It makes sense for solar as well because it produces a peak at noon but not much in the early evening when a.c. demand is still very high.
I have an off-grid family member running their lighting+IT loads straight off DC just so the inverter can sleep at night. Its efficiency is garbage with small loads.
However, all indications at the moment are that solar + lithium storage is winning this battle. We are seeing larger farms with more installed storage every year.
Perhaps CSP might flourish in certain niches, particularly those where waste heat and / or industrial high temperature processes are required in addition to electricity production.
Same is happening with CSP, just way less so in the US. CSP with MS storage is already cheaper on a LCOE basis than nuclear, esp factoring storage [0]. Lots of PV analysis never includes storage costs (and for obvious reasons).
I think Li tech is probably better served for individual consumer applications (because of all the constraints) than at the grid level.
Li is convenient for residential installations, but it's possible other batteries are better at industrial and grid scale.
Lazard's latest (2020) levelized energy cost comparison[1] shows the unsubsidised cost comparing like this:
Nuclear: $129-$198 MWh (1st chart)
Solar Wholesale PV + Storage: $188-$329 MWh.
This excludes the (huge) further price reductions in PV noted in the linked article. The wholesale solar does include flow batteries as well as Li storage.
The whole storage report[2] is worth reading[2].
[1] https://www.lazard.com/perspective/levelized-cost-of-energy-...
[2] https://www.lazard.com/media/451566/lazards-levelized-cost-o...
It wouldn't be impossible to move vats of molten salt at scale to different locations from collection for usage via HEX and typical Rankine power cycle systems (i.e. steam).
Here is a very efficient method for moving power: a copper cable.
Right, texas natgas producers should just run a copper cable directly to places like Singapore!
I think its because you missed
> ports
Other energy products are shipped on tankers: CL, NG.
Once they arrive at a certain country, they can be disbursed over roads and rail to plants; CI and residential alike.
Molten salt will vary depending on its heat capacity and delta_k above freezing point.
0.338048298841 MJ/liter for the typical NaNO3 KNO3 NaNO2 mixture (assuming a delta_k of ~300)
53.356955169699994 MJ/liter for something like CaCl2 45.5682%, MgCl2 54.4318% (Ionic halide, assuming a delta_k of ~1568)
In the future I expect even higher delta_k for molten salts.
[0] https://en.wikipedia.org/wiki/Liquefied_natural_gas#Specific...
I think we can go ahead and rule that out as impractical given we're talking about temperatures which are high enough to melt steel.
I don't think steel will be used as the only form of confinement (certain kinds of ceramics come to mind can handle it now[0]), so agree to dissagree.
[0] https://en.wikipedia.org/wiki/Ultra-high-temperature_ceramic...
Yes, if needed ("to the other side of the world" could very well mean from AU to SG, and not USA to SG [although that happens now in some cases to meet demand]). LNG and CL are shipped across the world so that it can make steam to make electricity now (as well as for heating).
I can imagine it might be hard for some to fathom that doing it this way may work out better at scale than trying to build CSP plants in regions with low irradiance/ limited land availability/etc.
There are plenty of places that don't have much sun that are already connected to places that do. They are are connected by copper wires.
There is no universe where trying to ship molten salt and keep it hot so it can turn water into steam is viable. Oil doesn't rely on heat gradients and doesn't lose its energy when put in a tank for a week.
A copper cable carries electricity indefinitely. Also remember that everywhere already has electricity from somewhere. Solar generation is not shipping from one place to another far away, it is offsetting power to somewhere in between.
Because it is an option, just like it is an option for those to place a bunch of PV down and not bother with storage costs for that. Some options will work for some and not for others.
And electricity needs to come from some source, ideally one that can provide more than +4hs and affordably.
> A copper cable carries electricity indefinitely.
All fields fall with distance from the source… shipping a source (like LNG, CL, and perhaps MS in the future) around to local nodes where it can be delivered with copper cables (after burning and combustion) is not beyond current logistics (and nor is factoring evaporation losses with LNG now).
> There are plenty of places that don't have much sun that are already connected to places that do. They are are connected by copper wires.
Right, tell that to every country that still imports hydrocarbons to burn for electricity/heat…
> There is no universe where trying to ship molten salt and keep it hot so it can turn water into steam is viable. Oil doesn't rely on heat gradients and doesn't lose its energy when put in a tank for a week.
I'm pretty sure that it will eventually be possible to have vacuum storage to keep it hot for longer than 17.5 hours with enough delta_k above freezing to provide a predictable amount of energy, so saying "There is no universe" would only apply to those like yourself who deem it beyond the realm of reason (which many things in this world tend to be for those who have walked it, thankfully progress isn't limited by such people in the long run).
Interesting that the notion of a multiverse you appeal to with "There is no universe" seems more realistic than the above…
> Solar generation is not shipping from one place to another far away, it is offsetting power to somewhere in between.
Solar collection is when you need to think about the infrastructure and the surface area needed if you want to do it on a large scale; perhaps beyond realm possibility of what some deem needed for any place in particular…
Electricity already comes from a lot of sources. It is pretty myopic to think people talking about solar panels means replacing all electricity generation with solar. Solar installations already pay for themselves in under 7 years in sunny areas. Peak electricity use is in the summer during the day, when solar would have the most impact and can be double the energy used at night.
> Right, tell that to every country that still imports hydrocarbons to burn for electricity/heat…
I'm not sure what you are trying to say here, no one thinks that all potential solar power is already installed, this is about what happens in the future with rapidly decreasing solar prices. Also I don't know why you are talking about entire countries, all the parent poster said was that 'they live in an area without sun'. You keep bringing up shipping molten salt around the world for some bizarre reason. Electricity grids are meant to route power.
> I'm pretty sure that it will eventually be possible to have vacuum storage to keep it hot for longer than 17.5 hours with enough delta_k above freezing
Lol, show me where this anything like this has made money. Coal isn't even as viable anymore and that is dug out of the ground and shipped by rail. Molten salt would have to be super heated and then enormous amounts of energy would have to be used to move it.
Do you actually think that molten salt, with a melting temperature that is double what it takes to start weakening steel as well as being corrosive (https://www.bbc.co.uk/bitesize/guides/zjb2pv4/revision/2) is going to be better to transport energy than copper cable (which is already run everywhere with electricity !)
Right, tell that to all the places rolling out CSP…
Very true
> It is pretty myopic to think people talking about solar panels means replacing all electricity generation with solar.
I don't think that all will be replaced with solar panels in the future, but considering the amount of energy that is absorbed by earth every second and the available surface area to collect and concentrate it, I think it possible that most of the energy will come from it (solar in general, not just solar panels).
I still think roll out will be slower in some areas and faster in others (mostly due to the incentives of all the different actors and their relative exposure to all the existing infrastructure).
> Solar installations already pay for themselves in under 7 years in sunny areas.
> Peak electricity use is in the summer during the day, when solar would have the most impact and can be double the energy used at night.
Yeah, its the short term time horizon (7 years vs 30 years) for profitability and looking at peak providing capabilities that attracts more projects in the US now (possibly because its a complement to all the existing hydro carbon infrastructure in place that wont be needed when more electricity generation can come from PV and CSP, that others places in the world don't have to the same degree and continue to build out CSP).
> Lol, show me where this anything like this has made money.
Cerro Dominador Project connected to the grid in chile in april this year with 110MW of CSP, so it's making money for someone in the chain (they also have 100MW of PV). Besides, I think nowadays cash flow matters more than present day absolute profitability, something that O&G producers have less of every year as demand growth slows due to higher end user prices and extraction costs rise.
> Molten salt would have to be super heated
That's what happens when you concentrate a large area of sunlight onto a receiver that contains the salt …
> and then enormous amounts of energy would have to be used to move it.
We already use "enormous amounts of energy" to move energy related products…
> Do you actually think that molten salt, with a melting temperature that is double what it takes to start weakening steel as well as being corrosive (https://www.bbc.co.uk/bitesize/guides/zjb2pv4/revision/2)
Not all molten salts are created equal, nor do I think steel will be predominately used in most casing that actually touches the salts with higher delta_k and high heat capacities, UHTCs come to mind.
> is going to be better to transport energy than copper cable (which is already run everywhere with electricity !)
You make it seem like I'm saying that copper cables wont be used at all, where as I don't think it makes much sense over long distances… even now grids need to have power stations every so often in between all the copper cables… power stations that have to consume some source…
I've seen some disingenuous discussions, but this is something else.
No, I still think it will be possible. Power stations will use molten salt instead of NG and CL now, and they wont have to replace all of their existing systems that convert NG and CL into electricity through steam turbines…
> I've seen some disingenuous discussions, but this is something else.
Yeah its something else, for some reason you are unable to comprehend the possibility of something being (or the scale of such) and chose to ascribe it to dishonesty… nothing dishonest about me thinking based on the research thats being done in many areas wrt theoretical and live systems that the above will not be possible any more so than appealing the notions something not being possible even within multiple universes (which remain firmly in the realm of untested theory)…
Luckily possibilities aren't limited to such ways of thinking… nor to the work that's being done now to build toward such a future.
If you want any idea of where the industry is at now, i suggest reading more [0]
They have no mention of what they are even going to sell the energy at, and no mention storage capacity (if any will be used at all) and not expected to come online for another 6 years.
We'll be able to learn more about it when they actually have an accepted bid from some one.
Yeah, in the latest Lazard's:
"Lithium-ion chemistries continue to be the dominant storage technology for short-duration applications (i.e., 1 – 4 hours),"
The last one I talked about on hn with someone else last year [1] was from 2018 [2] and the costs in this once for PV and PV + storage seem in line with that (or more expensive in the new report at the high end, but this new report only really covers PV and not anything else like the 2018 one, but at least this one really breaks down the storage).
Their nuclear costs then were $112-$189, and their solar thermal tower + storage were $98-$181 (I'm assuming 2021 costs would be lower here from all the low cost stuff that has came online since, but haven't seen a high level report yet).
Baseload (which i think nuclear really falls and I think what they would consider BTM? but maybe theres a whole sale market for it? hard to really tell with their storage definition capping out at 4 hours) still seems far away:
"The economic proposition of C&I behind-the-meter (“BTM”) projects remains challenged without subsidies", ranging from $214-$292 for C&I (no storage), $392-$507 (PV + storage), and $489-$662 for residential (PV + storage))
> This excludes the (huge) further price reductions in PV noted in the linked article.
Not sure we can consider panel costs (Perovskite, Bi-facial Panels, Doped Polysilicon and Bigger Wafers, Better Cells) speaking for storage costs, and the article really seems to be addressing panel costs. Though improvements in panels are always great.
[1] https://news.ycombinator.com/item?id=23588691
[2] https://www.lazard.com/media/450784/lazards-levelized-cost-o...
Projects like the proposed 4500km Australia-SG link[1] mean you get a lot of flexibility to build supply in places different to where the demand is. That solves a lot of the baseload issues.
[1] https://en.wikipedia.org/wiki/Australia%E2%80%93ASEAN_Power_...
Yeah, would like to see more improvements there. Running up to fundamental limits with that though. Perhaps large batteries can be charged and moved elsewhere, but that seems more feasble with molten salt tanks now (but would become more feasble when 4+ hour storage for chemical batteries gets cheaper).
> Australia-SG link
Yeah, really hard to look at prospective projects in terms of lcoe (maybe we'll get more insight with any bids for providing via that link in the future) so we'll really have to see and I don't see anything about MW or MWh costs listed there (just upfront) so i'm assuming only day time baseload.
Liquid Air or Liquid Metal batteries maybe.
However I would argue LFP based Li battery will likely win the cost battle for quite a while. And we are getting to a point in their scale where they are actually quite relevant for large storage.
Li production capacity has come a long way since Tesla battery in Australia.
I worked in the same building as them some years back, and they installed backup for the solar system in around 2016 there.
Some of the companies who wanted to build large scale flow batteries have had a lot of problems in the last decade and very few to non have managed to do it.
RedFlow seems to have some aspiration for grid scale but it seems to just assemblies of smaller batteries, rather then what most companies hope a flow battery for the grid to look like.
There are lots of advantages to smaller batteries and few (no?) disadvantages. I've never heard that being brought up before.
The whole main concept why so many people think flow batteries are so awesome for gird storage is that you can basically have a tank size can scale arbitrary.
The once that actually have some reasonably big pilot are Vanadium redox battery. There are multiple such batteries being developed.
There are also a number of startups who try these flow batteries but try to move away from Vanadium. Li-Ion batteries are collapsing in price so fast, that they have killed of many of these ideas, like the Zebra battery. So newer startups are looking at something cheaper.
But the approach of scaling many smaller ones might still work, it just basically defeats the main benefit of flow batteries in the first place.
But at grid scale there are lots of options for storage. Pump solar is the obvious option.
Esp if one ignores storage costs.
> But at grid scale there are lots of options for storage.
Agreed.
> Pump solar is the obvious option.
Disagree for CSP, but for hydro electric I'd agree.
What’s kept CSP limping along is you can use natural gas to supplement CSP while calling the project green. But from a pure cost basis it’s extremely expensive.
Really? Any cost stats for that, I'd love to see that.
From Lazard's 2018 [1] and 2020 [2] reports, PV+storage (and only 1-4 hours of storage) is still more expensive for "In-front-of the meter" on the low end compared to CSP for storage on LCOE basis, and 2.5x-4.4x more expensive on the low end, for "Behind-the-meter" applications.
> What’s kept CSP limping along is you can use natural gas to supplement CSP while calling the project green.
And we can also cast light on the manufacturing process behind panels and how the materials are gathered for both panels and batteries… either way improvements are being made all around.
[1] https://www.lazard.com/media/450784/lazards-levelized-cost-o...
[2] https://www.lazard.com/media/451566/lazards-levelized-cost-o...
People make the same textbook argument for "rational" actors in "markets"… as if all actors are "rational" or agree on what is "rational". As if everyone has access to the same information, as if everyone was equally invested in the same things, as if everyone wanted to protect the same things…
> but how do you explain that nobody is building them?
> Globally, solar thermal has almost come to a complete stop. So, how do you rationalize that it's a great technology but nobody wants it?
"Nobody"[0][1]
[0] https://en.wikipedia.org/wiki/List_of_solar_thermal_power_st...
[1] https://en.wikipedia.org/wiki/List_of_solar_thermal_power_st...
Luckily, that doesn't limit what other can explore and pursue :D
Look, panels are great. I've used them in the past in certain applications (esp those that wont require any storage, or have 24/7 access to sunlight like in some regions in space, etc), but storage (1-4 hours? lol) costs are a joke compared CSP offering 24/7 and providing electricity 17.5 hours without direct solar radiation[0].
[0] https://en.wikipedia.org/wiki/Cerro_Dominador_Solar_Thermal_...
That is starting but barley happening.
Utilities need to hit install some renewables and setting up some solar is easy. There likely is already a gas peaker somewhere so no need for storage.
I have not studied CSP in any detail so I can't say, but the reality is rational market actor in the current system does not optimize now for what the theoretical best solution is in 2040. They just need to hit their renewable quotas right now.
Suppose you want 1GW of solar 24/7. Well 4h is roughly 1/2 the output of an array over a day. So to get 1GW during the day you need a 2GW solar array. Now your 4h batter is 4h x 2GW = 8GWh. Which means your 4h battery can provide 1GW for 8 hours a day combined with your 2GW array providing 1GW for another 8 hours, and less than 1GW for even longer.
Of course the grid doesn’t actually want X GW 24/7, demand is higher in the day and lower at night. On top of this 4h is sized based on maximum output, you need extra solar to cover cloudy days, but extra arrays also come with more grid storage. Thus at scale 4h battery backed solar arrays can roughly meet all of the grids energy needs.
Another factor is panels can be pointed slightly east or west if you want power earlier or later in the day. The tradeoff is less total output over the day, but with expensive storage and cheap PV it can be more profitable that way.
Storage to buy cheap energy and sell it back is also a completely separate enterprise. At the moment it looks like lithium iron phosphate batteries will work the best in the near future since they are the cheapest when taking into account the battery life time.
There are some promising techs out there, like Liquid Metal Battery:
My car has enough capacity that it could easily do it . . .
I wish we could get Tesla to allow their cars to be used for energy storage, like the new Ford F150. I'm concerned they are worried about competing with their Powerwall business.