Used car batteries are much, much cheaper. So as more BEVs come on road, that's obviously one way forward. Also $100/kWh is today. However, given how rapidly costs have declined over the past three decades, nothing suggests that this is the final number.
Finally, these are just Li-ion prices. Na-ion and Fe-ion are also very close to deployment now. Both of them cannot match the energy density of Li-ion, but that's fine. Grid storage depends on cost, not energy density anyway
Storage batteries are going to fill gaps in generation swings, not power gigawatt scale grids for more than a few minutes. This is the crux of my point and i stated it as such.
A lot of smaller batteries is effectively equivalent to a single larger battery and that does not get around the problem from above - they simply cannot store enough energy for more than short periods. A handful of large hydro reservoirs store more potential energy than all chemical batteries in the world combined.
Meanwhile, extreme weather events are getting more common. Texas - for example - had a nearly 10,000 MW shortfall between their forecast and the actual, served demand. Is a battery every going to supply capacities like that for hours on end? No. During that cold snap, some wind turbines froze, some didn't see enough wind and much of the solar wasn't generating at full tilt. What then? Do we shrug our shoulders and say "no heat, too bad"?
Chemical storage batteries cannot power large grids for extended periods and that is the fundamental problem being ignored, not their cost.
The problem with HN is that people do not think outside their silicon valley and techbro bubble. Another comment here suggested that domestic loads can be served with a battery. Sure, if you live in California and don't need much heat. This is the problem - when someone points out that what is proposed is not viable everywhere and that a mix of solutions is needed, it gets downvoted because it doesn't follow the "renewable energy is the be all and end all" argument.
Define short periods. Also gas lines froze more than turbines did BTW at Texas, just FYI. Very few conventional sources can also backstop 10GW of demand mismatch. Which is why having a national grid is so much better, power from Arizona can be used to top up shortfalls in Texas and vice versa.
> Chemical storage batteries cannot power large grids for extended periods and that is the fundamental problem being ignored, not their cost.
You are being downvoted because this is just an opinion, not backed by any facts on the ground. Most solar installs today itself come with 4 hours of battery backup. Also, several posters and I have given cost figures on how much multi-day battery backups would cost even at current prices. So your insistence that batteries only last for a few minutes is getting downvoted, and calling people techbros isn't helping either.
I think there are two tracts to this discussion: industrial and residential energy.
Solar + Storage 100% can replace residential power usage. Ask anyone with a home battery and solar. Density is obviously a factor, but seems straight forward to solve.
Industrial energy usage, I think it’s seperate, but also is very moveable. Ie we can change how industry works to surge production at peak solar and wind, and then reduce when that power is not available.
Plus battery storage still works here (Not just lithium either, hydro, thermal, compressed a air, gravity store).
Where exactly do you see “unpredictable” energy problems that storage or shifting doesn’t solve?
> The problem with battery energy storage is not cost but scale.
I'd posit that it's mostly cost, otherwise it would be a solved problem now.
Certainly we can scale - even if it means putting low-cost batteries behind the meter (that solves, for a very modest sum, 90% of power consumers' problems). That's a horizontal scaling, obviously.
For the consumer types that need a lot of energy, then we need to come up with more imaginative solutions for their requirements. These types of turbines are almost definitely part of that.
Aside - complaining about downvotes is bad form. My feeling is your myriad comments on this article exhibit vehemence, and imply everyone else is ill-informed. It's possible we all have something to learn.
People here are indeed ill-informed. Silicon valley has a recent obsession with battery tech and seems to believe that batteries can scale like software. Spoiler - this is not the case.
As i said originally, the technical problem is scale. Even if cost does fall, no battery can power a gigawatt scale grid for more than a few minutes. Yes - minutes. During an extreme weather event, what will happen? There is a lot to learn but the solution - and this the part that people don't like hearing - is a mix of technologies. Batteries are a thin slice of the solution. We need more hydro generation, more nuclear, more wind and more solar PV. It is not a binary choice as many here seem to believe.
Overbuilding solar and wind have been long suggested. But your insistence that batteries will only provide power for a few minutes is honestly not backed up by any data.
> the solution ... is a mix of technologies
Certainly.
> Even if cost does fall, no battery can power a gigawatt scale grid for more than a few minutes. Yes - minutes. During an extreme weather event, what will happen?
In an emergency the primary focus isn't, for example, on powering offices, powering every single thing you can. The focus should be on powering homes for a relatively short duration at reduced consumption. The goal is survival, not keeping everything on the grid powered up 24/7. Two Powerwalls can power a home for two to three days at reduced energy consumption (rationing). There's no reason we can't get to that type of outcome as routine in the next 10-20 years, given the cost of homes today (nearing $400,000 median sale price) and the continual decline in battery costs (and that's assuming no great battery breakthroughs). There's no reason all new homes built in the US shouldn't come with the equivalent of at least three Powerwalls of energy storage 20 years out.
It's possible that the phrase 'behind the meter' isn't universally understood - what I meant by that is that we can put storage systems (Lithium, Flow, etc) on the customer premises - on their side of the electricity meter.
This gives the power consumer an energy cache, effectively, with 2-3 days of self-reliance.
That's why I suggested that your insistence that we need one stonking large battery sitting somewhere is misguided - we can distribute, today, relatively cheaply (compared to the cost of a house build / maintenance) short-term independence from the grid supply.
Yeah. And for big power banks, there are several options beyond Li-ion. There are sodium, iron and magnesium chemistries for one.
Compressed air storage is another option, that is quite cheap and scales extremely well. The one being built by Hydrostor now in California is 4GWh, for example [0].
Refs:
[0] https://newatlas.com/energy/hydrostor-compressed-air-energy-storage/ https://newatlas.com/energy/hydrostor-compressed-air-energy-...
One thing is pretty sure - we need to overbuild solar and wind. The total annual energy consumption in the US is around 15 PWh. Notably, the per capita energy consumption has essentially remained flat for four decades now.
This energy consumption thus translates to an averaged power requirement of 4TW. Notably, while peaks are local, however nationally, extremes are not that higher than average. The average capacity factor is a third for renewables, so 12TW of nameplate renewable capacity is enough for the vast majority of the time - to power the entire nation's complete energy requirements through renewables.
Let's go with a 50-25-25 mix of solar-onshore-offshore wind. Thus solar nameplate needs to be 6TW, and both onshore and offshore wind needs to have 3TW nameplate capacity installed each. Solar PV generates 10W per square foot. That's 22,000 square miles. A lot of that can come from rooftop solar. Onshore wind turbines are 2.5MW each, which translates to 1.2 million turbines, while with the offshore turbine designs that this article shows - this is 200k turbines. The cost of such turbines (onshore+offshore) with installation would come to 9 trillion USD. Solar would cost a similar amount. Battery costs for storing 96TWh (a day's worth of energy consumption - averaged) would be 10 trillion dollars more.
Thus, building the nameplate capacity for the entire US to be powered by renewables is approximately 30 trillion dollars, where the GND price tag originated. However, all of these are at current solar/wind/storage prices. Each of those three is falling, and if the market becomes this big, economies of scale will drive costs down further.
To put some numbers to this. France produced 537.7 TWh of electricity in 2020. That's 537,700,000,000 kWh. Let's say you want battery capacity that can cover this for one day - divide it by 365 - 1,473,000,000 kWh. At $100 per kWh that would cost $147 billion.
That's actually not that outrageous. In reality you would probably want more battery capacity than that, but at the same time you'd probably have other forms of power generation too. Those other forms of power generation probably aren't going to be off when wind power isn't being produced, which means that they don't need to be covered by the battery capacity.
France's budget in 2021 was $755 billion.
Edit: I forgot about batteries wearing out. If they have to replace the batteries once a year then this would still be prohibitively expensive.
Batteries easily last a decade now. LFP chemistries easily will do 1000+ cycles. Also 24 hours is a huge amount - you actually don't need more than that, especially if you have offshore wind in your mix, as they blow pretty much all the time.
1000 cycles / 365 days =~ 3 years, not 10?
I heard a solar project is over-provisioning the battery and planning to replace it every 5 years as they must be able to deliver some amount of MWh 4 years down the road.
And when a storm hits? No electricity?
There are currently 32 millions passenger cars in France, so for 1,473,000,000 kWh by 32 millions that's ... 46 kWh per car.
Most cars have more than that as battery so when all car are electrified, put together and filled up they can power France for a day without any other generation source.
France peak demand is about 100 GW so that's 3.2 kW per car.
It's a tiny amount relative to the power of the motor.
It can flow through a nearly normal plug (16A 230V).
And this is for the worst day of the year.