Hacker News Reader

Comment by jhoechtl | Hacker News Reader

Forgive me my lack of understanding of the englisch language, but I do not understand the distinction you make.

Hydroelectricity: Is this power you obtain by placing turbines behind a huge concrete wall of a large river? If so, how is the scenario of building a large concrete dam in the mountains where water is pumped up a hill to covert it into kinetic energy any more better?

I understand that hydroelectricity is used to supply the basic power need, say 80% and a water reservoir with turbines to cover power peak needs.

View on HN

> I understand that hydroelectricity is used to supply the basic power need, say 80% and a water reservoir with turbines to cover power peak needs

This seems like you either didn't understand what pumped storage is, or you've confused several related technologies.

A typical pumped storage scheme goes like this: We find a mountain with a high lake and a low lake. We dam the high lake so that its natural outflow (perhaps to the low lake) ceases. We dig tunnels between the lakes, and we put an electric pump/ turbine in the tunnels.

When we put electricity into the pump, water from the low lake is pumped up to the high lake (until it gets too full and we stop). This stores energy. When we let the water down the pipe through the turbine instead, the same water flows from the high lake down to the low lake, giving slightly less electricity via the turbine.

Unlike the conventional hydro electric power plant this is not really producing electricity, it is storing it, hence the words "pumped storage". We can use this to move electricity in time, which otherwise has to be done very expensively with batteries.

For example, when it's windy in the middle of the night in the UK, the French and Belgians buy some of our cheap electricity, and the stored power hydro facilities in Wales and Scotland also use that electricity to pump water up to their top lakes. The next afternoon, when peak electricity usage occurs as the sun goes down and people begin cooking evening meals, the pumped storage releases the water, making electricity it can sell at peak prices. In the UK the pumped storage plants are privately owned, so they benefit financially from this arbitrage.

Do you know the efficiency of using pumped storage? In other words, for every MW of electricity used to pump the water up the hill, how much do you get back in power generation when the water flows back down through a turbine? You will lose a bit of water to evaporation during this process, but it should be fairly negligible. Also all reservoirs lose some water by leakage into the ground. How does the efficiency compare to various battery storage techniques?

Batteries can get up to 95%+ efficiency, pumped hydro up to 85% efficiency so batteries win out in pure energy efficiency. However, when you take into account the economic efficiency then pumped hydro usually comes out very favorable. Battery prices have been falling rapidly over the last few decades, but simply damming up a mid-sized valley can store such a ginormous amount of water that is it hard to compete.

I would imagine energy density to be a bigger shortcoming (gravitational acceleration times height difference) -- especially for typically accessible height differences. Eg for 500m this is something like an order of magnitude lower than chemical batteries (which are themselves presumably at least an order of magnitude lower than gasoline). This would mean that we need reservoirs to be much larger than equivalent chemical batteries.

> simply damming up a mid-sized valley can store such a ginormous amount of water that is it hard to compete.

I would love to see an analysis of whether it is feasible to build enough such large scale reservoirs (and how many we would need) to store an order one fraction of the daily energy needs. (at city/country/world levels)

> This would mean that we need reservoirs to be much larger than equivalent chemical batteries.

Yes we know. It is still cheaper on a cost-per-kwh basis than batteries, by a significant margin.

> I would love to see an analysis of whether it is feasible to build enough such large scale reservoirs (and how many we would need) to store an order one fraction of the daily energy needs. (at city/country/world levels)

No it is not, there are not enough suitable sites in most places in the world to make this work for world levels. That said, it is entirely up in the air if there would be enough mineable lithium to make batteries for similar amounts of storage.

Efficient electrical energy storage at scale is currently unsolved.

I quite like what this startup is thinking. https://www.energybank.nz/. For offshore wind energy storage.

This Energy Bank system could possibly be practical.

Unlike "Energy Vault" (NRGV), a purely fraudulent investment scam. Energy technologies seem to be favorites of frauds (fusion especially so). It seems like nothing is so obviously nonsensical as to attract the attention of regulators.

There are, in fact, far more than enough suitable sites to store as much energy as we could ever care to store. Hydro power generation needs a watershed, but storage really needs only a hill.

But there are lots of different storage technologies, and costs are falling fast, so pumped hydro may be undercut in places.

Lake volume is just that: it's volume, plus the dam, times two for the lower reservoir. Battery storage facilities however mostly consist of maintenance access, scaffolding, temperature control and fire suppression, we don't just dump cells on a big heap and call it an energy storage solution. Pumped hydro is doing fine in density, particularly since we rarely think in volume for large facilities, we think in acres, and nobody would build a battery facility vertically stacked.

Lets take a pumped hydro I know of, https://en.wikipedia.org/wiki/Taum_Sauk_Hydroelectric_Power_.... It's 5,370,000 m^3 of water, 3600 MW·h for 8 hours to empty, 28,800 MW total. Cubic meter of water weighs 1,000 kg aka a metric ton. So 5,370,000 Metric tons of water.

Looking at https://energetechsolar.com/1mwh-500v-800v-battery-energy-st..., they're around .88 LB / AH, so 880,000 LB/MWh. 399,161 KG / MWh. 11,179,036,800, or 11,179,036 Metric tons of batteries. I don't know what the breakdown is but global lithium production was 100,000 metric tons in 2021, which is off by 2 oom.

My math may be wrong, and you do have losses of water from evaporation, but with pumped you can go up and down. But that's 11 million metric tons of lithium and other metals. Water falls from the sky (in most places still). Lithium has to be refined. You can recharge the battery, you can reload the water or pump it the other way. Tomsauk is mostly concrete and water (by volume). You do need almost a mountain for that head, or a mine. Concrete breaks down over time and can be maintained,. Batteries wear out as well but can be rebuilt. Water seems the easiest of the things to replace currently. The only things I can think of that are easier are provably gasses and maybe salt or rock.

This is all napkin math, and I may have missed a decimal some places. But they seem within the same order of magnitude for efficiency. But is it even possible to build batteries that big?

Of course when they got greedy with Tomsauk they ruind a great natural area. I'd love to be corrected on my math and assumptions.

If this still builds a dam, doesn't this have the same environmental concerns as any hydro project? Or is there something that makes this less of a concern?

Ideally you don’t build a dam.

Niagara Falls sits between Lake Erie into Lake Ontario. Which is the perfect location to use pumped hydro because all you need to build is a pipe between them and put a turbine inside the pipe. If you have excess energy pump water up from Lake Ontario to Lake Erie, and when you need power run that in reverse to generate electricity.

Move enough water and the water level on each lake will change, but 6 inches (15cm) isn’t going to change anything of note and that represents an insane amount of energy. Something like 500 GWh if I remember correctly. Unfortunately that’s literally the best case in the US, nothing else even comes close.

The scenario you've described still begs the question though. It seems the only difference between traditional hydroelectricity and pumped storage is how much you disrupt the natural flow from Lake Erie to Lake Ontario. The hard limit is when the Niagara Falls run dry. Anything more than that is basically spending electricity to create a treadmill for fish, which may be the environmentally responsible thing but we should be clear about that. And since fish probably cannot jump up Niagara Falls, it's probably more like spending electricity to run a really large water display for tourists and the occasional person who wants to go over in a barrel. But maybe that electricity use doesn't matter if we're getting more than we need from solar during sunny days, and just need predicable storage for nights and cloudy days.

I would think the stronger argument for pumped storage would be in a place where the high level of water did not naturally exist, and so the only way it gets up there is by pumping. But perhaps this still destroys too much of an ecosystem even if it's not a river ecosystem.

We let stuff migrate upriver years ago when they added a lock system back in 1829. At this point any environmental harm from connecting these lakes has already happened generations ago well before we where born. https://en.wikipedia.org/wiki/Welland_Canal

Anyway, the amount of water flowing over Niagara Falls is currently regulated hourly by treaty with excess flows above that level used to generate hydroelectricity. 100,000 cubic feet per second (2,800 m3/s) of water flowing over the falls, and during the night and off-tourist season there must be 50,000 cubic feet per second (1,400 m3/s) of water flowing over the falls. That excess is generally 50-70% of the rivers total, making the falls arguably just a really large and extremely expensive water feature used to attract tourism. https://en.wikipedia.org/wiki/List_of_Niagara_Falls_hydroele...

Adding a separate pumped hydro system really can be treated as an independent entity because we don’t just control the falls we can even turn it off when needed. https://www.dailymail.co.uk/news/article-1338793/Niagara-Fal...

Connecting two bodies of water for the occasional organism that wants to pass through is a distinct thing from disconnecting two bodies of water for say large numbers of fish that would otherwise swim up a river. The latter is a usual criticism of hydro power, which seems like it doesn't apply to Niagara because it's such a big waterfall.

I wasn't aware of the treaty, that makes sense.

The link about turning off the American side of the falls doesn't really support the implication that there is enough hydroelectric capacity to use up the entire flow of the river. The simplest explanation is that the flow was diverted over to the Canadian falls.

I assume they sent the water over the Canadian side simply due to the treaty. That said, they have excess capacity to handle blocking 1/2 the flow over Niagara Falls every night, but even doing nothing was still a non issue.

If you assume they blocked 1/2 the flow (1,400 m3/s) and didn’t use it for anything. That would still take 21 days to raise water level of Lake Erie by 1cm.

Pumped storage can also be used with existing hydro stations, just have a reservoir at the bottom to save water that can be pumped back up to the top of the dam. See https://www.nwcouncil.org/sites/default/files/MarkJones_1.pd... for the one at grand coulee.

California has dozens of hydro power reservoirs high up in its Sierra Nevada range, many built a century ago, with penstocks down to power stations far below. The penstocks have lately been fitted with pumps to push water back up.

Pumped storage does not, in fact, depend on finding conveniently placed lakes.

That is cheapest, but building a dike for a reservoir works fine too. A conveniently placed box canyon may need a dike only at one end.

> This seems like you either didn't understand what pumped storage is, or you've confused several related technologies.

What adds to the confusion is that Germans - OP used a common typo that hints at his native language - casually use "water power" ("Wasserkraft") both for hydroelectric dams ("Wasserkraftwerk") and pumped storage ("Pumpspeicherkraftwerk"). It's one of these fascinating little insights in how different languages form different train of thoughts.

I'm curious, what was the typo? The grammar hinted at German, but that's all I can see.

"englisch language" ... that the English 'sh' is pronounced almost identical to the German 'sch' is a common source for error.

Because you're not disrupting a river's flow and natural ecosystem, you only utilize the watershed some more during the initial filling period. After that, it's a closed system separate from the river ecosystem where only evaporation causes leakage.

That ship has already sailed as most power producing dams have been around for long time. Removing them would destroy the (now natural) ecosystem they have created. Plus dams have way more benefits to us than the extremely clean energy they produce - not just store.

This is the first time I've seen a claim that dams create ecosystems that we should worry about destroying. As far as I know, they disrupt fish and animal migrations, reduce natural carbon sinks, reduce biodiversity, and block sediment from providing nutrients downstream.

Is there any source saying dams create healthy natural ecosystems? A necessary evil I can kind of see, but a net positive for ecosystems? That seems doubtful.

I don’t know. I suspect the bias is towards the pre-dam ecosystem so it might be challenging to rationally compare them.

Certainly an ecosystem will grow up around nearly any situation.

there are many human ecosystems living below the dam though. Removing the dam means removing entire towns downstream in some cases.

That is not, in fact, true. Almost all dams have a river flowing below them, where it always did.

A very few places use all of the inflow for irrigation.

Yup my opinion only. To me water is life and if you let it flow to the ocean where it’s only use is to dilute saltwater you’ve wasted a precious natural resource.

Dams does not save any water except the initial filling. Their inflow and outflow match over a seasonal cycle.

Also, they tend to cost water instead: the larger surface area increases the amount "lost" to evaporation.

Perhaps. I suppose that would depend on size and depth of the reservoir and the makeup of the river. Also without a dam the water would be traveling the river mostly during the cold wet season with little evap. So yeah likely so. But in the end the cost would be lost saltwater. Probably a benefit for those that are worried about sea levels.

Yup that’s simplistically true. The difference though they do allow you to choose when it gets released. Around where I live that allows irrigation of crops during the summer months. That water does not go into the ocean. Also around here many of the municipalities draw water downstream from the dam as well. A portion of that travels through the sewer systems, gets purified then dropped back into the river.

This is not the case, removing dams still can come with huge benefits to the ecosystems those dams disrupted: https://www.mercurynews.com/2019/05/08/four-years-after-cali...

Read the article. Sounds like in that instance it was a good choice to remove the dam and had a good outcome. Well based on the information presented anyhow. Also I’m not sure if the main justification for removal pencils out for me. Yes some fish - 123 so far - were able to reproduce further upstream by 7 miles. What would have happened if the dam was still in place? Breed just below the dam? Do the fish really care where they do the deed or lay their eggs? Can’t believe that would be very true. Heck when I was their age I was pretty happy to do the deed and wasn’t too too awful concerned where it was at either. :D

No. First off, most dams silt up, not only not of being use in storing water but also not in creating power. Second, once they silt up, whatever ecosystem there is also destroyed and artificially created ecosystems are never natural and usually are bad for local fish populations. Artificial lakes can also host a number of non-native fish and are not any benefit to the native species that should be living there. Also, dams also block the migration paths of native fish.

Just think of this way, dams have a life span and for the most part, most of them in US have hit the end of their life spans and should be removed. They serve no purpose to anyone.

Right. Storing water for 50M+ people and generating enormous quantities of electricity. No use at all.

This is a pretty incomplete understanding of the issue, there are plenty of dams where silting is not an issue at all and will serve their purpose for centuries if properly maintained. There are also dams where silting is a regular problem. A so called 'sediment bypass' is often a feature of these dams, where water is pulled from the base of the dam through a bypass to clear out silting.

It all depends on the natural hydraulic dynamics of the dam - some are better placed than others.

We are, in fact, demolishing dams at an increasing rate.

Dams very commonly destroy fisheries worth far more than any value the dam can deliver.

Pumped hydro is burning energy to pump energy into a (usually artificial) reservoir.

Then, later, allowing that filled reservoir to drain through a hydroelectric plant and generate electricity.

This solves 2 major problems, compared to alternatives.

1) The scaling costs (with respect to energy storage capacity / volume) are extremely low. E.g. instead of building twice as much battery, you're just consuming additional empty space behind the dam.

2) There is no time-cost to storing energy in pumped hydro, which affords great flexibility in when it's stored and when it's extracted. E.g. stored on a bright day when the wind is blowing, extracted at night when there's no wind.

In essence, it turns spikey supply and demand patterns into more constant ones, which are cheaper and more efficient to service.

I think the idea is to start pumping up water at existing hydroelectric power plants.

If you do this using wind or solar power the dam becomes a giant battery you charge.

Where does this extra water come from?

It’s just borrowed

Hydro power work by water flowing down turbines from a higher body of water to a lower.

So you can typically pump the lower water back up, using wind or solar power.

Wouldn't this mean the downstream would have no water flowing through it?

Pumped hydro can be built entirely off rivers. There's a plan to put a reservoir for one on top of a mesa in the desert in Arizona, for example.

What does the location of the reservoir matter? If you're taking a river and pumping it up on top of somewhere (a mesa in your example) aren't you going to cause the river to basically run dry everywhere downstream of where your pump is?

There's no river at the bottom, either. I thought the "desert" part would make that clear, but I guess it had to be said.

Water in this system is part of the capital cost (charging it up when you start), then a minor maintenance cost (replacing evaporation). Otherwise, unlike primary hydro, there is no large constant flow out of the system, and no need to be on a river.

The water lost to evaporation is at least an order of magnitude less than water evaporated from a nuclear plant of the same levelized power.

Wouldn't you need two reservoirs in that case? Otherwise where is the water being drained down to take advantage of the potential energy?

Yes, of course. Off-river pumped hydro involves two reservoirs.

I think the confusion is a couple of comments up you said "There's a plan to put a reservoir for one on top of a mesa in the desert in Arizona" which is really confusing since of course they're planning to put a pair of reservoirs for one in the desert in Arizona. I guess only one is on top of the mesa though.

So at that point why use a river? Couldn't you just use ocean water? It's comparatively "free" and we aren't going to run out anytime soon.

With a high enough plateau near an ocean, that should work.

I think those places are really rare though.

You need a reservoir to store that water downstream as well, it can't just be a river

Ya, this is how they do it. Eg see https://en.wikipedia.org/wiki/Dry_Falls_Dam

Think of pumped storage like a battery: In areas with a lot of solar, pumped storage runs when there is a lot of daylight to store the energy for nighttime use. Pumps run when the sun is out, turbines run at night.

Or the opposite in areas with wind/nuclear. Pump it up during the night when demand and prices are low, then run turbines during day when demand and prices are higher. Arbitrage.

Reply on news.ycombinator.com