Life for a lot of people these days centres around power and money. But if you’re in the power business… well, it’s ALL about the money. Geoff Russell weighs in again on the never-ending crisis that is South Australia’s electricity generation.
Following my recent New Matilda piece on South Australia’s short hot spell, I had a Twitter exchange with somebody accusing me of telling lies about batteries. Their name isn’t relevant, but their claims are, because they are incredibly common:
“It [Hornsdale] is being expanded for more energy supply and others are being built! You will find batteries specifically for supply in other places – but do your best to lie by claiming these things do not exist!”
Note the lack of any list of examples of big energy supply batteries.
I replied that the biggest battery I knew about (in operation) was the 250MW Gateway battery in California. The link I just gave is to an article in Popular Mechanics magazine about this battery. Astonishingly, the author clearly doesn’t know the difference between energy and power. She writes:
“The facility stores up to 230 megawatts of power with plans to expand up to 250, both dwarfing and setting a new challenge for similar Tesla facilities in rural Australia and others around the world.”
You don’t store power, you store energy. It’s a critical distinction. Energy is power multiplied by time. Speaking of which, sorry, but it’s time for both a spelling and an energy lesson. File it under basic scientific literacy.
Watts is a measure of power and 5MW says you can supply five million of them; the ‘M’ is mega or million, and the ‘W’ is watts. There are rules about the capitalisation of these things. It’s just like spelling, except that small mistakes can have much bigger consequences; like the difference between mW and MW. 5mW would be 5 milli-watts… a million times smaller than 5MW!
Five mega watt-hours (5MWh) is a measure of energy; the h on the end makes all the difference. 5MWh is the amount of energy you’d need in a battery to supply 5 mega-watts of power for one hour.
If you had 5MWh of energy, you could, alternatively, design your battery to supply 5kWh – that is, 5 kilowatts for a thousand hours. The ‘k’ stands for thousand just as ‘M’ stands for million (mega).
There’s a very big difference between a battery that can supply 5MW for 30 seconds and one that can supply 5MW for 30 hours. Needless to say, what’s the difference between 230MWh and 230MW? It should be obvious now.
So a battery is normally described by at least two numbers; the maximum power output and the maximum energy it can store.
A 25MW/200MWs (note the s, for seconds) battery can supply 25MW for 8 seconds. That might be perfect for keeping something running long enough to start up a generator when an electricity grid collapses.
A 25MW/200MWh battery can supply 25MW of power for 8 hours. Enough to run somewhere like Broken Hill (population 17,000) overnight.
This isn’t rocket science, and if you are writing about electricity, you should at least know the basics. But over the past decade or so, I’ve seen a considerable number of (almost always) 100%-renewable advocates using a single number to describe batteries, and then soon after demonstrating that they don’t have a clue what they are writing about.
A full description of a battery’s capabilities would require more than two numbers, of course, but two is the bare minimum.
How many big batteries are there?
The functions of a battery depend mainly on the size of the grid it is attached to, if any. Some big batteries aren’t attached to a grid, but have a more specialised purpose. For example, you may have a data centre or hospital that you want to keep running for a few hours in the event of a grid failure. I that case, a big Lithium-ion battery might be just the thing. A diesel generator might be cheaper, but nobody gets bragging rights these days for a cheap and dirty backup solution.
The Global Energy Storage Database (GES), maintained by Sandia Labs in the US, contains data on nearly 1,700 significant storage projects around the world, including all kinds of promising (and not-so-promising) new or struggling or old battery technologies. If you select all the projects involving “Lithium” you’ll find just over 600, each using one of the following technologies; Lithium Ion Titanate, Lithium Iron Phosphate, Lithium Manganese Oxide, Lithium Nickel Cobalt Aluminium, Lithium Nickel Manganese Cobalt, Lithium Polymer… and just plain old Lithium-ion.
Why so many different Lithium based technologies? They each have differing characteristics. Most have far more of every other thing in the name than Lithium. Some are safer than Lithium-ion (Li-ion ) for specialist uses; some can be recharged more often; some are cheaper (but probably heavier).
While Li-ion batteries are incredibly safe, they can burst into flames at the wrong times, as distinct from those times when you really want a flaming battery. The US Federal Aviation Administration maintains a list of 300 air/airport fires caused by Li-ion batteries. And when the battery in an e-cigarette fails, the results can be both spectacular and deadly; think flame thrower in your mouth.
About 10 percent of the Lithium battery projects on the GES database don’t have an energy rating at all, but of those that do, half are below 400kWh, which is small. Many of these look like one-hour batteries – i.e. they can output their maximum power for one hour.
So it’s clear that none of these batteries provide grid-level storage sufficient to do something like keep even a modest-sized grid up overnight, when most of its wind and solar power isn’t available. And, if you check, you can see that the biggest of the batteries are on big grids; so supplying energy is clearly not their function.
Here’s a plot of all the Lithium batteries on the database, regardless of technology sub-type:
As you can see, almost all of the fleet are tiny. The few in the Hornsdale class are outliers (towards the top right of the graph). The 250MW Gateway battery isn’t on the database. The 300MW dot represents the Vistra Moss Project, a 300MW/1200MWh battery (300MW for 4 hours). The Vistra dot and the 450MW dot, along with the small cluster vertically below them are all projects which haven’t provided an operational date. This is the most comprehensive database on the planet for such projects, but no database is perfect.
The graph makes it clear how small most “big” batteries are. The Gateway and Vistra Moss batteries are both on the California grid; which generates about twice as much energy annually as the NEM (Australia’s biggest grid, including SA, VIC, QLD, NSW, and TAS). So both batteries have a maximum power rating of 550MW in a grid with a maximum power demand around the 50,000MW range. So together, they can supply not much more than 1 percent of demand at its max.
The bottom line? There are currently no batteries anywhere on the planet that can provide backup for wind and solar during still nights on even a small section of the grid; despite popular beliefs to the contrary.
Finkel’s grand plan
Remember back in 2017 when the then Chief Scientist Alan Finkel released his report into the Electricity Market? His prediction of the growth of renewables (wind and solar) and the decline of baseload implied a massive growth in batteries.
So how many hours of backup do you need for a gigawatt of wind and/or solar power? I’d suggest you need in the region of 4 to 16 hours to get through a still night without any baseload to lean on. In which case, over the next 30 years we’ll need to build rather a lot of batteries for the 40GW of wind and solar that Finkel anticipates.
Assuming there is still 10GW of fossil fuelled (or perhaps hydro) backup on the system in 2050, you’d need to build between 4 and 16 times 30 GWh divided by 30 years. Which is between 4,000 and 16,000 MWh of batteries each year… just in Australia. Note that I’m ignoring the power requirements for now; let’s just deal with energy.
And if you want zero carbon electricity by 2050, then you may not even have 10GW of backup; other than from hydro perhaps. Snowy 2.0 is only 2,000 MW; so even though it can store plenty of energy, it can only deliver it relatively slowly.
The GES database allows us to look at the current rate of global grid battery roll outs every year.
We can see that it’s only ever been over 1,000MWh per year, and then only once. And remember, this is a global figure, not a national one. So if other countries tried to follow a wind+solar+battery plan, then Australia’s required rollout would only be a couple of percent of what would actually be required.
Why is nobody building big batteries?
It’s obvious that people know how to connect up any number of batteries into pretty well any sized plant. So why has nobody done it?
Recall that batteries don’t make electricity. So normally, you have to buy the electricity to fill the battery. But if you own an electricity generator of some sort, be it the solar panel on your roof or a wind farm, then you can charge a battery for free when the grid is overloaded. Even so, you’re then stuck with electricity you can’t sell.
Remember the California grid, and those Gateway and Vistras batteries? There is so much wind and solar on that grid that they throw out 10s of gigawatt-hours every month. In one month of 2019 they dumped nearly 225 gigawatt-hours. So if you have a battery, you might be able to save some of this and use it later when there is a shortage. Some people may remember the Enron debacle; they deliberately engineered shortages to profiteer in this manner.
You can’t make money out of selling electricity from batteries at a cost-recovery price.
Let’s demonstrate this using Adelaide Hornsdale battery. OpenNem provides data that is easy to analyse. I saved about 280 days worth of data in 2020; some weeks I just forgot about it! During that period, the Hornsdale battery average about 150 MWh of electricity per day. Assuming that this $160m dollars battery lasts for 5 years and ignoring maintenance costs, then each kWh of electricity cost about 60 cents; assuming that charging was cost free. The wholesale electricty prices in SA are generally between 3 and 4 cents per kWh, so this battery wouldn’t be profitable if it was bought to supply energy; which of course it wasn’t. It’s there to help stabilise an increasingly unstable grid.
The story of the slo-mo train wreck that is the SA grid will have to wait for a separate article, but looking at the income of the Hornsdale battery is interesting.
Based on my 280-day sample, the battery made about $3 million from supplying energy in 2020. But RenewEconomy reports that the original $90 million battery price (remember, it has been upgraded for a further $70m since then) was recouped in just two years. How did they make $45m per year while only earning a few million each year from supplying electricity?
Think about a spinning top. If it’s big and heavy (and well made), it has plenty of enertia and can keep spinning for a long time. It can even right itself when given a bump. Coal-fired generators have huge lumps of spinning metal which provide inertia to the grid. When large increases in electricity demand occur on such a grid, they respond slowly, giving grid operators time to increase supply. Replacing coal with renewables makes the grid like a very light spinning top with bugger all inertia. Bump it even a little and it topples.
This provides a wonderful opportunity for batteries. They can add little bits of energy when the top is about to topple; just enough to keep it upright. And because the toppling of a grid is such a catastrophic event, you can charge a premium for preventing it… because preventing it has been made into a market.
Back in 1917, the price per MWh of these little nudges to keep the grid up reached $10,000. There is way more money to be made by keeping a flimsy top from falling over than building a proper top. The grid operators are aware of this price gouging and the SA grid is installing 4 synchronous condensers (at a cost of $166 million) to increase the inertia in the SA grid on a more permanent basis.
So what is a synchronous condenser? It’s functionally just a big lump of spinning metal… a coal-fired generator without the coal or the generation. Once these are all operational (by May 2021), there will be less opportunity for price gouging.
Making money out of batteries on a grid is complex, and as we have seen, it has nothing to do with the normal ways of making a profit, like reselling electricity for a bit more than it costs you to buy it. You have to hope your grid gets frequently stressed and that you can price-gouge on enough time slices during the year to make money.
It’s also useful to have renewable plants producing plenty of electricity when nobody wants it, because that way you can get electricity to charge your batteries cheap.
In fact, some market-based systems in some places will have wind farms paying you to take it! This is because they get payments for how much electricity they produce, not how much is used.
In short, batteries work best on grids having a mismatch between when energy is generated and when it is needed; particularly in grids that are in frequent crises… which is why the biggest batteries are in SA and California.
Generating energy in a cost-effective manner by well-engineered, long-lived technology and selling it at a profit is considered old-school in the new energy order. These days it’s all about setting up rules that can be exploited by a swarm of lawyers, who will find ways for costly and inefficient technologies like big batteries to make money.
Meredith Angwin’s very readable “Shorting the Grid” delves into the mire of the US electricity systems, some of which are depressingly similar to our NEM. Happily, the US still has enough management variability to permit comparison of the impacts of different kinds of grid management, and the market based approach hasn’t saved anybody any money. But it has certainly made plenty of money for lawyers and people selling gas and inefficient engineering.
The fact that some states have been shutting nuclear plants shows that climate change is an irrelevancy in the policies driving these management structures.
Mining, processing and big batteries
But even if lawyers did find a way to make buckets of money from rolling out truly massive batteries, there is fundamental physical problem making it really tough; a mining and material processing bottleneck. A battery rollout on the scale Finkel predicted or suggested (take your pick) would require an equally large expansion of some of the most filthy mining and refining processes on the planet. Most of which are opposed by precisely the same people who want the big batteries; and everybody else.
We aren’t just talking batteries here, but also wind turbines, solar panels, data centres, mobile phones and smart grids; to name but a few.
Read the articles I hyperlinked in the previous paragraph, or at least look at the pictures. You won’t see images like that on the side of any solar panel back-up battery pack. Battery advertising is as mendacious as meat industry propaganda. Nobody selling big Li-ion batteries is running a “Bought to you by the kiddies of the Democratic Republic of Congo”. The many organisations advocating for renewable energy have a blind spot as big as Antarctica about the entire renewable energy supply chain.
A recent book by French journalist Guillaume Pitron The Rare Metals War relates the history of the mining and processing operations behind the renewable and digital economies being off-shored to China or (less frequently) to other even less developed countries. The translation from the French has additional more recent information from the original 2018 French edition.
The mineral history of the past 30 years has seen China vertically integrate large chunks of critical supply chains. Normally, miners take prices, rather than make them. But with many critical elements in the renewable+digital supply chain, complex processing allows a reversal of this principle.
China offered companies cheap raw materials in return for companies relocating manufacturing to China. Many companies did this, handing China decades of processing expertise in the process. Greedy executives put short-term windfalls before long-term control over supply lines.
This makes the battery+renewable+digital expansion dependent on Chinese mining and/or processing. In 2016 the EU initiated proceedings against China’s restrictions on exports of graphite, cobalt, copper, lead, chromium, magnesia, talcum, tantalum, tin, antimony and indium; some of these are battery building requirements, and others are used in EVs or wind and solar farms. China has become a bottleneck.
‘Finkel-like’ demands for vast battery banks all over the planet imply a rapid expansion of the mining required, and there is evidence now that the Chinese, by which I mean the people rather than the Government, are getting rather tired of carrying the can on this.
Each year in China, there are an estimated 30-50,000 demonstrations from people objecting to the toxic pollution they are being subjected to in servicing global demands for renewable energy, electric vehicles and high tech gadgets.
Pitron cites opinions that it would take the US a decade to get back in the game by re-entering various rare earths and other metal production. And it wouldn’t just be a technical effort, but a political one. Many of the organisations most pushing for batteries would fight tooth and nail against the raw material mining and processing being done in their back yard.
Pitron details the actions being taken by many countries, including Australia, to establish mining rights over the ocean floor as an alternative source of the valuable elements necessary. The technologies needed to exploit the oceans are being developed but are nowhere near ready. The oceans might be just far enough out of sight that mining might be able to proceed at the scale required by the renewable energy industry.
The politics of China’s control of these supply chains will loom large in coming years. Perhaps the Chinese people will demand reform – there are some signs that this is already happening. But what of the Democratic Republic of Congo, the source of almost two-thirds of the world’s cobalt? Blood diamonds are a tiny issue next to the blood batteries in our phones and EVs.
You can read Pitron’s book for details of the other countries who feature in the critical supply chains; behind the renewable and digital revolutions is a litany of exploitation and suffering. But there is also some hope as poor countries realise they have some leverage via a critical mineral. Indonesia is the world’s largest producer of nickel, a critical battery material, and she’s been flexing her muscle with export controls over the past decade, with a ban between 2014-17.
The story of each of the many elements in the high-tech supply chain is complex and different.
So-called big batteries are only big if you haven’t a clue about the size of grids and the real functions of batteries in them. Not understanding energy units is a hint at the profound ignorance underlying the public’s uncritical belief that batteries will save the day. But expanded production is heavily dependent on supply chains requiring both increases in mining as well as complicated processing chains.
It never fails to astonish me how many renewable energy and EV battery articles and reports are written without mention of such issues. If you don’t look, then I guess there’s nothing there.
Donate To New Matilda
New Matilda is a small, independent media outlet. We survive through reader contributions, and never losing a lawsuit. If you got something from this article, giving something back helps us to continue speaking truth to power. Every little bit counts.