* We don't need the capability to store all the electricity generated. * You are comparing a number that doesn't accumulate (generation) and one that does (capacity sold in 2021). Batteries don't last one year, and I expect more capacity will be sold in 2022. The linked article notes a growth rate of 113% from 2020-2021. * Most importantly, there are many ways to shape demand and storage beyond EV-style batteries. E.g. my electric water heater is a "battery," my utility (Green Mountain Power) can turn it off for a few hours during peak loads.
We don't need the capability to store all the electricity generated.
Never said we would, but it's a useful number for comparison. We WILL need the capability to store a very large fraction of it if we're reliant upon energy sources with capacity factors of 40% and lower and largely simultaneous outages over large areas.
You are comparing a number that doesn't accumulate (generation) and one that does (capacity sold in 2021). Batteries don't last one year, and I expect more capacity will be sold in 2022. The linked article notes a growth rate of 113% from 2020-2021.
Batteries wear out too; useful lifespan is on the order of ten years, as my car is demonstrating to me. But the most interesting thing about the 36 minute figure is that running the world's electric grids on "renewables" would require something like three orders of magnitude as much battery capacity as was sold last year.
Most importantly, there are many ways to shape demand and storage beyond EV-style batteries. E.g. my electric water heater is a "battery," my utility (Green Mountain Power) can turn it off for a few hours during peak loads.
I lived over a year in a duplex where I'd switch on the water heater for a few minutes every other morning, and still have warm enough water to shower the next day. Then I lived for 3 years in a house where my utility had control of my air conditioner. But there is a crucial floor below which you cannot dip without major damage occurring. Municipal water systems and sewage lift pumps are two things which spell disaster if they shut down. They've got to have power 24/7/365, and if you don't have enough batteries to carry you over, you need something else.
I find the approach taken by the Natrium system by Terrapower and GEH to be very interesting. Instead of storing electricity in batteries, they store heat in tanks of molten salts. Whatever happens at Kemmerer, it is going to break new ground.
Long-duration energy storage (LDES) is a potential solution to intermittency in renewable energy generation. In this study we have evaluated the role of LDES in decarbonized electricity systems and identified the cost and efficiency performance necessary for LDES to substantially reduce electricity costs and displace firm low-carbon generation. Our findings show that energy storage capacity cost and discharge efficiency are the most important performance parameters. Charge/discharge capacity cost and charge efficiency play secondary roles. Energy capacity costs must be ≤US$20 kWh–1 to reduce electricity costs by ≥10%. With current electricity demand profiles, energy capacity costs must be ≤US$1 kWh–1 to fully displace all modelled firm low-carbon generation technologies. Electrification of end uses in a northern latitude context makes full displacement of firm generation more challenging and requires performance combinations unlikely to be feasible with known LDES technologies. Finally, LDES systems with the greatest impact on electricity cost and firm generation have storage durations exceeding 100 h.
Battery storage for bulk energy generation is barely a drop in the bucket on a global energy level. Anyone who understand maths and doesn't have axes to grind against nuclear power or other big scary energy sources, knows this. I guess very few people know maths, or don't have axes to grind.
I wonder if batteries can serve a more useful function by giving old EV batteries a second life as grid power banks. If everyone drives an electric vehicle, eventually quite a large store of used batteries would build up. They'd have insufficient range for an EV, but could serve a second life as grid power banks where they'd be used down to quite low levels of capacity - basically till they fail completely. Here the cost per kWh doesn't matter so much as the batteries would be written off already, and are simply given another life before recycling. Any thoughts on this EP?
Thoughts? It's been pretty much taken for granted by everyone, myself included, that large-scale scrappage of EV batteries would involve at least some second-life applications before outright recycling. Grid frequency control, peak shaving, riding through dropouts... the list goes on and on.
Well it would be interesting to quantify this. If the amount of scrap battery rivals the battery capacity in circulation in EVs it could be a big deal - we could easily be looking at over a billion EVs the next few decades, and over 10 billion kWh of scrap battery capabilities. Obviously, I'm donning my devil's advocates' hat here - I'm a nuclear guy.
I think I calculated this once, but it wouldn't hurt to revisit the numbers.
There are approximately 250 million LDVs in the United States. If all of them were PHEVs with 15 kWh batteries, they would have a combined total of 3750 million kWh or 3.75 TWh of energy storage, not enough to power US average grid load (which does not include even 10 million plug-in vehicles IIUC) for even 10 hours. The actual capacity would be less because the "second life" batteries would be in a degraded state. Useful lifespan is uncertain, and would probably be competing against recycling to reclaim things like lithium and cobalt.
If all 250 million LDVs had Tesla-class 100 kWh batteries, they'd total 25 TWh of storage between them. This would support the US average grid load of 477 GW for a little over 2 days if starting from full charge. Second-life batteries would store less than this.
There are things that these second-life batteries could be very useful for, such as buffering grid transients. Suppose your nuke plant had 15 minutes of storage on-site, kept at half-charge on average. You could ride out transmission-line trips without having to SCRAM the reactor for some minutes, allowing enough time to clear fault conditions and re-close breakers. Maybe you could time-shift overnight generation to the afternoon peak. But buffering for the many days required to handle the on-again, off-again nature of wind and solar generation? Not happening with any currently-available technology.
The main impending energy storage problem is seasonal storage for northern latitudes in the winter. Geothermal for houses/buildings plus better building insulation would take a big bite out of this if done right. Doing bespoke geothermal installations for individual suburban houses is inefficient. Community scale installations are wanted. Scale delivers discounts. Another option is huge molten salt heat storage--I think the physics favors a small number of huge storage operations over a large number of small ones.