Hacker News Reader

Comment by kijin | Hacker News Reader

Most of the energy that goes into pumping water uphill can be recovered when the water is released downhill through the turbines. Existing pumped-storage hydroelectric plants have efficiencies around 70~80% which is way better than half.

Besides, you get free power if your upper reservoir naturally gets a lot of inflow or precipitation :)

View on HN

Also Lithium ion battery have a fairly fixed ratio of capacity and power, of about 1MW to 1MWh. That makes them expensive for storing large amounts of power. With pump storage you're much more flexible, just build bigger reservoirs. very large pump storage assets can be optimized even seasonally.

With physical energy storage you also don’t need to worry about the environmental effects of lithium nor do you need to worry about charge cycles sure some maintance is required but the majority of the cost is going to be in the reservoir and that can stand for centuries.

This is some pretty good efficiency. Lithium-ion batteries are more efficient for sure, but also quite a bit more expensive, I'm guessing.

Storing electric energy by just raising some mass seems like a pretty good concept - I wonder why It's not adopted further.

You need a lot of land and specific terrain; still water isn’t carbon neutral(!); and the nature of the things means that if it suffers structural failure the death toll makes nuclear accidents look like almost nothing.

https://www.sciencemag.org/news/2016/09/hundreds-new-dams-co...

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

Lithium batteries are considerably less carbon neutral than physical energy storage when you account for all the energy needed to mine, produce and repurpose them after they fail.

While dams have failed in the past if built right they can last for centuries.

Also unlike batteries where the majority of the cost sunk is in the batteries themselves with physical storage like this the majority of the cost will be in the construction itself which means that as better pumps and turbines become available you can upgrade your system at a marginal cost.

" if it suffers structural failure the death toll makes nuclear accidents look like almost nothing"

You compare a hypothetical dam disaster, with nuclears record so far - its clearly worth reconsidering that comparison. Nuclear plants also consume large amounts of water, and other resources so could not be considered carbon neutral by your own measure.

Reservoirs are large lakes, creating them entails ecological disruption but not consumption or degradation in the long term.

Not sure where this comparison is going? One dam disaster (Banqiao Dam) caused more death and destruction that almost all other disasters combined. In fact failure of nuclear plants has had remarkably few verifiable deaths.

And compared to the energy one nuclear plant produces, dams seem to have vastly worse death-per-benefit statistics.

Damming a river can cause ecosystem collapse downstream. That's degradation, as I understand the word. And its definitely long-term.

I hoped merely pointing out that discrepancy would suffice - how ben_ws comment compared the history of nuclear accidents to the potential dangers of dams.

Potential scenarios are very important in determining probabilistic risk. Simply showing "it hasn't happened yet" is not a valid way to assess risk. This is the reason it costs more to build and keep nuclear plants safe, than it does dams.

Nuclear proponents always point to the Banqiao Dam disaster, and it's always a misleading argument because that dam was built to control downstream flooding as well as provide power and its failure was probably attributable to how the flood control aspect was mismanaged - by holding back too much water for too long out of fear of downstream flooding, they ended up overfilling the reservoir. If China hadn't bothered with hydroelectric power, the dam would still have to have been built and all those people would still have died when it failed, you just wouldn't be able to use their deaths to make nuclear power look better. (Also, nuclear power didn't exist back when that dam was put into operation, and given the quality of engineering that went into it that's probably a good thing.)

Rationalizations can be made for nuclear as well. But it doesn't touch the evidence that dam failure has and may continue to be the far greater risk.

> You compare a hypothetical dam disaster

No, I’m not.

"the nature of the things means that if it suffers structural failure"

What actual dam did you mean to refer to there?

The one I linked to in the original post:

"""Its failure in 1975 caused more casualties than any other dam failure in history at an estimated 171,000 deaths and 11 million displaced."""

Though I could also be talking about the 2,209 who died in the 1889 South Fork Dam incident. (“””Blamed locally on poor maintenance by owners; court deemed it an "Act of God". Followed exceptionally heavy rainfall.”””)

Or the 423 in Malpasset dam, France, as recently as 1959.

Or the 238 in Canyon Lake Dam, 1972.

Or the 1,800 to 25,000 (estimates vary) in the 1979 Machchhu dam failure.

But the one referenced as an example in your original post DID fail - and you wrote "IF it fails". You described a potential failure "in the nature of the things".

It is natural to consider the nature of things - apply equal consideration to nuclear power and your comparison will be more valid.

I literally do not know how to parse your sentence in the way that interacts sensibly with what I wrote.

Interact candidly - don't avoid your error while trying to parse.

You wrote "IF it suffers failure..." so the "it" was not the example of failure you gave, "it" was hypothetical . This is why I asked you what "it" was - not so that you could elaborate on what you could have been talking about, I asked to give you a fair chance to realise "it" was yours to define - a hypothetical dam, one defined only by "the nature of things" and a "death toll makes nuclear accidents look like almost nothing"

Stored hydro & nuclear solve entirely different demands. Nuclear's core strength is providing a strong, reliable, but relatively constant base-line.

Wind/Solar can supplement (or even replace it) depending on costs and geographies, but their generation can be very spiky, and you don't get to pick the spikes (or troughs).

Stored-hydro rounds off the trifecta by providing spikes on-demand, but also by absorbing over-production from other sources.

Nuclear looks wonderful in a "spherical cow" calculation, but the reality is that nuclear strongly prefers a constant production, but demand is never constant. Under-production is obviously undesirable, but over-production is no easier to deal with.

So it's not really about which is better - it's balancing the weak points in each system against the strengths of the others.

When I was in the generation industry we called these base and peaking plants. The 5 nuclear plants were base loaded, a portion of the fossil plants were the load following units then there were the required number of spinning reserve plants. These spinning reserve plants, which are running and using fuel but not producing electricity are what utilities would like to replace first.

There was an article some months back about a concrete block lifting system that could be used for a similar purpose, but obviously a lot more compact and lower capacity.

It has been adopted also flywheel energy storage at scale is by far more common than lithium batteries are today it’s also cheaper and more reliable.

Because it scales poorly, is not easy to deploy, and has miserable energy density.

Reply on news.ycombinator.com