> It doesn't always need the hill: underground cavities work to pump water up out of, and to drain into.
Do you have any good numbers on real world projects? I'm very happy to be wrong here, but all the numbers I can find are either lies from nuclear shills or using existing watersheds. Most also only focus on the cost per kW which is higher than batteries and not the relevant metric (as batteries can drain in a few minutes) for season-long storage.
Also a quick back of the envelope seems to suggest emptying and filling lake Baikal could store as much energy as about a billion tonnes of chemical storage. This seems like a reasonable upper bound which would indicate pumped hydro is about an order of magnitude short of solving the problem. Current battery production is nowhere near (total cumulative seems to be about a megatonne chemical equivalent even if it is more than doubling annually it'll take over a decade to catch up), but this is expected because batteries are optimal for short term.
Overall by gut feel it seems a more feasible to make and store ten cubic kilometers of chemical fuel worldwide than move 200,000km^3 of water around.
Most current pumped hydro uses existing dams because duh. But nobody is building those anymore, for reasons you note. Existing hydro power dams were expensive because they needed to be deep to store years of water, and concrete because deep water has high pressure. They destroy ecosystems because that is where the water comes from.
Dedicated pumped hydro storage is typically quite shallow, with an earthen dike (if needed at all), and the only place with high pressure is at the bottom end of a penstock. It does not need to store years of water; just a day's worth is useful.
To follow up, fengning is on an existing river, uses an existing lower reservoir, has favorable geography, cost between $1.8 and $3 billion somewhere and stores 40GWh with 3.6 GW power.
This makes it better than existing batteries for ~1 day time scales and roughly on par with the upcoming generation of things like sodium batteries.
It's not a clear indicator as it's obviously optimized for power, but these all seem to be big advantages specific to the site which would indicate that an artificial reservoir would have trouble competing even against batteries.
I could believe that you might improve storage/cost by a factor of 10 if you found a suitable reservoir by reducing power, but that seems to back up my initial assertion that you need specific geography and to significantly change the ecosystem fairly well.
As such it seems like it is not much better than a battery for displacing fracked methane, oil, or nuclear for mediating seasonal variability (which is what synthetic denser-than-hydrogen fuel is for as it is optimized for approximately zero cost per capacity at the expense of the highest cost per joule with competitive cost per watt).
Plus batteries still have a 10-20% efficiency benefit.
Pumped hydro, batteries, and tanked chemicals are not the only storage media.
We also have underground compressed air using existing deep cavities, and undersea compressed air. And underwater buoyancy, drawing floats down toward seafloor-mounted pulleys, using a winch and motor-generator on shore. Demand will not exceed our capacity to make cables and floats.
But the real answer is that there is nowhere even close to as great a need for long-term storage as you imagine, just as there is not much petroleum stored today. Petroleum is extracted and delivered continuously and reliably. Myriad tropical solar farms will synthesize ammonia year-round, shipping anywhere needed on demand, so storage is needed only until the next shipment arrives.
And HVDC transmission lines will move power from where it is being produced to where it is not, over 1000s of km, at a wholly tolerable loss rate. Much of this will move power eastward from afternoon production and westward from morning production, but also generally fill in for local production and storage shortfalls everywhere.
So Finland can have ammonia shipped in continuously all winter long, just as they ship in petroleum and NG today. Transmission lines will compete for that business.
Further looking into HVDC, it seems to be about $500k/km. Over distances of 4000km, that's $2/Watt ($2.20 with losses) or about the cost of producing the electricity in the first place.
Better than fuels at present, but inherently fragile so not a complete solution. Also those are just claimed building costs not including operation (and assuming it lasts about as long as the pv), and the natural monopoly spawned always winds up being a massive tax money sink, so it's not clear that building thousands of hvdc lines is going to work out.
What is inherently fragile in HVDC compared to conventional transmission?
AFAIU it is even less fragile because it's usually 'point-to-point', which means to integrate it into the grid you need very modern substations with the ability to 'transform' the DC into the AC of whichever pre-existing grid by means of mass cascaded https://en.wikipedia.org/wiki/Insulated-gate_bipolar_transis...
Which in turn makes the grid around these substations smart, the more, the smarter. Because you have much better ability to switch and regulate(diversion from same frequency in AC-grid due to dynamic load or failure) much faster.
Think of the difference between the large external 'power-brick' for older laptops vs. the small switching power-supplies for contemporary notebooks. Just in reverse.
> What is inherently fragile in HVDC compared to conventional transmission?
Nothing? The point is transmission is extremely fragile. To physical conditions (weather, accidents, faults). To market effects (massive price gouging during those failures due to lack of buffer). And to market failures (harmful monopolies always form around private utilities and neoliberals always privatize utilities).
Then if the oceans are involved, cost becomes a non-starter.
> We also have underground compressed air using existing deep cavities, and undersea compressed air. And underwater buoyancy, drawing floats down toward seafloor-mounted pulleys, using a winch and motor-generator on shore. Demand will not exceed our capacity to make cables and floats.
These are more technologies that compete with batteries, not fuels. And poorly at that. CAES in ideal sites might compete with batteries for a while yet, but the others do not. Even a single truck full of ammmonia can store the equivalent to tens of thousands of cubic metres worth of displacement storage. A cubic metre float in a 500m deep body of water can store 5MJ, or about the same as a battery you can lift with one hand and buy for a few hundred dollars.
> But the real answer is that there is nowhere even close to as great a need for long-term storage as you imagine, just as there is not much petroleum stored today. Petroleum is extracted and delivered continuously and reliably. Myriad tropical solar farms will synthesize ammonia year-round, shipping anywhere needed on demand, so storage is needed only until the next shipment arrives.
Moving energy 2000-4000km as not-electricity is strictly a harder problem than storing it for 6 months and is solved with a subset of the same solutions. The main upside is energy input is cheaper and there is less idle time. This is definitely part of the solution but comes under the same heading.
HVDC systems are a solution for most of the issues, but distant solar isn't completely uncorrelated. Also no country is going to stake their survival on a system with cascading failure modes that can be triggered by anything from war, to a heat wave, to a blizzard, to a cyclone, to a forest fire, to political games. Thus capacity for months of backup is still required.
All of this is moot anyway because your other comment led me to find sources stating green ammonia is already within 50% of cost parity with fossil fuels for the cheapest solar energy sources and ammonia fuel cells are now viable as well with the same technology. Between that and thermochemical storage, fission is obsolete immediately and fossil fuels only need one more price shock for the transition to start.
Also none of this supports your original assertion that pumped hydro meaningfully exists in a non-ecosystem-altering way.
No they don't. Proposing international transmission from the equator over a hanful of many thousand km long lines in competition with your neighbors in all directions as a sole source of winter energy is far more fragile than a mixture of local fossil fuel reserves, nuclear, local production, and imports from any direction from immediate neighbors.
PVs solve net energy needs. Perovskites might even make them do so without needing strategic mineral reserves. But they don't really provide energy security, and keeping fossil fuel infrastructure working for 1 month in 30 is extremely costly.
> You are always free to invent falsehoods about pumped hydro, as about anything else.
Then show the real numbers. I did. Demonstrate it being viable as a significant portion of primary energy in a typical country as a new project.
A day's worth puts it in the power limited regime where $100/kWh 4C batteries are already close to viable, and can use the 50x higher power per $ and higher efficiency for minute by minute arbitrage to offset costs. Do you have sources for real projects that can beat $60/kWh capacity and $600/kW power (what you'd be competing with by the time construction finished)? Moreover it also needs to beat hydrogen or methane storage (electricity->chemical-electricity) which (sans capex for tanks because I can't find good numbers, but I think it adds about 20%) is about $100/MWh out and $1000/kW using current technology including energy and projected to fall to somewhere around $40 and $500 in realistic timescales.
To make it impossible to mine more fossil fuels even for a mixture of slave-driving sociopaths unrestrained by law and theocrats actively seeking apocalypse we need to be able to use a MWh at night in mid winter that was produced at 2pm in summer for less than around $40 and then do it another billion times without hitting some resource limit. Hydrogen with storage is shockingly close, and if synthetic ammonia/methane or metal hydride get over the line, noone will look at fossil fuels again.
You seem to be telling me that pumped hydro is already there, but I can't find a decent source agreeing with you (or any numbers dealing with this use case for that matter).
What I am saying about pumped hydro is that it is deeply mature technology, with no surprises in store. All the equipment is essentially unchanged for decades, except for control-system electronics. All that is new is reservoirs not fed by watersheds.
There are actually dozens of shallow reservoirs behind earthen dams way high up in California's Sierra Nevada mountain range, many almost a century old, constructed with bulldozers that used pulleys instead of hydraulics, hauled up there on fantastically bad cart-track roads. The reservoirs feed penstocks down to Pelton wheels thousands of feet below. For storage, they just attached pumps to the penstocks to push water back up.
> What I am saying about pumped hydro is that it is deeply mature technology, with no surprises in store. All the equipment is essentially unchanged for decades, except for control-system electronics. All that is new is reservoirs not fed by watersheds.
And I can attach a washing machine motor to my water tank. If it produces basically nothing and costs double the alternative, it's not relevant to the discussion.
> There are actually dozens of shallow reservoirs behind earthen dams way high up in California's Sierra Nevada mountain range, many almost a century old, constructed with bulldozers that used pulleys instead of hydraulics, hauled up there on fantastically bad cart-track roads. The reservoirs feed penstocks down to Pelton wheels thousands of feet below. For storage, they just attached pumps to the penstocks to push water back up.
So there are no new projects which not destroying an ecosystem that are on cost parity with batteries then?
Also as an aside, how pathetic is capitalism as an organizational system that we can't achieve the types of things that were done with horses and carts and pulley bulldozers in the past?
This is always claimed, and is always false. Please do not repeat falsehoods.
Existing hydro generation needs a watershed. Pumped hydro does not. All it needs is a disused hilltop and earthen dike, and not always the dike.
It doesn't always need the hill: underground cavities work to pump water up out of, and to drain into.
Batteries will always be the most expensive alternative. They will be used in limited amounts, mainly for very short-term (overnight) storage.