Tesla big battery moves from show-boating to money-making

Graph courtesy of Dylan McConnell from Climate and Clean Energy College.

The Tesla big battery has been the big news of the Australian electricity sector since its formal opening at the start of December, grabbing headlines for its many interventions in the energy market and the first-of-its-kind demonstration of the next generation of technologies.

The Tesla battery, formally known as the Hornsdale Power Reserve, is perhaps the biggest development in an industry that has seen little real change in nearly a century – its impact is probably more profound than even the spectacular cost reductions in wind and solar power.

For much of its first weeks of operation, most of the Tesla big battery’s activities were a mixture of tests, trials and just a little showboating – going from zero to 100MW in 140 milliseconds, jumping in ahead of contracted generators to arrest falls in frequency as big coal generators tripped off-line, and just generally being smart and fast.

Some of its early manoeuvres were spectacular but didn’t appear designed to be making money, focusing more on testing its capacities and abilities, particularly its speed of response and reliability, as much for the satisfaction of its builders (Tesla) and owners (Neoen) as for curious third parties.

That phase seems largely complete, and in the midst of the heatwave and the gyrations in supply, demand and pricing last week, it seems that Tesla is now able to demonstrate its ability to cash in and make money for its owners.

This was particularly visible in the price spikes that hit the South Australia market on Thursday and Friday, as this graph at the top illustrates.

The blue indicates generation, and mostly coincides with high prices, particularly the big spikes on Thursday and Friday afternoons. The red indicates charging, or load, and this occurs mostly during lower priced periods.

Another view of this data is presented below, showing the actual price achieved during the buying (charging) and selling (generation). It’s hard to be sure, but it might have made around $1 million over the two days from the wholesale market.

The advantage of the Tesla big battery over other sorts of storage, and particularly over fossil fuel generation, is its ability to switch on and off in an instant, meaning it is capable of responding immediately to change in demands and pricing.

The Hornsdale operator uses around 30MW/90MWh  in the wholesale market as a form of price arbitrage, balancing wind output from the neighbouring Hornsdale wind farms, taking advantage of spikes and dips, generating when supply gets tight, and making micro adjustments to supply and demand.

The rest of the battery – around 70MW, 39MWh – is contracted by the South Australia government specifically to deliver “network” services, such as frequency control at a time of system faults and problems.

Tesla has demonstrated its ability to respond quickly by jumping into the market in other states in recent weeks, such as when the Loy Yang A units failed on repeated occasions and when a unit at the Eraring coal generator in NSW also failed.

It jumped in again, last Thursday, possibly for a little more showboating, or testing, when the Loy Yang B unit tripped and pushed frequency outside its normal operating band.

The fact that the battery focused on network services carries less storage capacity is because the services they provide are usually of short duration – such as arresting changes in frequency within 5 minustes.

Exactly how the contract with the South Australia government works is not clear, like so much of the opaque electricity market – but some insight into its capabilities was given on those same days by its offering to the frequency control and ancillary services (FCAS) markets.

These three graphs above – again from McConnell at the Climate and Energy College, and sourced from NEM data – are a good illustration of how the battery can divide its services between several different needs.

There is no judgment here on the success or otherwise of its offerings, but more of an illustration of its multiple levels of operation. So far, the battery has pleased its owners (and developers), and has exceeded expectations.

This is important because the unique proposition of the battery, of course, is that it’s speed, particularly from a standing start, its modularity (it can be built up from very small units to a very large one), and its speed of construction (less than 100 days in this case).

It makes a mockery, for instance, of the Coalition’s derisory criticism that the battery is no more useful than the Big Banana in Coffs Harbour, or that it could only power the whole grid for just a matter of minutes.

What it does do is point the way to the future, and the transformation to a smarter, cleaner, faster, cheaper and more reliable grid than what we have now.

More batteries are on their way – in South Australia, the Northern Territory, Queensland and NSW – and more will follow once the market rules are adjusted to catch up with the technology.

Comments

96 responses to “Tesla big battery moves from show-boating to money-making”

  1. Joe Avatar
    Joe

    Premier Gladys, please get us a Big Battery in NSW!

    1. Michael Murray Avatar
      Michael Murray

      I thought you had a Big Stadium instead 🙂

      1. MrMauricio Avatar
        MrMauricio

        Yep-same as Perth!!!

        1. Joe Avatar
          Joe

          ….the Perth does look quite ‘glitzy’ from the outside…at least what I saw on the TV from over here in Sydney.

      2. Ren Stimpy Avatar
        Ren Stimpy

        The stadiums will be named Pointless and Futile. As in “Tonight on Nine Parramatta takes on Manly at Pointless Stadium”, “Followed by the Roosters and Penrith at Futile Stadium”

      3. Joe Avatar
        Joe

        ….err, not just yet. There is one knockdown and rebuilder happening at the moment with the ‘New’…. Parramatta Stadium. Then we have the proposal for 2 more knockdowns and rebuilders, one at Homebush, Sydney ( Olympic Stadium ), the other at Moore Park, Sydney ( Sydney Stadium ). The rebuilders..how Big….is the….BIG question.

    2. rob Avatar
      rob

      Soz Joe….. but you are stuck in the backward 1900’s state of NSW ……just keep burning your Black juice ……while we here in S.A. make it all happen…..What pisses me off is that we pay more for our power here in S.A. than you do when you import so much from QLD!

      1. MrMauricio Avatar
        MrMauricio

        Your main supply comes form gas-which is expensive.Private operators of those power stations game the market as well.Change is a -happening-in the wind tho!!!

        1. rob Avatar
          rob

          main supply? We are on average over 50% clean ……..and yes I hate the red line(gas)

      2. Joe Avatar
        Joe

        NSW is indeed ‘King Coal Territory’. It will change but a liiiiiiiiittle bit slower than the good State of SA.

    3. Brunel Avatar
      Brunel

      You are better off voting for a minor party that wants to install a big battery and tram lines.

      1. Joe Avatar
        Joe

        We are getting the Trams in CBD Sydney. I guess The Big Battery will just have to wait until all those Tollways,Stadiums and ill planned Metro Rails are all completed.

        1. Brian Tehan Avatar
          Brian Tehan

          …or the current NSW government is voted out.

          1. Joe Avatar
            Joe

            …its another looooooooong 12 months before the next NSW State election…plenty of time for Gladys and her mob to create more damage.

          2. Brian Tehan Avatar
            Brian Tehan

            Yes, they recently sold off 6 hectares of wildlife reserve and precious Koala habitat in Port Stephens (the Mambo Wetlands), to a developer, for the princely sum of $250,000. Habitat that’s absolutely essential to their continued survival in this semi-urban/bush area. The developer wants to develop it because it’s worth, probably, 50 times what he paid for it.

          3. Joe Avatar
            Joe

            Yep, the environmental vandalism begun by the Baird is faithfully carried on by the Gladys. Any patch of Greenery in NSW is fair game for these Liberal Destroyers.

  2. Sophia McGrane Avatar
    Sophia McGrane

    what a ripper!
    modern day storage/power generation acting exactly like it should as part of a (hopefully sooner rather than later) modern day National Energy Grid
    thank you Jay Weatherill and Tom Koutsantonis!

  3. Garth Power Avatar
    Garth Power

    At the time of the trip, it was dispatched as 63 MW of fast raise (6 seconds) as well as slow and delayed raise. Would have been interesting if they had been dispatched as Raise Regulated at the time and not the contingency raises.
    There were dispatched as Lower Regulated but that wasn’t going to help the scenario.

    For the average FCAS price for the TI was for the entire day and if they were dispatched at the price the battery would have taken in $24,755.78 and only has to run if something large trips. or $9 Million PA.

    Not bad considering you just have to be in standby mode.

    Raise/Lower Reg they get a signal every 4 seconds if required so they would have to dispatch.

    1. itdoesntaddup Avatar
      itdoesntaddup

      Interesting calculation. But that’s surely just for FCAS pricing on the day of the trip, not the average revenue across 365 days. It’s hard to see those prices being maintained unless they join the cartel.

      https://www.aer.gov.au/system/files/20170828%20FCAS%20%245000%20report%20%28SA%29.pdf

      1. Garth Power Avatar
        Garth Power

        They actually looked lower than the norm, would expect them to easily pull in $10 Million P.A as it can deliver so much support.

        They also get paid to pull power out under Lower Regulation. When you see it in load between 27 and 0 its likely that is under FCAS support. Then they sell the power back they were paid to take in the energy market. Its harder to calculate out as they are constantly bidding in that market but its also valued a lot higher at like $15 on average. If they are dispatch 1/2 the day at $15 that would be an additional $5,400/day and they get a recharged battery to sell the energy back in the energy or raise reg market.

        If you had exactly the same set up you could cover the entire FCAS market NEM wide with about 5 of them.

        In that article they require 35 MW of Raise/Lower regulation while Vic-SA inter connector has one or more cct out. The battery can do 30 so they would have to source 5 from someone else. As it can deliver while charging or discharging it should hold the prices a lot lower.

        1. itdoesntaddup Avatar
          itdoesntaddup

          Can you point to any public data sources that might allow us to look at the FCAS market?

          1. itdoesntaddup Avatar
            itdoesntaddup

            Thanks – I’ll take a look and see what I can untangle.

            Excellent site BTW – you’ve really put in some good work.

          2. eljmkt_nemlog Avatar
            eljmkt_nemlog

            Thanks. Great to see your detailed analysis and discussion. You are using the site exactly as it was intended. More graphing configurability and features under development with the aim creating graphs like the ones developed for your analysis.

  4. Robert Westinghouse Avatar
    Robert Westinghouse

    Excellent news against a background of LNP incompetence….Well done SA. Well done Elon Musk. Viva la batterie!!!

    1. maeon3 Avatar
      maeon3

      Musk’s Gigawatt sized battery packs need to pepper the surface of the earth as Tesla owners create huge shocks to the power grid as they all tend to plugin between 5PM and 6PM every day, causing intolerable drain shocks to the grid. The electrical grid was never meant to be a wattage replacement for the gasoline and diesel delivery system.

      100m people plugging in their Teslas at 5:25PM is going to cause brownouts and blackouts all over.

      1. PeterA Avatar
        PeterA

        It’s called a schedule.

        When I get home I plug my car in at 6pm. But guess what! I’m not dumb! I don’t charge at peak, I have it wait till my electricity providers off-peak period.
        Why pay 15 dollars to charge a Tesla to full when you could pay 7?

        Which is -you guessed it- cheaper and when electricity demand is much lower due to industry being offline. If it becomes a problem then providers can change the peak times or enable the use of controlled loads, which is a centrally managed off peak system controller by the provider.

        All of this is already enabled if you have smart metering installed. Which I should have been installed several years ago in most places

        It’s almost like you’re spreading hullshit given the reality of things!

        (I have a Volt, not a Tesla fwiw)

        1. Goldie444 Avatar
          Goldie444

          Exactly PeterA, and if I charge during the day, my solar panels are doing the work.

      2. Justin Avatar
        Justin

        Speaking from all your inexperience. You just got served.

      3. Mike Shackleton Avatar
        Mike Shackleton

        Some people will have access to charging facilities during the day, some won’t need to charge at all. It’s not a stretch to foresee the car’s battery and charging management system scheduling your recharge in concert with everyone else’s needs so as to smooth out the demand from 6pm to 6 am. Most people won’t need to charge for the whole night.

      4. Nick Kemp Avatar
        Nick Kemp

        I think you mis-spelled moron when you tried to create a username

    2. Barri Mundee Avatar
      Barri Mundee

      And may the batteries go forth and multiply!

  5. Craig Allen Avatar
    Craig Allen

    So is it profitable enough that it would make financial sense for Neon to set about enlarging it and for others to do the same. If not, how much cheaper to batteries have to get?

    1. neroden Avatar
      neroden

      Neoen has already commissioned another one (see this website) and has stated that they look forward to producing more along with Tesla

    2. John Smith Avatar
      John Smith

      They can likely get better results by having other batteries in other locations, instead of increasing the size of this one. Being distributed helps a lot unless your transmission/distribution system is amazing.

  6. Chris Jones Avatar
    Chris Jones

    Great that the battery is performing as planned, and better. However I note the NEM-watch widget has a feature for pumped hydro on the demand side, but despite my obsessive monitoring, I have never seen either Tumut 3 (NSW) or Splityard creek (Qld) registering any pumping on the widget. Are they not reporting or are they just not pumping and releasing?

    1. Paul McArdle Avatar

      Does not happen every day, but when it does will typically happen early in the morning (remember that’s when electricity is cheapest, these things run by “buying low, selling high”).

      See the time-of-day average for Nov/Dec/Jan from NEM-Review v6 here: https://uploads.disquscdn.com/images/d5d5167083e0423737877d29efa680786df043751f6179200561371274439c27.png

      Hence you need to watch the widget in the middle of the night to see anything there…

      1. Chris Jones Avatar
        Chris Jones

        Thanks Paul! I’ll have to set my alarm 🙂

    2. David Osmond Avatar
      David Osmond

      You can see the splityard ck pumping at the following links (untick everything except Pump1):

      http://anero.id/energy/hydro-energy/2017/december
      http://anero.id/energy/hydro-energy/2018/january

      1. Garth Power Avatar
        Garth Power

        The problem with Anero is its half baked, they have only one of the two Wivenhoe pumps and none of the others, they have Hornsdale Power Reserve Unit 1 generator but not the load.

    3. Garth Power Avatar
      Garth Power

      Wivenhoe/Splityard draw a little in the past few days so they will look to top it up at some stage but it does not run very often. Tumut 3 also gets fed from upper Tummut and its running atm so unlikely they will pump.

      Shoalhaven Scheme gets a bit of a run, its a ~240 MW capacity and was running yesterday pumping last night.

      1. David Osmond Avatar
        David Osmond

        you can also see the last few days at the following

        http://nem.mwheeler.org/stations#PUMP1

    4. Ray Miller Avatar
      Ray Miller

      Or you can go to the AEMO online site (price and demand) and look at Queensland, they have been pumping with about 240MW at the lowest demand part of the day around 6am Queensland time, especially for a hot day coming. http://www.aemo.com.au/Electricity/National-Electricity-Market-NEM/Data-dashboard#nem-dispatch-overview

  7. orko138 Avatar
    orko138

    What specific market rules need to be ‘adjusted to catch up with the technology’??

    1. Garth Power Avatar
      Garth Power

      Depend on who you speak to, its very complicated.

      1. orko138 Avatar
        orko138

        Dont mean to be rude, but ‘its complicated’ is a bit of a cop out answer.

        Seriously, what rules need to change??

        1. John Smith Avatar
          John Smith

          Biggest one is the settlement period. But, in addition, allow the battery to be compensated for Fast Frequency Response as an alternative to synchronous inertia.

          1. orko138 Avatar
            orko138

            FFR is a good example. And yes, it does need more clarity about how this service is provisioned. Its far from clear whether this should be a paid service, as it could be a mandatory requirement for connection – many new wind operators can provide this service too. Further, it is still not clear whether we actually need a change in rules to provide this service as AEMO could potentially manage this service.

        2. Garth Power Avatar
          Garth Power

          Lets not worry about what needs changing and start from a fresh piece of paper for the set of rules. You need to do 100 things to get the grid/power production to a certain place and its like 50 on the list. What takes place on the top will determine the new set of rules.

          1. orko138 Avatar
            orko138

            As nice as that would be, that is not going to happen. At least, not any time soon.

    2. neroden Avatar
      neroden

      Well, the settlement period for starters.

      1. orko138 Avatar
        orko138

        Batteries already in market, more coming. Yes 5 min settlement will improve this, but it’s not make or break. Tesla/Neoen made a cool $mil in 2 days according to Giles. What other rules need to change to catch up??

  8. PaulC Avatar
    PaulC

    The question is now whether this is succeeding in addressing the gaming of the FCAS market? [Gaming described in exquisite detail by Giles yesterday: “When generators jack up prices 100-fold…” ]

    And that question is not just applicable right now, but also if and when the market players adapt their strategies to the new entrant? [I can’t imagine the existing players are too happy about a new player backed in some form by a state government.]

    What amazes me is the total lack of transparency in these markets. If you actually published the details of market operation such as the proportion of costs from market manipulation including FCAS gaming and rorting around poles & wires on peoples’ bills , I think there would be a mob with pitchforks outside parliament and gen-tailer offices!

    That would perhaps stop some of the “alternative facts” circulated about RE being intermittent and expensive, and the cause of high bills. It’s all too reminiscent of Manus/Nauru – let’s just hide the truth, block independent scrutiny, and promote some convenient lies instead.

    1. Garth Power Avatar
      Garth Power

      Tas Hydro makes a lot from this market, half the problem is when a generator output is too high or low (depending on if its supply raise or lower) it can’t be enables as its run out of running room.

      Take Vic market for example, very little FCAS in the raise from the coal as they are running at max so they can’t raise any further. And everything has a turn down.

      The joke part is the Contingency FCAS, it should just go with the rules stating that the generators must support this anyway. If you read my post below the battery looks like it will take in $9Million PA just from contingency which dispatches so infrequent and if anything does dispatch its normally the Fast Raise or Lower (6 sec). The slow and delayed (60 sec and 5 min) rarely if ever get utilised as the fast will cover the gap and if the slow or delayed does, the grid is in very bad shape and possible close to a blackout due to generators tripping.

      1. Mike Shackleton Avatar
        Mike Shackleton

        That’s interesting, I would have thought that the HVDC link between Tassie and the mainland would have made providing FCAS to the rest of the country impossible.

        1. Garth Power Avatar
          Garth Power

          The FCAS was part of the design criteria. The other DC do not have the same capability. You have to also remember that Basslink itself is basically a registered generator and load (depending on direction of operation) and sells at the Vic price and as its contracted to Hydro Tas they receive the Vic price at the DC converter station at the back of Loy Yang power station.

  9. Ray Miller Avatar
    Ray Miller

    Now just think what would happen using a substation battery storage, you can then add more grid services to the existing and growing list.
    Then seeing how modular the system is just replicating the successful design decreases the costs.

    1. BushAxe Avatar
      BushAxe

      It’ll be the next bandwagon the networks will throw into their expenditure plans to justify increasing network fees.

      1. Ray Miller Avatar
        Ray Miller

        I think you missed the point, its about reducing costs, especially capital expenditure marking more use of existing assets. Yes the management of the network assets has an extremely poor record adding significant costs, we need a high level of transparency.
        With the Lakeland PV and battery storage for North Queensland (when it is commissioned inland of Cooktown) will be closely watched for all those value added services especially in a remote area and substation (as it is located next to it).

  10. Cooma Doug Avatar
    Cooma Doug

    This article is really great to see.
    The graphs are making me very optimistic now about the effect of these innovations on the industry.
    Ive been thinking for a while now how electric cars could function this way whilst on the grid. Also home and industry solar plus storage sites.

  11. Tom Avatar
    Tom

    The only problem with the “making money” bit from wholesale arbitrage is that the more batteries there are, the more their combined output will depress the price spikes, hence the less money they will all make.

    This may be a barrier to new entrants investing.

    1. vielepunkte Avatar
      vielepunkte

      true, but I think its rather that while batteries are expensive, the marked provides subsidaries until they no longer needed.

    2. BushAxe Avatar
      BushAxe

      Ultimately the market will decide as battery storage becomes cheaper as to whether storage or peaking generation is more cost effective. I think most can see where it’s heading.

      1. Tom Avatar
        Tom

        I hope we’re right about where it’s heading.

        I can foresee that the primary function of storage in the future will be to increase transmission line capacity factor to relatively remote wind and solar sites – wind in particular. Transmission lines make up half our energy bill after all.

        For example, there might be a 300MW wind plant in remote NSW connected by a 300MW transmission line. At an optimistic capacity factor of 30%, it will only supply an average 90MW.

        If you built 600MW of wind capacity connected by a 300MW transmission line, you would not lose half of your energy. You would only lose half your energy on the occasions that your wind turbines were generating at maximum capacity, which is not very often.

        When they are generating at 10% or 30% or 50% capacity (most of the time), you would not lose any energy, and even if they were generating at 66% capacity you would only be losing 25% of the energy that you are producing at that time.

        Therefore a small amount of storage at the site of generation would be all that is needed to capture a lot of your excess power that is surplus to the transmission line capacity, smoothing your effective output (making it more “baseload”?) for very little extra expense.

        This is completely the opposite of the big, centralised Snowy 2.0 pumped hydro storage, which is going to require massive transmission investment which will be largely underutilised.

    3. Cooma Doug Avatar
      Cooma Doug

      Yes indeed…that is how markets work. The other factor is the cost of assets and that is falling quickly.
      Rapid response energy, negative and positive will increase the asset value
      of solar and wind on the market as the costs are falling. It will flatten the profile via the load shifting.

  12. neroden Avatar
    neroden

    $500,000/day in profit, $50 million cost. That’s 100 days to break even. My God. This is money printing.

    1. Ren Stimpy Avatar
      Ren Stimpy

      A simple lack of competition due to entrenched systemic economic in-breeding. Whoever the Charles Darwin of economies was we need to learn from him now – we need more competition! If the Libs can’t supply it because of their conflicts of interest then they need a good solid boot ASAP and let the other mob have a go.

      1. earl colby pottinger Avatar
        earl colby pottinger

        In a years’ time it could be making more money than it costed.

        At that point all the other power companies are going to ask why they don’t have one. Worse, the power generators may find themselves competing against pure battery owners who only care about making money without gaming the power system.

    2. Andy Saunders Avatar
      Andy Saunders

      I’d be careful about taking one somewhat exceptional day and multiplying it out – the average day isn’t like that.

      1. André Balsa Avatar
        André Balsa

        Of course, but a 1% return on a 20-year investment in a single day is still impressive anyway you want to look at it.

        1. Andy Saunders Avatar
          Andy Saunders

          Oh, of course, and good luck to them (and of course that’s only the portion from the wholesale energy market – there’s other revenue coming in too).

          But those days don’t happen every day. And if the facility is super-profitable (return above the cost of capital) which you and I strongly suspect it will be) then there’ll be quite a few players piling in, which will degrade the revenue potential (the spikes won’t be so often or so extreme – which is a good thing).

          1. André Balsa Avatar
            André Balsa

            That’s just the market reaching equilibrium and when it does so, everyone will be the better for it, I guess.

  13. Cooma Doug Avatar
    Cooma Doug

    The ability to go from charge to discharge in the blink of an eye is a new product and I dont think they have worked out the full potential yet. When there are a dozen of these on the grid the value of everything on the grid will be significantly changed.

    I cant stand those political words “when the wind/sun dont blow/shine”. Thankfully it will be getting harder to say.

    1. john Avatar
      john

      As you know putting in more of these and further more storage will just level out the supply and kill the 3 days making money business model in place at present.
      On the storage story we honestly need to do this as soon as.

    2. fehowarth Avatar
      fehowarth

      Big country. I am sure the wind will be blowing, sun shining somewhere of the grid.

      1. Brian Tehan Avatar
        Brian Tehan

        If WA had more large scale solar and you had a HVDC connection to the eastern states, that would solve the evening peak. Then again, east coast storage might be cheaper.

        1. Andy Saunders Avatar
          Andy Saunders

          Well, it may solve the evening peak in the eastern states. Won’t help so much in WA!

      2. Joe Avatar
        Joe

        The COALition have their heads buried so far underground that they wouldn’t need whether it was sunny, windy or day, night. And of course you are right that the sun is shining and the wind is blowing somewhere across Australia and a truly National Grid will tap that.

      3. Andy Saunders Avatar
        Andy Saunders

        There is a certain amount of non-correlation of solar and wind resource. But there are also periods where the correlation is high eg winter highs over the continent causing low wind generation virtually everywhere.

        1. itdoesntaddup Avatar
          itdoesntaddup

          Non-correlation is not good enough. You need anti correlation. Think of it this way – if output is perfectly correlated, you get a nice sloping line comparing the output of one producer with another. If it is random, then the correlation plot is smeared all over the square, and you have significant probabilities that both producers will have simultaneous low output or simultaneous high output: in energy systems that means you still either need 100% backup via storage or alternative, dispatchable sources. With perfect anticorrelation, you guarantee constant average output, as lower output from one producer is offset by higher output from the other.

    3. hydrophilia Avatar
      hydrophilia

      Yes, the temptation is to suggest something about “where the sun does not shine”.

  14. john Avatar
    john

    BUT but the Energy Minster said only today that the Tesla battery is not of any use like he said it can not carry the whole of Australia for an hour.
    Well that would work for idiots who listen to him.

    1. Joe Avatar
      Joe

      John, rule No, 1,,….just ignore anything that the Joshie says. When he isn’t making things up he just tells lies.

  15. itdoesntaddup Avatar
    itdoesntaddup

    I’ve been tracking the performance of the battery going back to November 30th, using 5 minute data from the excellent nemlog.com.au site. Of course, 5 minute data have gaps of 4 mins and 59.999 seconds (they appear to be instantaneous readings, not averages of the 5 minute period). I’ve found it fascinating to see the process. Here’s the first 8 days:

    https://uploads.disquscdn.com/images/802a93bca438eb4cf753df815bbeb6b54a79686a570cdaf2b10d73126ce41806.png

    The orange MWh line is simply cumulating the net charge/own use and discharge data, assuming that the battery does the same thing for 5 minutes. The RRP is the regional price for South Australia at the Torrens Island 60kV bus. Notable is the discharge at 70MW on November 30th – famously called for to help handle a hot day when supply was tight. Price volatility that preceded the discharge disappeared afterwards: the unkind might claim that they waited until the market tightness was due to dissipate before discharging, but I’d score this a clear win for the battery.

    The orange line reaches a maximum that appears to be close to the 129MWh battery capacity during this period: however, that does not take account of the steady draw that takes place (just visible on the X axis, but more easily seen in the shallow upslope segments of the orange line) – I assume that this load runs the air conditioning to keep the equipment cool. It runs at 200kW. There’s a deep discharge test on the 4th December of about 90MWh. Charging was limited to 30MW throughout. It’s also obvious (and unsurprising) that there was no attempt to operate commercially in this period, with price spikes being ignored.

    Overall, the battery supplied 236.6MWh and consumed 345.6MWh, for a net charge plus consumption of 109MWh. 8 days of 200kW for A/C would account for 38.4MWh.

    1. Ren Stimpy Avatar
      Ren Stimpy

      For sure, appreciate your observations, and will endeavour to try to keep up.

      1. itdoesntaddup Avatar
        itdoesntaddup

        200kW seems eminently reasonable for A/C. There are 108 battery containers, and 36 containers for inverter electronics if you look at a picture of the project:

        https://brandsanews.com.au/wp-content/uploads/2017/12/Blog770px-Hornsdale-Battery2_.jpg

        The question is whether that is enough when it gets really hot – a subject that is worth looking at in the light of the recent heat wave.

  16. itdoesntaddup Avatar
    itdoesntaddup

    Here’s the next 8 days:

    https://uploads.disquscdn.com/images/643fbe622c0265b72b7b36d69819ea6010cbe6c9140a6c555f5efec27206f750.png

    The major events in this period are testing of the battery performance envelope: the first 100MW discharges (proving the battery is up to spec in that regard), and charging at up to 80MW. There’s also a deep discharge (apparently 136MWh, beating the spec) and recharge (perhaps worryingly consuming 162MWh – implying only 84% round trip efficiency) on the 14th.

    Of course, there was also the event that attracted all the attention at 1:59a.m. on the 14th: the Loy Yang trip which saw the battery halt its low level charging to supply 7MW for a few minutes, before going back to what it was doing Completely unnoticeable among the other actions of the battery in this period: the only hint is the price spike.

    Over the period supply was 471MWh while the battery consumed 640MWh, for net consumption/use of 169MWh.

    1. Ren Stimpy Avatar
      Ren Stimpy

      Actually going by those numbers net consumption/use was 471MWh = 74% utilisation of energy @ 471MWh / 640MWh. Which is pretty good for storage of energy.

      1. itdoesntaddup Avatar
        itdoesntaddup

        Well, we’re not adjusting for any net charge or discharge over the period, but even so, 74% is more like what you get from pumped hydro, and rather less than what you get when you don’t pump the hydro, but replace it with wind or solar when available, and use the hydro only when it isn’t. It’s also some way below the sort of efficiency claimed for a Tesla Powerwall plus inverter, which is 85/88%. Inverters do tend to be less efficient at low power outputs. At grid scale, I’d be looking for 97% round trip efficiency through the inverters, and 95% for the batteries, so certainly over 90% for the system. Perhaps the batteries are still bedding in and will attain optimal performance after a few more cycles.

        1. David Osmond Avatar
          David Osmond

          Stats from Nov 24 to Jan 21 indicate 4,240 MWh discharge, 5,215 MWh of charge, implying a 2-way efficiency of 81%

          1. itdoesntaddup Avatar
            itdoesntaddup

            Thanks for adding it all up… I have provided numbers for each chart I have displayed here, and was going to do the sum. Overall, I’d rate that as a disappointing performance, but perhaps influenced by testing and initial charging. I shall be watching to see if it improves.

            Another factor is that the overall capacity factor is low, with less than 5MW of supply to the grid. However, I would expect that battery management would entail rotating the use of individual batteries and inverters, so that the inverters were used in their more efficient high throughput regime, and the batteries are all regularly cycled so that they get even use. With 36 inverters, each can supply about 3MW – there is no need to run all of them at 5% which would probably be rather more inefficient.

  17. itdoesntaddup Avatar
    itdoesntaddup

    The next chart covers 16-29th December, when the battery operation seemed to settle into a more regular mode of operation after some further testing of 100MW discharge on the 16th. However, its charges and discharges (which were confined to +/-30MW) seemed to follow an irregular pattern , so I wondered if it was acting to balance out variations in output from the Hornsdale Wind Farm, while now has all three phases operational. The wind output over this period shows substantial variation between periods of steady output at full capacity, and lulls in the wind with zero power output (even small net consumption from the grid presumably to keep the rotor shafts from sagging by turning them slowly), and gusty performance. There are also some fairly sharp ramp rates. Here’s the battery layered on top of the wind farm:

    https://uploads.disquscdn.com/images/d65ddaa1c30c0d75b3bd066d5c92e3ce51754b0739d21fa49e65d199ef241fc1.png

    There is no obvious correlation with the wind farm at all: indeed, the battery seems to add volatility to the output when the wind farm is operating at capacity. There is correlation between low wind farm output and volatile/high prices, and also the battery managing to be somewhat more commercial in its operation.

    I’d be interested in insights as to what the battery was really doing in this period, and why its behaviour was so intermittent, rather than the more solid block, constant charge/discharge that it had been operating before.

    Stats on the batts: total charge/use was 1,812MWh (5.39MW on average), and supply 1,578MWh (4.70MW on average) for net consumption of 234MWh. Average value of supply: A$102.90/MWh at RRP; average cost of charge/use: A$75.45/MWh (time average of RRP A$82.16/MWh; average value of wind farm output: A$70.38/MWh). Net cash generated by battery: A$25,634.

  18. itdoesntaddup Avatar
    itdoesntaddup

    Having failed to find a correlation with wind farm output, for the next period I took a look at the battery alongside the flows on the Heywood interconnector.
    https://uploads.disquscdn.com/images/7557323efa6f050010a775fcc5c9c850465cdd44a6da5668a907027d349f6715.png

    Throughout, the battery flows were limited within +/-30MW, and the battery broadly operated in timeshift cycles, some with amplitudes of around 100MWh. Towards the end of the period we see export to Victoria on the Heywood interconnector reach a constrained 500MW, driving price in SA to minus A$1,000. The battery did take some advantage, but not nearly as much as it could have done: first time around, it only sucked up about 46MWh (including spending some time inexplicably discharging), with charge rate limited to 24MW much of the time. Why it did not charge up at 80MW until it was full is a mystery. Also, the event was fairly predictable, since it was a consequence of surplus wind and solar output hitting the export constraint on the connector: the battery should have ensured it was empty beforehand.

    Stats on the batts: 1,590MWh of supply, 2,013MWh of own use/charging for net 423MWh of charge/use. Cash generation of A$69,479.

  19. itdoesntaddup Avatar
    itdoesntaddup

    With all the excitement over the A$14,000/MWh prices, I took a look at the battery and the Heywood connector once again – but putting the battery on the same scale as the interconnector:

    https://uploads.disquscdn.com/images/b812434b7942a64b6e9e299278e763981c222b984de46f78faef8b694950a06b.png

    The price differences between SA and Victoria go off the scale, and merely serve to highlight the period when supply was tight. Once again, the battery persists in operating in the narrow +/-30MW charge/discharge envelope. We switch from feast to famine at noon on the 16th, with the interconnector supply to SA over 20GWh net over the remaining period – which would have required about 160 Hornsdale sized batteries already stocked up to match. Constraints on interconnector throughput are easily seen, pulling pricing away from any arbitrage margin, and showing flat tops at the export limit. Note that there were lower limits of 250MW on the 18th/19th at times – presumably a loss of a circuit on the interconnector. Over the period, the stats on the batts are 585MWh of charge/own use, 449MWh of supply, net consumption of 136MWh – and cash generation of A$722,647.

    That’s a very hand injection, but it depended on a combination of circumstances – soaring demand in hot weather, constraints on the interconnectors to make good the lack of local supply from wind farms, and a major generation outage in the middle of it all with inadequate backup available since the closure of the Heywood coal plant. Not sure that Australians want to see sky high prices every week or even every month just to remunerate the battery. It needs to find adequate sources of revenue without these incidents.

  20. John Burnett Avatar
    John Burnett

    What no congratulatory speeches from the federal Coalies? “No I was incompetently wrong” comments from Captain Blackout (Frydenberg)? Or the Potato or Fizza or most rediculous treasurer in Australian history ScoMo? Can’t believe how this inept lot continue to fall over themselves in an attemp to get an award for most idiotic politician ever.

  21. Interested Observer Avatar
    Interested Observer

    “The Tesla battery, formally known as the Hornsdale Power Reserve, is perhaps the biggest development in an industry that has seen little real change in nearly a century”.

    A century ago we had small inefficient coal fired power stations located in cities with coal shipped to the power stations. Since then we moved to large efficient coal fired power stations located close to coal mines and the development of high voltage transmission systems, the development of a natural gas industry and gas fired plants, the rapid development of gas turbines and combined cycle technology and the rapid development of wind turbines and solar technology – yet the Tesla battery is apparently the biggest development in an industry that has seen little real change.

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