They claim 1m in revenue from their existing product which is:
"1.75 megawatts (MW) of peak power output; a 2.2 MW charge rating; and 10+ megawatt-hours (MWh) of storage capacity"
How does that equal 1m in revenue? You have to pay to charge your air-battery during low-cost hours and then discharge it during high cost hours to make the difference in revenue.
The issue is, you can buy 2 megawatts of power 24/7 at average datacenter energy pricing (~5c/kwh) for $2400/day or 876k/year. Who is paying more than that for less energy and only at certain hours?
Is there some energy company somewhere that sells power during the day at like 50c/kwh and 1c/kwh at night???
These kinds of utility-run storage facilities usually buy and sell on the wholesale spot market, which swings much more than retail prices do. Somewhat similar economics to gas "peaker" plants, which only fire up when supply is tight and prices spike, making it economical to run a higher-cost generator.
The Hydrostor plant is in Ontario. Here's an archive of hourly spot-market electricity prices there (in dollars per MWh): http://reports.ieso.ca/public/PriceHOEPPredispOR/. So far in 2022, prices have been as low as $0.00/MWh (during a few hours overnight on Jan 5) and as high as $230.87/MWh (during the morning of Jan 11). In 2021, hourly average prices even exceeded $1000/MWh ($1/kWh) twice, on March 28 and October 10.
I would need to see the math, because nothing adds up to even 1/10th of their claimed revenue. Best case, they bought 10mwh on jan 5th and sold it at the perfect time on jan 11th. This would give them 2k in revenue in 6 days.
Wheres the rest? And what about if you cant nostradamus predict the best times to buy and sell?
Sometimes utilities only need a few minutes of power to cover small gaps while ramping up/down powerplants, or if a cloud passes over a solar generation facility. Since these volumes are very small but critical, the effective price per MWH is extremely high.
When Tesla did their grid scale battery installation in Australia, this is how they made their money. Not via energy arbitrage, but with selling frequency response services. I have no idea if this is the same thing, but there are other ways to make money with energy storage as well.
Still seems extreme. IF my math is right, they would need to buy and sell 10MWh every day at a 27c/KWh markup. Given they they deliver at ~2.2 MW, This means that they would have to hit this markup for a 4 hour duration, every day.
They are likely operating on the wholesale markets, which are market priced. For the western grid there is a significantly wider price range of prices. On peak days in the summer, prices can go up above $1,000/mwh in the late afternoon for several days. During the late fall/early spring, there are negative prices in the late morning (like $20-$50/mwh) due to extremely high solar production and low demand, so utilities are literally paying you to use electricity. It's not inconceivable to get to $1 million a year on 1 mwh of storage.
A lot of people look at these and think "that doesn't seem like it'd be remotely worth it for something that size" but it's really hard to explain just how badly California has painted itself into a corner on distant generation, huge variance between excess renewable capacity and peak load, and transmission congestion. The numbers don't make sense to people.
Pretty much all the solar plants we're putting in now have combined battery storage systems either for voltage support or "peaker" sales opportunities. In some places (like literal islands) it's about capacity smoothing or resilience, but the driving force behind all the investment is the US western interconnect with arbitrage opportunism starting to make real money and looking to make much, much more over the next decade.
If government-sachs is betting on this, they've probably got an even more pessimistic take on the CAISO market than I do.
Just read the constant stream of news about it? They keep shutting down power plants in the name of the environment, but just shift the generation outside of the state. It's a numbers game that's meaningless as long as they still consume power, but it lets them virtue signal all over the place to people who don't pay attention to details - which sadly is most of us. How many times when we are give stats do we really wonder about the context around them?
Im familiar with energy markets. Im also familiar with the fact that any serious seller or buyer of energy locks in yearly pricing contracts and is unaffected by these swings. Any serious datacenter or industrial buyer is not relying on spot pricing. These energy markets also dont exist in every state.
If you are familiar with energy markets then by all means tell us the percentage breakdown of electricity consumed that is covered by long term contracts vs the spot market.
There are times where the rooftop solar and base load generation for a utility is greater than the total electricity demand. Demand and supply for the grid have to be balanced, or else you will end up tripping plants and/or damaging equipment. Since they can't "turn off" rooftop and most grid scale solar generation, and they can't really turn off and on their base load plants throughout the day, their best option is to try to sell the excess electricity on the wholesale market. The problem is that the other utilities are in the same situation, and they need to offload electricity. Since no one wants the electricity, wholesale prices will actually go negative. Basically a utility will pay another utility to turn off a base load plant and take the excess electricity from them.
This is why grid scale energy storage is such a hyped technology - if you can store that energy when supply is way too high and dispatch it when renewable generation is low (no wind/solar), you can maintain high renewable percentage in the generation mix. Without it, you can only really go up to a certain percentage of renewables.
Buy low and sell high. They can find surplus power and buy it cheap, then wait until there are unmet demands and sell high.
These crazy things called computers make creating and automating markets like that pretty easy :)
It's win win - the people with surplus power can at least get some money for energy that would otherwise likely go to waste (large base load power plants can't ramp up and down on a dime) and on the flip side if you have an energy provider that only occasionally needs capacity in excess of their inherent generation even though the immediate cost is higher for spot power, it can still be far cheaper in the long term than building more base load generation (often fossil fuel because renewable still isn't a viable replacement for base loads). And running base load generation under capacity is wasteful for emissions and overall efficiency.
Once we start to crack the whole energy storage thing renewables will be A LOT more valuable since we will be able to start to use them for base load. Storage allows you to create reliable energy delivery that's predictable - that's essential if we ever want to eject fossil fuels from the grid. Or we could stop being childish about nuclear power - frankly I think we can solve the storage problem long before we can convince people to be rational about nuclear :p
It probably seems crazy for the uninitiated, but there are a lot of efficiencies in markets that occur naturally over time. You see distortions in the market from CA's stupid and politically motivated choices - what seems crazy to you is the market adapting to the model that CA is forcing. Is it nuts? Yup. Is it necessary? Nope - these wounds are entirely self inflicted. Pretending the world is different than it actually is doesn't solve anything and the people of CA are paying a hell of a lot more for electricity because they are trying to live in a pipe dream for where we are today.
IF 10MWh is their daily cycle, they are selling 3,650,000 kWh/year.
To hit $1M per year, they would need to be selling ~27c/KWh above their buying price.
I get that you can have a big spread if you are being paid to take power, and paid again to sell it.
I think that’s it basically. New England consumers pay ~$.22/kWh. We’re going to have offshore wind energy in the next decade that might conceivably have surplus power quite a lot. It’s not beyond the realm of possibility that these plants could have a deal with the power plants to take extra power for free so long as they have guarantees on filling the expected gaps or meeting expected spikes in usage. Everyone wins.
They claim 1m in revenue from their existing product which is: "1.75 megawatts (MW) of peak power output; a 2.2 MW charge rating; and 10+ megawatt-hours (MWh) of storage capacity"
How does that equal 1m in revenue? You have to pay to charge your air-battery during low-cost hours and then discharge it during high cost hours to make the difference in revenue.
The issue is, you can buy 2 megawatts of power 24/7 at average datacenter energy pricing (~5c/kwh) for $2400/day or 876k/year. Who is paying more than that for less energy and only at certain hours?
Is there some energy company somewhere that sells power during the day at like 50c/kwh and 1c/kwh at night???