There's little reason to build massive batteries at one spot, unless you are repurposing an only transmission line.
Instead, a good chunk of grid storage is getting deployed right at the generation site of solar (and some wind), which allows more efficient use of that transmission line.
Instead, we should be looking for large amounts of total install. However, this still won't happen much until it's actually needed by the grid, which starts to happen at much higher amounts of renewable generation than most states are using.
The tech is there, it's being deployed at massive scale where needed, and it's dropping in cost as fast or faster than predicted.
The tech is not here. The scale of grid storage required to fulfill just diurnal storage - let alone days or weeks to offset seasonal variation - is far beyond what batteries can provide. To put this in perspective, the US alone uses 12 TWh of electricity per day. The world uses 60 TWh per day. Both of these figures are going to increase, as poorer countries develop and want amenities like air conditioning. Also, as transportation and industrial processes are electrified. By comparison, global battery production is around 500 GWh per year. Yes, this will increase. But most of that production is going to electronics and EVs, not grid storage.
This is why proponents of a primarily wind + solar grid assume that hydrogen, ammonia, compressed air, giant concrete weights, or something else will make energy storage nearly free. Delivering the required storage scale with existing technologies isn't feasible, so people just assume that some other heretofore unproven technology will be orders of magnitude better.
The point of the various 100% solar, wind, battery projections that exist is that no new tech is needed.
Things won't be 100% Solar, Wind, Battery because other minor techs like nuclear, hydro, tidal, biomass or whatever already exist to some degree and can be part of the system. But current solar, wind and battery tech is enough, we just need to build it. The first 80% is the easy bit, with the greates payback, so there's no need to wait around.
Batteries cannot feasibly achieve the scale required to even out diurnal, let alone seasonal, fluctuations. The amount of batteries produced is nowhere near enough to satisfy demands for grid storage, and it'd massively set back electric vehicle adoption. Even as battery production ramps up, it's mostly going to go the EVs. Furthermore, electricity demand is going to go up too as people move from gas heating to electric heating and combustion vehicles are replaced with EVs.
"No new tech is needed" is a pointless statement if it can't reach the required scale. You might as well say "just build more dams". We don't need any more wind or solar. Just build dams everywhere.
This sort of argument does not pay attention to numeracy or the existing plans for battery production within the 2020s.
We can't build more dams because there really is a hard limit on the geographical sites. With batteries, we already have commitments for factories to build 1TWh/year within the US alone by 2030. Worldwide production will be several times that.
Average US electricity production is 500GW, at 8-10 hours that's only 4TWh. With batteries lasting 20 years, only need 200GWh/year of production to fill that diurnal need.
Batteries are cheap and scaling at a scale that we couldn't dream of scaling our construction capacity. Our limited construction capacity should be reserved for high speed rail, subways, and housing in urban centers.
Battery production will increase, yes, but so will electricity demand as transportation, heating, and industrial processes are electrified. Right now electricity use is only 37% of total energy use [1].
> Average US electricity production is 500GW, at 8-10 hours that's only 4TWh
Again, it's 12 hours for diurnal storage not 8 hours. More than 12 hours during the winter, actually. And diurnal storage isn't the only type of storage that's necessary. Factor in storage to even out seasonal fluctuations and you're looking at days maybe even weeks of energy storage. And again, 500 GW is going to turn into 1,300 GW as the rest of our energy use is electrified.
Batteries don't last 20 years, not even close. Diurnal storage is going to be cycled daily. A typical lithium ion cell lasts 300-500 charge cycles [2]. You can prolong this by limiting depth of discharge but this has the side effect of reducing the usable capacity. Let's be generous and assume 2,000 cycles that's only 5 and a half years.
200 GWh per year is still a massive amount of batteries. We're talking about over a third of global battery production to provide 8 hours of storage for just one country. And again, in reality we need more than 8 hours of storage and batteries don't last nearly as long as you claim.
There's a reason why plans for a primarily renewable grid assume that compressed air, hydrogen, or something else will account for the majority of storage: batteries aren't available in sufficient quantity, and deploying grid storage at any significant scale will severely reduce availability of batteries for EVs.
In a decade of following plans for 100% renewable grids, compressed air has never made it as a conceivable solution. Similarly hydrogen almost never gets on the list, and it's only there assuming some sort of currently non-existent tech advances far in excess of what is ever allowed to batteries. And yet, batteries make the list on all these plans!
As for your link, this is just flat out misinformation for grid batteries. Might apply to phones:
> How Many Cycles Can You Get Out Of A Lithium-Ion Battery?
>A Lithium-Ion battery's average life span is 2 to 3 years or 300 to 500 charge cycles, whichever comes first. As we put it, a charging cycle is a duration of utilization when the battery is fully charged, completely drained, and wholly recharged.
Industrial grid storage is rated to daily discharge for more than 10 years, with warranties typically around 12-14 years and expected life far afterward. 5000-10,000 cycles is more realistic.
Every five years, battery production capacity is growing 10x, prices drop dramatically. What is this limit? What is the reason it's suddenly going to stop? There's no answer other that I can see other than "I thought this 10 years ago so it's probably true too."
Hydrogen (or e-fuels) make it into plenty of plans. The important point is that it's for the last part of the journey to 100% RE, and makes little sense as long as natural gas is still being used. But for that last part, it's likely to be very useful.
> Industrial grid storage is rated to daily discharge for more than 10 years, with warranties typically around 12-14 years and expected life far afterward. 5000-10,000 cycles is more realistic.
I'd be very, very interested in these lithium ion batteries that have a life span of 10,000 cycles. The only way this would be achieved is with a very small depth of discharge, which severely reduces usable storage. I at least provided a link to back up my claims, yet you accuse me of spreading misinformation despite not doing anything at all to back up yours. Alternative chemistries like lithium iron phosphate achieve 3,000 - 5,000 cycles [1] at 80% depth of discharge. They last 5-10 years, not 14 [2]. But that's a new batter chemistry with smaller share of the battery market than typical lithium ion, and they also have smaller capacities than lithium ion and lower max power output.
Batteries are already being bottlenecked by input materials. Manufacturing accounts for only a quarter of a battery's cost [3]. Scaling out batteries is already becoming a problem of resource extraction. Even if the manufacturing cost is optimized to zero, the cost of inputs are still there.
The cost of a new car went from a quarter of a million dollars in 1900 to $25,000 in 1920. Would it be reasonable to observe that the cost of a car was falling by a quarter every two decades and predict that a new car would cost $6,250 in 1940, $1,500 in 1960, and $100 by 2000? Why would this scaling stop? What's the limit? Why didn't cars keep dropping in price?
But if you still need to support the complete gas, hydro.... infrastructure for those few months when wind/solar/battery is not enough, how cheap are those renewables then really?
Gas plants have their main cost in running them. So if they only fill the gaps, their cost is reduced too. But yes, they are not to be considered "profitable" but rather as part of the infrastructure like powerlines. But as solar and wind costs only a fraction of gas when running, this is a good trade off.
The numbers in your post do not march your claim of the text not being there. Scaling production of existing tech 10-100x, or even 1000x, will surely result in new discovery too, but it we only need a few orders of magnitude increase in production, that's proof that the tech is there.
Compare this to nuclear. Let's increase our production levels 100x. Where does that leave us, assuming that it was magically economically acceptable to electricity customers to pay higher prices than necessary. 15 years for 2.2GW is about 150MW/year. 100x would be 15GW/year. That's nowhere close to being where we need for a full energy transition in the US.
Nuclear, if it figures out its huge problems with construction, will be a small player to help with climate change. But in the year 2023, we know the big players: solar, wind, and batteries. There's no more time for anybody else to scale to catch up. Nobody else has a tech that can compete with such fast dropping costs. The numbers and pace of change are hard to compete with.
> Compare this to nuclear. Let's increase our production levels 100x.
Nuclear power already produces 10% of the world's electricity [1]. A 100x increase leaves the world with 10 times as much electricity as is needed, all coming from a decarbonized energy source.
Wind and solar are cheap because we don't currently have to even out their intermittency. Take away peaker plants and then intermittent sources become way more challenging. Solar produces energy in a sinusoidal pattern daily, requiring at least 12 hours of storage for truly non-intermittent solar plants. It also fluctuates over the course of the year due to weather and inclination of the Earth [2]. Wind power similarly sees fluctuations over the course of the year [3].
A 100x increase would get us back to what we were building half a century ago. Current nuclear production has been nonexistent, which is why this one reactor is such big news.
Instead, a good chunk of grid storage is getting deployed right at the generation site of solar (and some wind), which allows more efficient use of that transmission line.
Instead, we should be looking for large amounts of total install. However, this still won't happen much until it's actually needed by the grid, which starts to happen at much higher amounts of renewable generation than most states are using.
The tech is there, it's being deployed at massive scale where needed, and it's dropping in cost as fast or faster than predicted.