Whenever grid batteries are discussed, it seems that one of their problems is that they’re only good for about four hours. And as a result, the typical installation has 4 hours worth of total energy when discharged at the maximum rate. For example, 50MW/200MWh and 100MW/400MWh batteries seem to be very common sizes of grid batteries.
So where did this 4 hour battery rule come from? Do grid batteries turn into a pumpkin if discharged for more than 4 hours? Is there some physical reason they couldn’t make 100MW/600MWh batteries and have a 6 hour battery? Or is it that someone once did an optimization study and concluded that 4 hours was the best use of batteries and that became battery dogma?
That does not sound like a “rule” so much as a technological limitation of the vanadium batteries or whatever it is they are using. If you had higher-tech batteries, the maximum discharge rate, specific energy, etc., would be different.
There’s nothing that limits batteries to 4 hours. You can run them at half the power capacity for 8 hours, or a quarter for 16 hours, or whatever you like.
Four is a common number because batteries can very comfortably discharge at that rate. And the support electronics are not too expensive.
Whats sorta funny is that 2-hour packs for the same capacity are more expensive. Seems counterintuitive at first… but it just means that they had the same collection of cells but had to add more power electronics to achieve a higher peak power. Go ahead, check out Tesla’s configurator to see for yourself: https://www.tesla.com/megapack/design
Their 2-hour pack actually has slightly less energy, but costs more… because the peak power is 9.6 MW instead of 4.9 MW. Those extra power electronics cost money.
But no one seems to be doing that. They all seem to start with a 1:4 ratio of energy to capacity. When they expand (as in the article I linked to above about Moss Landing), they also increase the energy to keep the 1:4 ratio.
There are a few different ways of looking at this. One is just: given that you have a pile of battery cells, with power electronics designed to discharge them completely in 4 hours, how much do you save by taking away electronics so that they take 8 hours, 16 hours, etc. to discharge instead?
Given that the cells are already the dominant cost, and that the structure and cooling and fire suppression and other things are also significant, there are probably diminishing returns to stripping more electronics out of the unit. And it means the pack can no longer handle certain problems.
From my link above, you can see that doubling the power from 4 hours to 2 hours costs about 7% of the pack cost. Therefore one might expect that halving it from 4 to 8 would save 3.5%. And reducing it further to, say, 1000 hours would only asymptotically reach 7%. It’s not a huge savings.
But one of the biggest uses for grid batteries at the moment is for flattening the duck curve–that is, timeshifting peak solar generation at noon to peak power use in late afternoon (4-5 pm). A 4-hour battery is a good fit for that.
Other uses are for grid stability, handling transients like generators turning on and off, and smoothing the load from renewables so that transmission lines can be used at peak efficiency. These uses may only need a 2-hour battery or less.
I should have added in the UK and Ireland most grid batteries are between 30 min and 2 hour. The shorter duration are good for when a thermal generation plant trips and it takes time for another generator to be fired up. The 2 hr are OK for the tea time though 4 hours are now coming online.
There are of course differences in the grids. In the UK and Ireland most renewable generation is wind and it is typically windy or calm for a few days at a time, therefore demand net of renewables is usually lower overnight that in the middle of the day.
Sometimes, when an OP asks a question that makes no sense because it refers to something or someone I’m unfamiliar with, it just means I’m not in the intended audience for the question or the thread, and I can safely ignore it; but sometimes it means that the OP is poorly worded and has failed to provide needed context. I thought this thread was the first kind, but it’s good to have that confirmed.