Deals hatched over Twitter generally don't work out that well. Elon Musk might have found an exception.
The Tesla Inc. chief executive officer promised Atlassian Corp. co-founder Mike Cannon-Brookes last week that he could install a 100-megawatt-hour energy-storage plant within 100 days at a battery-pack price of $250 a kilowatt hour.
That would help balance out the unstable local grid in South Australia state, where spot wholesale electricity prices over the past two months have risen as high as $13,440 a megawatt hour and dropped as low as minus $146.29/MWh.
If that sounds like science fiction, blame the speed at which the lithium-ion battery market is developing.
The logistical challenge has already been more or less solved. Tesla last year installed an 80 MWh system in California in just over three months. A 100 MWh system in 100 days is more of a stretch, but not by much.
What about that $250/kWh price? Anyone with a vague sense that large-scale lithium-ion battery packs were around twice that price recently would be right. In 2013, they were selling for about $599/kWh, according to Bloomberg New Energy Finance. Thanks to the dramatic reductions in battery costs, that dropped to $273/kWh last year -- and BNEF calculates Tesla's Gigafactory could now make the packs at about $240/kWh.
Given the way the industry is progressing, Musk has little to lose from pricing his deal keenly. He expects the energy-storage market "to grow maybe twice the rate of the automotive business," according to the transcript of a Feb. 22 investor call, but faces a problem from the fact that his costs are, if anything, high relative to Korean and Chinese competitors.
With many investors still skeptical that utility-scale battery storage is even viable, getting Tesla's name on some high-profile demonstration projects like those in California and South Australia will be a big help in winning new customers. Even if Musk were to forsake his profit margin to win the contract, Tesla shareholders have shown themselves to be comfortable with selling cars below cost in the pursuit of long-term market share.
Could such a system make money for the asset owner, though? In South Australia, the answer right now appears to be yes.
Energy-storage systems profit by arbitraging volatility in the electricity market. As long as the gaps between the trough prices at which they buy and the peak prices at which they sell are big enough and the cost of the storage system itself is low enough, it's a fairly straightforward trade: Buy low, sell high, and make sure you have a good handle on how the daily cycle is going to change.
The volatility of the South Australian electricity market makes it peculiarly well suited to this business. Let's assume that Tesla's utility-scale batteries can charge up and discharge about 5,000 times before being replaced, and estimate that the system is only about 85 percent efficient -- in other words, that for every 100 kWh put in, you only get 85 kWh back out.
Let's also assume that Tesla's battery pack price of $250/kWh is about half the total system cost once you include the expenditure on building a plant and connecting it to the grid -- pretty much in line with the $400/kWh to $600/kWh quoted by Tesla's vice-president Lyndon Rive last week, in the Australian Financial Review article that sparked Musk's Twitter discussion.
A back-of-the-envelope calculation suggests operating such a system would cost about $117/MWh, or A$155/MWh, on top of the amount it pays for electricity. So any day on which peak prices are at least A$155/MWh above the trough, it should be able to make money. So far this year, that's happened on 31 out of 71 days, with the median daily range at A$136/MWh.
The long-term problem for battery-storage systems isn't the cost of equipment, but the tendency of all arbitrage trades to cannibalize the volatility from which they seek to profit. South Australia offers an attractive opportunity for Tesla because of the extreme variance in its electricity prices -- but if the system gets installed, there'll be an extra 100 MWh of demand at times when prices are low, and an extra 100 MWh of supply at times when they're currently high.
That will represent a net benefit for the state's electricity consumers, but it will make the economics a lot more challenging. Musk is right to be confident that he can install a battery-storage system quickly, but that's only half the challenge. In the electricity trade, as in any other market, there's no such thing as a free lunch.
This column does not necessarily reflect the opinion of Bloomberg LP and its owners.
To contact the author of this story:
David Fickling in Sydney at firstname.lastname@example.org
To contact the editor responsible for this story:
Matthew Brooker at email@example.com