Explainer: What the Tesla big battery can and cannot do | RenewEconomy

Explainer: What the Tesla big battery can and cannot do

Print Friendly, PDF & Email

Details of how the Tesla storage array is configured, what it will do and won’t do, and what it will cost.

Print Friendly, PDF & Email


There’s been a lot of discussion about the Tesla big battery since the company announced last Friday that it would build a 100MW/129MWh lithium-ion battery storage installation – the world’s biggest – after winning the state government tender.

The battery will be owned and operated by French renewable energy developer Neoen (which stands for “new energy”), and will be located right next to the 309MW Hornsdale wind farm currently being completed near Jamestown.

But what exactly is it? And what can it do? And how much will it cost? We try to answer some of the main questions, and dispel some of the myths.

How big is it?

At 100MW/129MWh it is the biggest lithium-ion battery storage installation in the world. As Tesla founder and CEO Elon Musk says, it will be three times larger than the next biggest (although bigger ones are on the way).

Why is storage needed?

Two reasons – grid stability, and managing varying renewables like wind and solar.

Imagine a brick wall built without insufficient cement to keep it together. That’s probably a pretty good analogy for how our ageing grid, with its slow moving machines, operates at the moment. Battery storage can be the glue that helps keep it together, mostly through incredibly fast response to system faults – such as transmission failures and tripping gas plants that can make the grid sway like a drunk walking down the road.

And storage will also be able to smooth the output of wind and solar, and allow those operators to store excess output for use at times of peak demand, or when the sun is not needed and the wind is not blowing.

What can this array do?

Battery storage can do a bunch of different things – smooth out the output from renewables when clouds come over or the wind dies down, store power for sale at peak times or high prices, or for night-time use, and even remove the need for new grid infrastructure.

They can also provide a whole range of what are called “network services” – and providing frequency control and “synthetic inertia” are the top of that list. In all, there are about 20 different things it can do in its “value stack” – think of it as the Swiss army knife of the energy system. In practice, one installation may focus on only one or just a few of those pitons.

Will it solve South Australia’s electricity problems?

Some, but not all. It will provide much needed network support, so if there are more network failures, or if the big gas generators trip again like they did in March then battery storage will play a critical role in keeping the grid stable and the lights on, particularly with its fast response capabilities.

It can respond to faults in an instant – in milliseconds – while the big gas generators are still asleep in the back of the bus.  The gas generators take at least six seconds before responding at all, and up to 10 minutes to get dressed.

Over the longer term, South Australia and other grids will need more battery storage, but also more forms of storage that is  longer term (hours and days) and bigger capacity. This could come from pumped hydro, solar thermal, and even hydrogen. Battery storage will likely remain the quickest to respond, and the cheapest over short periods, so will play a critical role.

Would it have prevented the System Black?

Possibly. The battery storage systems has been configured on what would have been required to ride through some of the events that happened last year, including the system black. That event threw up a whole bunch of issues, including the way the system was managed, the un-known fault ride through settings on wind farms, and the lax governor controls on the big gas and coal generators, and a bunch of other things.

But work by Lloyds Register found that a battery of around 100MW could have been sufficient to do the job of dealing with those sudden changes in frequency and voltage in those dramatic seconds when one tower after another collapsed and three major transmission lines failed. Whether the system would have been able to ride through the remaining issues is not known, but batteries would have given it a fighting chance.

baseload to smart grid

How will it change the energy market?

Enormously. This is as significant as the start of the shift from analogue to digital, from a landline to a mobile, from film to digital photos, from switching from a manual gearbox to an automatic, or from a petrol car to an electric vehicle.

It  will change the way we think about energy, from the old baseload, peak load paradigm, to a new one based around the cheapest form of new generation “base cost renewables” (wind and solar), and then flexible or dispatchable generation to fill in the gaps. (See graph above).

That transition won’t happen overnight, but this battery storage installation – and the many others planned for the state and elsewhere, will allow for significant new investments in wind and solar.

In South Australia, this level of wind and solar had reached saturation point with wind and solar accounting already for 50 per cent of demand. But storage will allow those penetrations to jump significantly.

More importantly, it reduces the ability of the big fossil fuel generators to dominate the market and manipulate prices as they have. South Australia has paid dearly for this lack of competition for more than a decade, and the rest of the nation is starting to pay the price too, hence the huge rises in electricity bills.

How is this battery storage array being configured?

Of the 100MW/129MWh in this array, around 70MW of capacity is contracted to the South Australian government to provide grid stability and system security. It will likely mostly provide frequency and ancillary services (FCAS) when needed (such as a major system fault, generator trip or transmission failure).

This part of the battery is designed to last 10 minutes, which sounds short but is long enough to keep the grid stable while slower machines such as gas generators can respond.

The other 30MW of capacity will have three hours storage, and will be used as load shifting by Neoen for the Hornsdale wind farm, where it will be located.

Please note: The 30MW and 70MW ratings mean exactly that – it is the capacity at the connection point.

How much will it cost?

Those details are confidential and media estimates have ranged from $50 million to $200 million. The best indication may have come from Musk himself, who reportedly said that failure to meet the 100-day-we’ll-finish-it-or-it’s-free offer would cost the company about $US50 million. But that refers only to the component contracted to the government, so the 70MW part. So, it is likely to be a number less than $A100 million, but not a lot less.

What is South Australia’s contribution?

We don’t know, but it did flag a sum of $15-$20 million from its newly created Renewable Energy Technologies Fund when the tender was first announced. It is also the contract partner for the grid services, so presumably is making some payments for those. Weatherill has put the total cost to the government – grant and payments – at $50 million over 10 years.

Will it make money?

Yes. Neoen’s strategy is for long term, low yield in investments, and this fits the bill. The battery storage plant being installed by Vector in Alice Springs – to double the amount of solar in what is already the country’s solar capital – will deliver a pay back in less than 5 years on the savings from reduced gas generation. And it is clear that solar and wind couple with storage is cheaper than gas power now.

When will it be finished?

The government wants it finished by December 1 and Tesla and Neoen have committed to doing that. Ironically, the 100-day-or-it’s-free offer will not start until the connection agreement is sealed with the Australian Energy Market Operator. That will likely take a couple of months, so the December 1 deadline will fall before the 100-day deadline.

Will they reduce prices?

Almost certainly. The existence of a new FCAS provider, contracted to the government, will lower FCAS prices, which have jumped sharply in recent years and added tens of millions to consumer bills. Weatherill puts it at $48 million a year. Now, a major new player will enter the FCAS market, so restricts the ability of the existing FCAS providers – the gas generators – from setting any price they want.

Also, the introduction of a new player into the wholesale market – the 30MW component – will have a dampening effect on prices. Not only is there a crucial new player in the market, but it doesn’t take that much to change prices. Really high prices are set when maybe a single, high cost generators sets the marginal cost. If the battery storage can shift that “bidding stack” to the left then it will reduce prices.

Will there be more battery storage installations?

Yes, but by whom and where is still not certain. A whole bunch of different people have projects – Zen Energy, Lyon Group, Carnegie Clean Energy, Reach Energy, but most will reassess their plans in the light of this, and see where their project fits in.

We will see new storage in Victoria, a tender in Queensland, the 5MW capacity in Alice Springs, and lots of behind the meter storage – in households – linked together as “virtual power plants” doing much the same thing as the Tesla battery. Sonnen, Reposit  and AGL are doing this and others will surely follow.

Why did Tesla win the tender?

Probably because they offered the cheapest price, but not necessarily. Premier Weatherill said: “There were some great local bids and some exciting ideas but this was just head and shoulders above all the other bids in terms of price, reliability, capability of delivering by 1 December.”

RenewEconomy understands that by partnering with Neoen, they could get access to a connection point which needed no further upgrade, and who were already working on connection points.

And with Tesla and Elon Musk, you can’t buy this much publicity. You get the sense that the mainstream media would get excited if Musk announced the opening of an envelope.

Look at the media interest a week earlier when Neoen announced the country’s biggest battery – 20MW/34MWh – to be built alongside a 196 wind farm to power the country’s largest glasshouse for vegetable growing, that was going to be built overseas because of the high grid and gas prices. The project, attracting $560 million in investment and 1,300 jobs, got virtually no mainstream media coverage.

How long will the batteries last?

RenewEconomy understands the batteries carry guarantees for 15 years, although these guarantee contracts allow for some level of degradation each year. So, for instance, after 15 years, the batteries may only be operating at around 60-65 per cent of their initial rates capacity, so will then be replaced by newer, more efficient and cheaper equipment.
How many houses will it power?
A common question, but an irrelevant one. Deputy prime minister Barnaby Joyce says it is just one grain of sugar in a teaspoon. As we try to get our heads around that analogy, we should remember that if you have a flat tire, you don’t buy a new car.  This battery array may be small, and may only have the ability to power a few homes for a short period of time, but that’s not what it’s there for.
It is there primarily to deal with extreme events at the margins of normal operations. Peaking gas generators have long been designed to operate for just a few hours of the year.
These batteries will have similarly specialised services. Much of the capacity will be used to provide system stability, and the rest to hedge the output of the wind farm and shift output from low to high demand. In the end, though, they are likely to operate every day, as opposed to only a few days a year for some peaking gas units.

tesla storage south australia


What happens when the wind doesn’t blow?
One common misconception is that if the wind doesn’t blow then the battery cannot be charged, and so once used (discharged) will sit there uselessly. Not so. The battery has a separate metering so it can charge from the grid if the wind is not blowing.
 Will it help address other power system issues?
Last week saw the first time AEMO invoke constraints on some wind farms because it did not have enough gas generators on the grid. AEMO has taken a safety first approach, particularly in relation to the need for “fault current,”, and says that if four gas generators are not operating when wind output exceeds 1200MW then wind will be curtailed back to that level.
That measure is seen as heavy-handed, and is having an impact on pricing. It is thought that this battery storage will help AEMO understand how this system works, and on how to use measures and develop more sophisticated and cheaper ways of ensuring stability in the network.
 How many people will work on the project?
About 100 people will work on construction, about the same as on the construction of the third stage of the Hornsdale wind project. Indeed, it will be more or less the same crew, with the addition of Tesla engineers. This is the biggest thing for Tesla Energy, so they will be sending enough to make sure it is done properly. Maybe Even Musk might be looking for a motel room in Jamestown, although he is known to sleep on the factory floor.
How big will the installation be?
About 1 hectare. the batteries are installed in waterproof cabinets rather than in containers or special buildings.
Print Friendly, PDF & Email

  1. Charles 3 years ago

    Fantastic summary. Almost every time I was online over the weekend I found another article about this battery system and found myself correcting morons in the comments saying it was a scam because it would only support the state for 5 minutes. In the future I can link direct to this article!

    • Joe 3 years ago

      …and Bananaby Joyce should be first on the list to receive the link.

      • Marcus L 3 years ago

        He probably wouldn’t read it. he doesn’t like facts.

        • MrMauricio 3 years ago

          only alt. facts-learned strategy form Trumpelthinskin

      • Trent Deverell 3 years ago

        I once had Ken’O’Dud (LNP) tell me Solar has got no guts to pump enough water…. [lol]

        Maybe we should see these fools appointed to be fuse plugs or just tie them wastes of oxygen up to HV busbars in the sub and watch the fireworks show !!!

        • Goldie444 3 years ago

          He was mostly likely trying to use the solar power direct from a panel, the DC power.

  2. Keith 3 years ago

    Thanks Giles, nice summary. Not sure that it will change a lot of the press coverage, which is trying to talk the story down, but hopefully the essence will seep through.

    • Trent Deverell 3 years ago

      I had the Gladstone Observer tell me I should stop sledging the LNP members over their love affair with all things Coal, Adani etc….

      Now that Giles has penned a well formed and 100% factual RE article I have a longer handle to take a harder swing at the bogans.

  3. George Michaelson 3 years ago

    The time difference between recharge and reuse. Will it limit its availability?

    Is there an intention to have full discharge of the revenue raising part, the state government part, or both on some cycle? Can the contracted part be discharged by having the remainder stand in, and thus earn revenue?

    Can it be functionally useful at 50% charge state? 30%? can it meet contract if it has to back up twice inside one hour, to perform its functional role in FCAS as well as supply of energy?

    Is it operated at eg 75% capacity so it can also absorb surplus power but stand ready to supply volts if needed? Is it deliberately discharged to provide a ‘hole’ to fill, from other sources?

    (I’d expect the answers to be good ones btw. This is not trick question time, I’m trying to understand how a battery stack plays. Simply having it there but not discharging it seems silly)

    • neroden 3 years ago

      I figured it out — see my comment.

      • George Michaelson 3 years ago

        Thanks! your other comment makes sense to me.

      • Mike Dill 3 years ago

        I think that it is even better than that, as the batteries can use the wind farm as a load if absolutely necessary. If I am thinking correctly, the synchronous momentum of the turbines can adsorb and discharge small amounts over a few seconds to minutes.

        • jeffhre 3 years ago

          The equivalent of motoring for a hydro-power plant with pumped storage?

        • neroden 3 years ago

          I think it’s important to emphasize that Tesla can under all circumstances meet contractual obligations, but can *usually* do better than contracted.

          There’s a large margin of safety in the design.

  4. Neville Bott 3 years ago

    Great work again Giles.

    This will be a great reference point when the skeptics run their misinformation campaigns.

    • Trent Deverell 3 years ago


      Outstanding work Giles… and I am proud to contribute a modest monthly donation to see good journalism happen.

  5. MikeH 3 years ago

    >”So, for instance, after 15 years, the batteries may only be operating at around 60-65 per cent of their initial rates capacity, so will then be replaced by newer, more efficient and cheaper equipment.”<

    In the Q&A, Musk suggested that any degradation over time could be dealt with by progressively adding additional PowerPacks to the installation to bring it back up to the original spec rather than junking the original PPs. Still a long way off so who knows what the most cost effective solution will be but not the huge problem that the critics claim it is.

    • neroden 3 years ago


      See my description in the comments of how I think this works. Each individual battery is sort of contractually sliced:

      So think of the capacity of a battery as follows:
      2% base anti-bricking protection, always full
      9% “always full” until the government needs emergency power
      70% for Neoen to fill and empty as it likes
      9% “always empty” until the government needs to absorb excess power
      10% anti-overcharging protection at the top, always empty

      I may not have the allocation between anti-bricking protection and anti-overcharging protection right.

      But anyway. If the batteries degrade in total capacity, they can simply add more batteries until they meet the contract requirements for both the government and Neoen contracts. (I’ve never actually seen li-ion batteries degrade in *charge rate* as opposed to capacity; it seems like the same thing would work, though.)

      • MikeH 3 years ago

        And by limiting the Neoen slice to 30MW/90MWh, they don’t grab all the arbitrage opportunity for this one battery install leaving some incentive for other investors like the Lyon Group or Zen Energy.


      • Mike Shackleton 3 years ago

        Just from what we know about Tesla batteries in their cars – You only charge them to 80% under regular use. If you need the full range of the battery, say you’re driving Melbourne to Sydney, then you tell the car your plans and it charges to 100%.

        So for the stationary application – using predictive modeling for consumption and generation, you can keep the batteries in a protected mode for 99% of the time, only charging up to 100% if the model calls for it.

        Also, the key word in his statement is MAY – “So, for instance, after 15 years, the batteries may only be operating at around 60-65 per cent of their initial rates capacity, so will then be replaced by newer, more efficient and cheaper equipment.”
        It is probable that some units will still be working with minimal degradation and some will be worse. But we’ll never know until we have a real-world test to see how it pans out. This is incredibly valuable IP for Tesla. Perhaps there is a knowledge sharing clause in the contract to make it more attractive.

        • Brunel 3 years ago

          ARENA is testing batteries including the Powerwall to see how many cycles it can actually do!


        • Dan Paulson 3 years ago

          You can’t make any judgements based on the car batteries. Different chemistries and cycle lives, as well as energy.

        • Greg Hudson 3 years ago

          You don’t need to charge to 100% to get from Melb to Syd. 80% will get you from Melb to Euroa (easily). Next stop is Albury, then Gundagai, Goulburn, then you are in Sydney. New SuperChargers are popping up all over the place. Melb-Adelaide to be completed by end of 2017. A new one in Nowra implies they are working on the coast road from Sydney to Melb. No doubt we will see many more. Musk is rumored to have said that we will be able to SuperCharge our way around Australia in the not too distant future (2018/19?). Check the map here:

    • BushAxe 3 years ago

      No doubt this will be an invaluable research project for all involved, I’d expect Tesla to trial new batteries here so in the future as power density of new cells increase so will the total capacity of the site.

      • Mark Brown 3 years ago

        Oh good more costly ideology experiments for SA as we are already getting .40c more per GST $1 than this state earns. Plus our huge State debt which is growing. Most of you must be University Students or receiving easy Government Money.

    • Barri Mundee 3 years ago

      I assume the lithium would recovered and be available for either new batteries or other uses?

      • Be 3 years ago

        Tesla has said it would reclaim the lithium, but cobalt and nickel almost always get recovered from batteries because of the high value. Big battery packs also are more likely to get recycled.

  6. George Darroch 3 years ago

    “if you have a flat tire, you don’t buy a new car.”

    This is a fantastic analogy.

    • Trent Deverell 3 years ago

      Well funny you say that because in Iraq if a truck blows a tyre or develops a mechanical fault delivering fuel or oil to the army they immediately torch the truck and load.

      So the moral of story is that it really does happen when fossil fuels are involved.

  7. MG 3 years ago

    So can this battery supply the fault current needed in order for AEMO to remove the “must have 3 or 4 gas turbines online at all times in SA” policy?

    • neroden 3 years ago

      It should. I honestly don’t know what AEMO means by “fault current”.

      It should be sufficient to stabilize frequency and voltage instantly, and for 10 minutes, under most circumstances, with a few obvious caveats which you probably know.

      — Wind turbines and solar panels need to have the correct ride-through settings (I believe this is already being done).
      — The gas turbines would still need to be on standby, ready to start on demand and be up and running within 10 minutes.
      — The capacity is not large enough to replace a full Heywood interconnector or Murraylink failure, though this is only relevant if power is flowing into SA and it’s often flowing out.
      — The capacity is not really large enough to replace SA’s largest single gas turbine. If any single gas turbine *smaller than 70 MW power capacity* suddenly fails, this will handle the next 10 minutes. That’s most of the gas turbines.

      But if one of the oversized turbines at Pelican Point, Torrens Island, or Osborne is operating and full capacity and fails, they’re each bigger than 70 MW, so the capacity would have to be replaced in less than 10 minutes.

      That said, Pelican Point hasn’t been running anyway and didn’t turn on in the last blackout, so it’s kind of irrelevant. And apparently AGL plans to close half of the Torrens Island generators and replace them with individual smaller generators.

      FWIW the largest single individual generator in the system appears to be 200 MW (at Torrens Island). So it would only take two more of these projects (70 MW x 3 = 210 MW) to cover a failure of one of the Torrens Island B units.

      • BushAxe 3 years ago

        Pelican Point generally runs at 50% capacity (1x240MW) it hasn’t run at full capacity for some weeks (I suspect there’s issues with the second unit). In low demand/high wind periods it shuts down completely and leaves TIPS A or B to provide fault response by running 2 or 3 units at low capacity (20% ea). In high wind periods I can’t see why Neoen wouldn’t fill their 70% and use it to provide fault capability as it wouldn’t be needed for dispatch at the time.

        • neroden 3 years ago

          Yes, I strongly believe that they would do so. However, with only a 70MW + 30MW = 100MW grid connection, if Pelican Point were running at 240MW and suddenly tripped out, 100MW rate isn’t enough to make up for it (even if there is enough energy in the batteries).

          This shows the problems with single very large turbines, as opposed to multiple smaller generators. If any smaller *single* generator failed, it could be covered.

      • Cooma Doug 3 years ago

        The protection systems on a grid are complex. They need to be, above all else, able to descriminate and accurately determine fault location and disconnect it from the system, attempt immediate reclose, then lock it out if the fault lingers. In order to determine these things the grid needs to be capable of consistant adaquate response and deliver current into fault areas that correctly reflects the situation. In a complex grid, faults can be undetected and linger without accurate protection response and lead to a cascading breakdown of the system. So we need accurate and rapid system reaction to faults as well as demand and we need the capacity in the system to deliver those specifics.
        In some countries there are faults on grids that linger undetected and energy thieves as well.

        • neroden 3 years ago

          Aaaaah, “deliver current into fault areas” is what AEMO means. Thanks. (In the rest of the world, “fault current” is a synonym for “short circuit current”)

      • JonathanMaddox 3 years ago

        “Fault current” is the maximum current that might flow in the case of a short circuit fault. Synonyms are available fault current, short circuit current and prospective short circuit current.

    • Trent Deverell 3 years ago

      Well 3 or 4 gas turbines would be in the order of 600-800MW, or in rough numbers equal to Victorian inter-connector…

      Maybe AEMO’s thinking on such a policy is not so much about the variability of renewables, but what happens when the interconnector suddenly trips off and takes 800MW of capacity instantly.

      If a gas turbine is off-line it can’t just go from zero to hero in milli-seconds like battery storage can, but what 100MW of battery would cover is a nominally loaded Torrens Creek or PPGT unit suddenly tripping off. .

      There are similar rules that govern how much and from where power is generated and/or dispatched depending on what Kogan Creek is doing…. the bigger the source that can drop off-line then complicates the contingency needed.

      • neroden 3 years ago

        “the bigger the source that can drop off-line then complicates the contingency needed.”

        Yep. This is why distributed is more robust than centralized!

  8. John Boland 3 years ago

    This is a very useful summation of the attributes of not only this particular system but also what is in store with more store, and don’t excuse my word play. One point that should keep being emphasised is in reference to the BIG BLACK. You state “That event threw up a whole bunch of issues, including the way the system was managed, the un-known fault ride through settings on wind farms”. When the system in SA was managed by ESIPC, one of the things that was emphasised to me (I did quite a lot of work for them) was that part of their job was to have knowledge of the ride through capabilities of all generation facilities attached to the grid. Perhaps, and I just say that reservedly, that went by the board when they and others joined with AEMO.

  9. David leitch 3 years ago

    To the extent the battery is used for frequency control it may not even be kept fully charged. In other jurisdictions the battery might be kept say 70% charged to deal with under frequency or over voltage as well over frequency, under voltage

    • MG 3 years ago

      What about frequency regulation – is it going to do that too (continuous charge/discharge as instructed by 4 sec AGC signal)?

      • neroden 3 years ago

        I don’t know, but now that I’ve figured out the design, I would expect it to do frequency regulation. It should be super easy.

  10. neroden 3 years ago

    Fascinating. Thank you, Giles.

    Government part contracted: 70 MW * 1/6 hour = 11.7 Mwh storage
    Neoen part contracted: 30 MW * 3 hour = 90 Mwh storage
    Total contracted capacity = 101.7 Mwh storage
    Actual capacity: 129 Mwh storage
    Percentage of capacity contractually required to be available = 101.7 / 129 = 79%

    This makes sense; they usually won’t have to go into the top or bottom 10% of the battery charging range.

    “C rate” of charge/discharge when using the goverment part if the whole system can discharge : 0.54 (which is EASY and lower than a Tesla Model S discharges its batteries). This should always be possible, even if the Neoen side is “empty”, using the reserve top/bottom 10% of the batteries, which is about 12.9 Mwh. Still leaves 1.2 Mwh to avoid “bricking” by going to zero.

    I see how clever the design is here. There are a whole bunch of parallel batteries. The bottom of each battery (from 0% full to 1% full) is reserved for anti-bricking. The next 9% of each battery is reserved for the government if discharge is needed, allowing a very high power rate discharge with a low C-rate. The next 69% of each battery is allocated to Neoen to fill and empty as they need. The next 9% of each battery is reserved for the government if charging is needed (absorbing power), allowing a very high power rate charge with a low C-rate. The final 1% is reserved to avoid overcharging/explosions.

    • BushAxe 3 years ago

      Yep, this will address the runaway costs of FCAS, introduce some fast response backup and give renewables a great idea as to how storage will work for them. It kicks some big goals.

    • Scottish Scientist 3 years ago

      So think of it as follows:
      2.6 Mwh (2%) base anti-bricking protection, always full
      11.7 Mwh “always full” until the government needs emergency power
      90 Mwh for Neoen to fill and empty as it likes
      11.7 Mwh “always empty” until the government needs to absorb excess power
      12.9 Mwh (10%) anti-overcharging protection at the top, always empty

      What’s your source for those figures?
      Claiming a “129MWh” capacity is more mis-selling or hype if only 11.7+90 = 101.7 MWh is actually usable.

      Of the 100MW/129MWh in this array, around 70MW of capacity is contracted to the South Australian government to provide grid stability and system security. It will likely mostly provide frequency and ancillary services (FCAS) when needed (such as a major system fault, generator trip or transmission failure).

      What happens if and when the Torrens Island Power station and its 1,280 MW develops a “major system fault, generator trip or transmission failure”?
      70MW is not going to cut it if they lose Torrens. Lights off.

      The other 30MW of capacity will have three hours storage, and will be used as load shifting by Neoen for the Hornsdale wind farm, where it will be located.

      So the excess MWH are sold to Neoen who doesn’t need a high power rate but can profit from capacity.

      Actually they do need much more storage regeneration power and more energy storage capacity than 30MW / 90MWh.

      As I have explained previously, your average wind farm needs an energy store with regeneration power of 1/7th capacity (which is 45MW – or 44MW if the farm is only 309MW as claimed here rather than the 315MW they claim on the farm’s website) and they need 1.5 days, or 36 hours at that regeneration capacity, or 45MW x 36hours = 1,620 MWh.

      Especially 90MWh is only 5.5% of 1,620 MWh, which is grossly and disgracefully little capacity that will fail to provide any kind of a respectable 24/7/52 on demand power performance from the wind farm / battery pairing.

      It’s too expensive to do grid energy storage by battery, except for the shortest timescales, where yes batteries are useful but you can equally well use many other short term energy storage methods, which I won’t go into right now.

      And by the way, pumped-storage hydro is no slouch off the mark and is routinely used in the UK for responding to demand spikes.

      This whole scenario cries out for pumped-storage hydro and it is about time that Australians started listening to their own scientists who are telling them that.

      This battery is not fit for purpose.

      • neroden 3 years ago


        Battery capacities are always quoted “gross”, without excluding “bottom reserve” or “top reserve”, because those vary by APPLICATION. You really don’t know much about the battery field, do you? Industry standard quotation procedure.

        The entire battery is usable. They have simply *contracted* with the South Australian government for a *particular application* which requires keeping some of it empty.

        The key issue for most battery installs is that at the very bottom and very top of the battery, it charges and discharges quite slowly. It charges and discharges fast in the middle of the charge range.

        Being *specifically contracted* to charge and discharge fast (high kW), it becomes necessary to stay in the middle of the range, which requires extra battery kWh.

        I’m making an educated guess (because I *am educated* in the battery
        field) based on Giles’s key information regarding the *contracted
        abilities* of the battery.

        There are *two contracts*. It’s very clever here. It’s like cracking oil and selling gasoline to one person and asphalt to another. The batteries have a specific natural ratio of kW to kWh. This isn’t the ratio which their customers want; each customer has a different desired ratio. What Tesla figured out is that you can sell 70 kw / 11.7 kwh to one customer and 30 kw / 90 kwh to another customer out of the same battery, whose real kw / kwh ratio is actually roughly 1 / 4.

      • Jacob 3 years ago

        Why does the wind farm need to provide 24/7/52 power? Do we ask taxi drivers to ensure that they always have a back-up driver around in case they want to go to the bathroom or take a day off work? No. Our transportation system has trains, buses, taxis, and private vehicles. Buses and trains are contracted to provide ‘baseload’ transportation and the government subsidizes that in the public interest, but taxis are still a valuable part of the industry where they can quickly allow a customer who misses the bus to get from A to B. No one ever suggests mandating that taxis make schedules like buses.

        Windfarms are taxis.

  11. Paul McArdle 3 years ago

    Thanks Giles, best coverage I have seen so far.
    Goes some way to answering the questions I’ve been pondering since the tweet-storm of March, including these:


  12. john 3 years ago

    Even with this small size against the usage per day of the state the fact it should have a very strong effect on gaming the price structure just may encourage more and more backup that can be called upon to stop these practices.
    The sooner more renewable energy generators are implemented over a large area with in the grid together with storage that can work as a bridge to stop high price flow on to the end user the better for everyone.
    With the implementation of RE and as much storage as possible it is just possible; I think is going to happen and that SA will have a lowering of retail price of energy.

  13. Andy Saunders 3 years ago

    “would cost the company about $US50 million”

    Presumably that’s Tesla’s cost, rather than what it priced to sell at. And US$50m is around A$67m…

    “Tesla found and CEO Elon Musk” Although that’s commonly repeated in many places (usually as “co-founder”), I believe actually he financed the Series A round rather than founding Tesla (the founders were a bunch of engineers obsessed with high-tech motors).

    • Dan Paulson 3 years ago

      Musk won the right to call himself a founder, in a lawsuit with the CEO he displaced.

    • neroden 3 years ago

      There was a legal settlement regarding who gets to be named “founder” of Tesla. 5 people. In reality, as opposed to law, it was founded by 2 of them.

  14. Brunel 3 years ago

    Would it not make more sense to install the 90 MWh of storage at the point of use? Or in 9 different substations in Adelaide?

    • Michael Turnnidge 3 years ago

      Oh yeah! Once again we perpetuate the centralised power source. Why?

      • Catprog 3 years ago

        Probably because the wind farm needs open areas not the middle of the city.

    • juxx0r 3 years ago

      We could totally make this 9 times more complicated, or we could prove it works and do it nine more times.

    • JonathanMaddox 3 years ago

      Batteries are already installed at substations and I’m sure it will happen more often in future. But space is at less of a premium in rural areas, and there’s plenty of concentrated “use” for storage services at the wind farm site as well.

  15. Michael Turnnidge 3 years ago

    Great informative explanation about how this thing will work, apparently written by someone who knows.thankyou, I will surely share amongst my friends and colleagues.
    My wonder is about the contract and more about how he won it.
    I wonder if this may be a leadin to enable him to build his next megga battery factory here right on top of the mineral deposit needed for the batteries. I am led to believe that Australia has one of only three deposits.

    • juxx0r 3 years ago

      Australia has heaps of lithium, in lots of places dotted around the country. We have graphite in similar distribution, nickel and cobalt in all the likely places.

      We are the second largest lithium producer, possibly set to become the largest when the hardrock guys hit their stride.

      • Trent Deverell 3 years ago

        But our country’s leadership is one generation disposed from the jurasic era, and they takes their ques from the giant man-baby and his swamp of fosilised millionaires that occupying the white-house…

      • Greg Hudson 3 years ago

        We may have the lithium, but from what I understand, the Chinese are buying up as many mines as they can…

  16. Rod 3 years ago

    As many will be aware, we in SA nearly had another system black in March? this year when 3 of 4 TIPS B units tripped due to a substation fire.
    The voltage excursion caused every rooftop solar array to disconnect!
    An hour earlier, when PV input was higher, it would have been lights out.
    If this battery can prevent this happening again it is worth every cent.

    • Trent Deverell 3 years ago

      According to George Christensen, Federal MP for Dawson (or Adani/Carmichael) he recommends South Australian’s should go buy petrol generators from Bunnings.

      To be over-simplistic 100MW of petrol bangers would cost north of $32M, without extension leads and all the obvious stupidity.

      ….so it goes $50M is good price for the big Tesla kit, and if it can deliver FCAS and peaking capacity for under $250MWh that is entirely competitive!!!….

      But we all know George is a bit of mathematical dunda, to not see that just 8,000 Powerwall2’s at $10k would also be another way to provide a similar storage capacity whilst additionally proving very useful every day for next 10-15 years…

      To think that those PW2’s would be entirely paid off within half a decade and saving the owners money there after, you gotta wonder about the leadership in Canberra and size of brown paper bags they collect from the MCA, QRC and Coal Lobbyists……

      • Dan Paulson 3 years ago

        I would like to see the math that has the PW2 units paid off in a half decade.

  17. Jon 3 years ago

    The 70MW/10MWh Govt Reserve Capacity will only be used in an emergency (this is the govt insurance policy which will be matched with their new gas generator).
    The Tesla battery is actually only about 80MW with a limited use “boost mode” to provide the 100MW required by the tender.

    • Trent Deverell 3 years ago

      But as a bonus will the PowerPack SA-edition come with a Sport, Super-sport and “Insane” modes like the totally trick Tesla cars do !!!

  18. Mike A 3 years ago

    Why does Elon say 100MW instead of 100MWh? Is this AUD or USD dollar figures? Anybody know?

    • neroden 3 years ago

      MW are power (how fast you can charge or discharge). MWh are energy (how much you can store). They are TWO DIFFERENT THINGS and they both matter.

  19. Robin_Harrison 3 years ago

    It would be great if Tesla used its newsworthiness to promote the fire safety of their power packs with the test they’ve already done and do another one with PW2. Otherwise the fossil fools will hamstring storage rollout with their ridiculous fire regs.

  20. Gabi Gerrie 3 years ago

    Thanks for that clear explanation. So the batteries are just a safety net as it were to buy time for the grid to sort itself out if needed.

    • neroden 3 years ago


  21. Mark Brown 3 years ago

    All you pro renewables together with this totally incompetent Labor State Government were crying out that the State wide blackout was caused only by “major Electric towers were blown over”. You lot still try to say renewables are cheap while we now pay the dearest electricity prices in the entire world while we export billions of tonnes of cheap coal. You lot have no business & common sense!
    Say what you like but facts are proving you nuts wrong!
    Ask the main question that has put us on this economically disastrous path ” how much carbon pollutants we will save & will that change anything?

  22. Michael Murray 3 years ago

    A number of people have been pushing this page over on The Guardian website to combat the ignorant commentators. The Guardian has put up something informative of its own today


  23. Caffined 3 years ago

    There should be no mystery to this, there have been several other Powerbank projects with known costs.
    Musk himself has tweeted that Powerbank costs have been reduced to $250 /kWh at pack level (less inverters, installation , etc)..for 100MWh + orders.
    Prior to this, Tesla had a Powerbank system cost calculator on their site where you could estimate a total cost. Even then , at the higher prices, a 420kWh powerbank system with all ancilliaries inverters, controls etc , was costed at $145,000…installed.
    That put the installed cost at approx $350,000 per MWh.
    So the SA 129MWh system cost should be around us$45 m (Au$60 m).
    But remember that was at the “old” pricing levels, and before Musk later stated he was going to share half the cost of the system himself because of the learning Tesla will gain from this exercise !
    So i dont see this project costing SA more than Au$30-35 million (+ site costs) .
    Not a big deal in the scope of Energy utilities….
    …..and a fraction the cost of a vote on Gay Marrage. !

  24. Aluap 3 years ago

    Nice article, except that I didn’t understand the image “Imagine a brick wall built without insufficient cement to keep it together”.

Comments are closed.

Get up to 3 quotes from pre-vetted solar (and battery) installers.