peak, you can meet demand in the trough too. All other power types are added to the grid to increase affordability, reliability or green credentials.

Neither wind nor nuclear are suitable to use for peak power.

This assumes that you're power generation always matches demand through time. If you had a sudden spike at night that you don't usually have (like in Texas where people kicked on their heat unexpectedly) and you were using solar, the mismatch would exist and you wouldn't be able to meet the load.

Exactly. You need natgas peakers (that aren't frozen) or charged batteries to deal with this. Nuclear, wind and solar are all unsuitable for this situation.

What do you think of the claims of user "corty" elsewhere in this thread, who seems to say that nuclear isn't unsuitable for this situation?

The primary problem with nuclear is the fact that it's insanely expensive. If you're not running it flat out 24/7, then the cost per MWh goes up dramatically.

Of all the other energy sources, it's the only one where we have to put serious consideration into warning people not to dig up the waste, and how to keep them from doing it thousands of years from now.

Or turbines burning hydrogen.

> and also energy storage.

Precisely, that's my point. The costings of the OP ignore that part.

It also needs to factor the cost of producing batteries and manufacturing the storage.

"Base load" doesn't really exist any more. Yes, there's a minimum below which demand almost never dips, but the implication of base load is that this should be produced by the type of power plants that provide a continuous supply, like nuclear or coal. But these are extremely inflexible and can't react quickly to changes in demand or supply by other means. As the share of renewables grows, the position nuclear is in becomes worse, not better. More flexible solutions are needed that don't need demand scheduling days or weeks ahead.

Nuclear isn't as inflexible as you think. A usual change rate in power output is around 10%/min, so you can go from half to full in 5 minutes. Far away from days or weeks. Here is a nice comparison: https://de.m.wikipedia.org/wiki/Lastfolgebetrieb

On a synchronous grid? Unless the capacity of the nuclear plant is tiny (which means the economy of scale is lost - who is building a nuclear plant in the area of 10s of MW?), you instantly have voltage instability if you lose supply, and yes, bringing the network back to synchronicity can take days or weeks. This is the same issue we have with coal, which is part of the reason we are trying to diversify the grid (large synchronous machines are not all bad because they do maintain inertia in the network, which means it is easier to maintain voltage stability, but machines like synchronous condensers can provide this without needing to pin our energy requirements on it).

Yes, on a synchronous grid. Yes, everyone does it. Yes, it works.

You seem to be thinking in a binary manner, the plant is either on or off and there is some big central regulator doing the switching. But that isn't how it works.

Plants, through their grid connection, 'see' the grid frequency of 50.00Hz. If the frequency now deviated to 49.99Hz because consumption increased, the generators will be slowed down a little. Not a lot, because they intentionally do have some momentum reserve for very fast variations. Now that out generator has slowed, your nuclear power plant (say a BWR) reacts to this: a slower generator slows the coolant flow to the reactor. The load following will turn up the coolant pumps, cooling down the reactor a little more, leading to less steam bubbles, more moderation, more fission, higher power output. This higher power output leads to higher steam throughput, speeding up the generator.

If grid load decreases and frequency goes up to 50.01Hz, the reverse happens. Faster generator means higher coolant flow. Load following control slows down the coolant pumps a bit, steam bubbles increase through higher temperature, less moderation, lower fission, lower power and steam production, slows down the generator. Things are designed to stabilize at 50.00Hz. The exact amount of power the reactor produces is dynamically scaled up and down to keep the grid frequency stable.

For a PWR, there is some more control logic involved because steam bubbles aren't normally present in a PWR. So load following works by changing control rod positions depending on the grid frequency, which isn't quite as fast, but still in the order of minutes.

So load following is very gradual and non-binary. The myth that you can only on/off power plants is just a myth for almost any kind of plant.

That's really interesting.

I am against nuclear power but upvoted your comment anyway.

Reply on news.ycombinator.com