Unless I'm misunderstanding you, you're not correct. You seem to be saying that the "cost of storage" numbers already include averaging the storage cost between storage and not storage, but I can't get anything like that to work our to reasonably plausible numbers. Check page 8 of the PDF of their very slightly fuller explanation: https://www.lazard.com/media/451418/lazards-levelized-cost-o...
The project is cheaper than wholesale arbitrage because they don't have to pay wholesale prices to charge the battery, inverters are used for multiple purposes, and most PV solar designs already throw away some electricity, so that's free.
Also on that page, doing some napkin math on the project lifetime MWh of 1,260,000, and the purchase price of $600-$1000/kWh is higher than most grid batteries deployed these days. These are very conservative numbers!
Lazard's numbers are difficult to compare to most actual storage payment contracts, because every one I've seen is structured around paying for capacity, in kW-months, rather than straight payments for every MWh of energy dumped on the grid. But the capacity payments end up working out to numbers that are on the low end of Lazard's cost estimates when translated.
I'm sorry, it was a mistake. You're not counting it twice, it is indeed a figure just for storage, with the assumption of daily cycling.
If we use the ratio of the plant (i.e. 200 MWh storage for 720 MWh generation at a 30% capacity factor), that would leave us with (81-140 * 0.27) + 29-38 for the generation, for a total cost of $51-76/MWh.
But that's still based around the idea of cycling the storage daily. Solar performs poorly on cloudy days, so you can either overbuild it significantly, build days of storage instead of hours, or keep the old power plants around as backup. Either of the three is going to bump up that cost significantly.
It blows coal out of the water if you don't have to worry about days you're not producing power, but you can't build a grid around that. And coal is mostly being eaten by natural gas for price reasons anyway.
Did you factor in a discount rate with your back of the napkin math? It might explain why the purchase price looks so high.
The project is cheaper than wholesale arbitrage because they don't have to pay wholesale prices to charge the battery, inverters are used for multiple purposes, and most PV solar designs already throw away some electricity, so that's free.
Also on that page, doing some napkin math on the project lifetime MWh of 1,260,000, and the purchase price of $600-$1000/kWh is higher than most grid batteries deployed these days. These are very conservative numbers!
Lazard's numbers are difficult to compare to most actual storage payment contracts, because every one I've seen is structured around paying for capacity, in kW-months, rather than straight payments for every MWh of energy dumped on the grid. But the capacity payments end up working out to numbers that are on the low end of Lazard's cost estimates when translated.