Network polemics: Why ENA is wrong on battery storage costs

John Bradley wrote an emphatic response to my prior article comparing the cost of installing a battery in Queensland with Ergon’s claimed costs per kilowatt of network capacity.

For those wishing to get a sense of things without dipping into the technicalities, the vehemence of Mr Bradley’s response to my note rather gives it away. As the Player Queen representing Hamlet’s mother says, “the lady doth protest too much, methinks”.

Mr Bradley’s main point seemed to be that I had taken the present value of Ergon’s network charge in perpetuity and compared it to the capital outlay for a battery that he said will last for only 10-15 years.

Is this a valid criticism? I don’t think so, but lets first check whether my conclusion would change if we compared the cost of a battery to network connection for 15 years. And lets just look at Ergon’s eastern region where network charges are 2.5 times lower than Ergon’s western region.

tesla storageIn this case, 15 years of network access would cost, as a present value, $2,500 per kilowatt. This is still about 10% more per kilowatt than AGL will sell a battery to you today. And, as we noted, Tesla’s batteries are expected to sell for around half the installed cost of batteries today. So, no Mr Bradley, my conclusion remains unchanged.

But it seems reasonable to consider more pessimistic assumptions about network costs. The analysis so far assumes no increase in Ergon’s costs in future. Yet Ergon has a well established track record of continually increasing costs in real terms.

Even if Ergon is able to get a grip on their expenditure, the prospect of continuing declines in network usage will mean prices rise even if expenditure does not. Even if we assumed just a 1% annual increase in prices, the present cost of connection to Ergon’s network rises to $2,700 per kilowatt for 15 years.

Of course a full pairwise comparison of grid v off-grid needs to consider many factors including access to solar and the cost of buying electricity (excluding network) from the grid.

Here too the picture is strongly in favour of distributed generation. But there is no need to take a categorical line here: in some cases disconnection is likely to be more encouraging and for others less so.

But there can be no reasonable doubt that networks are under pressure. The ENA knows this full well and for this reason is pushing regulators to reduce asset lives, reflecting their response to this asset stranding risk.

Indeed Mr Bradley writes admiringly of Ergon’s investment in batteries in place of network. So he would have us believe that if a regulated monopoly spends its customers’ money on batteries that’s wise, but if customers do the same that’s unwise. Draw your own conclusions.

 

 

 

Comments

15 responses to “Network polemics: Why ENA is wrong on battery storage costs”

  1. Terry Leach Avatar
    Terry Leach

    Actually the quote was Hamlet’s mother talking about the Player Queen. But that’s by the by. The insult to the English language is ‘2.5 times lower’. Surely you mean ‘40% of’ or ‘60% lower’? The author makes a good point but his knowledge of literature and grammar leaves a lot to be desired.

    1. MaxG Avatar
      MaxG

      🙂 Looks like you are a more mature citizen 🙂 Well, I reckon we had the best education, our kids and grand kids will never experience. So, what I am trying to say (very friendly and smiling): you need to adjust your expectations 🙂

    2. Bruce Mountain Avatar
      Bruce Mountain

      Terry

      I am suitably chastised. On 2.5, yes 60% reduction.

    3. John Herbst Avatar
      John Herbst

      Technically, it could also mean that eastern prices are 28.57% of western prices, which would make one price ‘2.5 times higher’ than the other. You’ve assumed he meant to say that one price is 2.5 times the other. It’s not inconceivable that he meant it this way.

  2. Peter Grant Avatar
    Peter Grant

    Incisive as always Bruce. This is a time bomb for the networks. What is so surreal in all this is the Nationals stridency against renewables. Arguably, given the high grid cost regional and rural customers could demand high capital subsidy to disconnect and so defray the ongoing cross subsidy embedded in network tariffs.

  3. Ian Avatar
    Ian

    According to the article in the 5th paragraph network access per KW for 15 years is $2500 and the battery about $2300 for a KW lasting the same time period. Is that a battery with a small KWH rating able to discharge one KW quickly in a few minutes or is it one with a high storage capacity of many KWH with a slow discharge rate of only 1 KW? I think he means a battery with a nominal rating costing $2300 per KWH. The battery’s function is to store home generated electricity, usually solar, when it is produced, ie in the day and then discharged at night. this means the battery only has 1 KWH available for use in a day( best case scenario 100% discharge with a battery life of 15 years or 5400 cycles. ‘must be a Kryptonite battery to have this good a spec.)

    The network access has a similar cost, he says, but their KW rating is exactly that: 1KW for as long as you need it, in other words up to 24 KWH in the day.

    The network, according to the Author’s of the article own figures, is up to 24 times cheaper than the battery.

    By the way, respectfully, not to rub it in, KW is a unit of power, that is similar to horse power, how many horse- equivalents do you need to pull a carriage?. KWH is a unit of energy, how much grain do I need to feed those horses to pull that carriage for an hour? A battery is generally rated by its KWH or more specifically Amp Hours for a given voltage. A car battery may be able to crank out several KW of power over a very short time, but it won’t store very much energy. A huge lithium battery may store many KWH but be rated at releasing this energy over 20 or 100 hours.

    One has to get a grip on electrical units or one may be tempted to buy an electric Chery with a fantastically rated motor but with a tiny battery! Able to accelerate up the steepest hills but unable to take you anywhere. To reiterate the horse theme, KW is what gets you out the stalls and KWH is what gets you around the track.

    #%£@$ I know this and I haven’t done physics for at least 20 years! We can’t really get to the meat of the problem if we are still choking on the milk.

    1. Terry Leach Avatar
      Terry Leach

      Networks provide capacity (KW), generators provide energy (KWH). The author seems to have his units correct, and his prices seem to be around the mark. Tesla Powerwalls will be selling for around $1500 AUS per KW, 5 year life, and prices are expected to drop dramatically. The cost is about half that per KWH of storage. And while we’re being pedantic, that should be ‘for a short period of time’, not ‘over’. And 24 times cheaper? Oh dear. It’s been a long time since you studied grammar.

      1. Ian Avatar
        Ian

        Thanks for your reply. My point is this: in the article the cost of the battery is 10 % less than the network capacity of 1 KW both compared over a similar time period. 10% difference is not a lot in this estimation. Now the network is able to provide 24 KWH in a day for each 1KW capacity, whereas a battery will store 1 kWH in a charge / discharge cycle. In other words in a day a battery can provide up to 1KWH of energy per 1KWH rating and the network 24 KWH for its rating of 1 KW. 1:24 . The network is cheaper by a factor of 24 ie it is 24 times as cheap. Or if you so desperately want correct grammar , the battery, according to the article’s figures, is 24 times as expensive as the network in usable energy. I think you were trying to paint over this very nasty fact by trying to correct my English. Like you , I don’t care for the networks, and I would love to have a home battery system to store my solar generated electricity but at the moment the network costs are still cheaper than battery storage. This may not be the case in a couple of years. Thank you for the Tesla powerwall figures @ 1 cycle a day 5 years =1825 cycles. From my reading lithium batteries should last longer than that, but then again they do need to be nursed with a DOD of < 80 %. Neglecting all that : $1500/1825 = 82 cents/KWH per charge /discharge cycle. My network charges a network fee of $1.50 a day plus a KWH fee of 22c. A Tesla Battery would be cheaper if night consumption was less than about 2KWH and progressively more expensive at higher consumption levels. There may be other costs involved in installing a Tesla powerwall which would push up its overall cost of installation and make it less viable, pity if that 's the case.

        By the way, Mr Guardian of and Final Authority in English Grammar, Mr Can't-see-the-meaning-for-the-grammar Guy, the original article quoted $189 per KW per year cost to provide network capacity to a house. For a day that would be 50c. The question is: How much capacity would the network need to provide the average house cover for their needs, given that the pattern of consumption is one of peaks and troughs, but the average daily consumption is not much more than 24KWH? Is Ergon's $1.50 connection fee,which is three times the average capacity required, justified? What does the $189 /KWyear cost actually include? Is it just hardware namely the power stations and poles and wires or does it include the administration and consumables? I take it that the total cost is $189/KWyear to supply 1KW continuously 24/7/365 which is 2c/KWH. Peaks and troughs of consumption will make this figure higher obviously as a higher capacity is needed to cover the peaks, and less revenue comes in during a nadir of consumption. I should imagine peak capacity would be 3 or 4 times average consumption for a non solar house and many times more for a solar house. Why I ask these questions is to try understand costs involved in generating power and to find out if other forms of storage on the electricity grid would be viable, such as wind, which often has its peak in the evening , and hydro, in the form of once through or as pumped storage. FF has to go, that's a given, but how cheap and how viable would be wind and hydro? Would the Ideal of 100% renewables in all its distributed generation and storage forms be cheaper than the current FF system? The FF system is incredibly cheap if it runs at capacity constantly, belting out the KWH,but it relies on a steady load to soak up the energy generated. It's not so cheap when the load is variable, especially when rooftop solar greatly increases the variability. That is probably why the utilities fight so vehemently against solar. With the changing landscape in energy consumption, has the FF lost its competitive edge to such an extent that large scale storage becomes viable? Judging from the various champions of renewables and of FF sparring off and taking sides the answers are not really known.

        1. Ian Avatar
          Ian

          Excuse the verbosity, but it’s worth analysing this problem of network costs a little more. If $ 189 is the network cost per KWyear to deliver electrical energy to a household’s meter, and if the peak rate of consumption is 5 KW , then the networks cost to deliver that capacity is $189 x 5= $945. For the sake of the argument, any KWH not consumed is wasted. The average household consumes about 30 KWH in a day, which if they are not on solar, would translate to 10950 KWH a year, to deliver this amount would cost the network 945/10950= 8.6c/KWH. Another house with the same consumption produces and uses say 2/3 of this power, but still needs the same peak capacity from the network. 10950 x1/3 drawn from the network. Cost to the network is now 26c/KWH. Millions of identical dwellings with solar would cement this increased cost to the network, other consumers with different peaks and troughs would reduce this effect. Throw a battery into the mix and see what happens. A network battery installed near the point of consumption, in the home or in the suburb could smooth the peaks and troughs so that in effect peak consumption is reduced to average consumption. for a non solar suburb 2c/KWH, not including the battery. A 5KW peak would probably not last for more than 1/2 hr, say 3KWH battery per household, battery last 10 years ,costs $2300/KWH: $1.90 per day to save 6.6 c/KWH or 6.6x 30 = $2.00 per day. In a solar neighbourhood two factors come into play 1 similar peak of 5 KWH lasting 1/2 hr, 2. draw of 10KWH over 16 hours would require about 3.3 kWH battery to smooth the flow. So to save 26-2c = 24c/KWH for 10 KWH =$2.40 per day would cost the network $2.10 per day. That’s not really worth while. Bring the battery price down to $1000/KWH and suddenly the economics would favour network battery storage at the house or suburb level. Have the suburb placed in a distant rural area and the battery storage becomes very much viable.

          This line of thinking is exciting stuff. The cost of rural electrification is far more expensive then conglomerated areas in the cities. We are told the cost to the network for rural capacity is 2.5 times that in town. Even the most expensive lithium batteries appear to be viable alternatives to long distance transmission lines. Australian designed and manufactured large capacity flow batteries at reasonable cost together with solar could supply rural towns very cheaply . Rural properties could be given incentives to construct their own mini grids. Divesting the rural electricity grid could make city electricity far cheaper to supply and the knock-on effect for the networks would be improved viability of electric vehicles.

          Perhaps others could rip this analysis to shreds and shed some more light on the subject.

          1. Terry Leach Avatar
            Terry Leach

            You seem wedded to calculating network costs in terms of kilowatt hours, which is not correct. A network has a fixed cost. That network capacity is measured in KW, MW or GW, depending on size. Forget all about the time period and the amount of energy consumed; they are irrelevant in calculating the cost. You cannot validly calculate the cost of the network in cents per KWH.

            Now the price is something else altogether. How to pay for that fixed cost? Do you spread it out evenly over every KWH consumed, as you have done? Do you charge more during peak periods to shift the load and therefore reduce the need to increase the capacity and therefore the cost? Do you charge everybody a fixed fee? Do you charge everybody a fixed fee based on their peak consumption? Do you charge everybody a fixed fee based on their peak consumption during peak periods? The answer is all of the above, somewhere in Australia.

            This is an argument that is raging at the moment, particularly in regard to solar PV owners who use the network less, still contribute significantly (at the moment) to the network peak, but provide a public good in providing clean energy and putting downwards pressure on energy prices by adding to supply.

            And then you add in distributed storage when it becomes significant, which will see solar owners dramatically reduce their network utilisation, particularly in peak periods. It’s going to be a spirited debate.

            Separate from the network costs is the cost of generation, which also needs to be able to provide peak power and these costs are not fixed. The wholesale prices skyrockets during peak periods, over $13 per KWH, but can turn negative if there’s too much sun, wind and they can’t dial the coal down quickly enough.

            Your electricity retailer needs to take all of the above when setting retail tariffs.

            Now when you think about a stand alone solar storage system, you need a battery system that can provide peak power, and store enough energy for when the sun isn’t shining. Currently systems are custom built to meet both, and are very expensive. But in the future you might buy two Tesla Powerwalls to provide peak power capacity, even if you only need the storage of one. Or vice versa.

            And you’re right, it could be cheaper to subsidise remote areas to go off grid, rather than subsidise their grid access. But the real argument is going to be about the grid in cities, if solar storage takes off and there are many customers utilising the grid less. Increase the fixed price of network access and you encourage grid defection, bringing on the death spiral. If batteries have the same cost curve as solar PV and smart phones, we’re in for interesting times.

  4. James Hilden-Minton Avatar
    James Hilden-Minton

    What we are trying to do here is put an upper bound on the economic value of a grid connection. It is all well and good if a utility can account for the cost of a service, but what is the value of the thing. We can bound economic value by looking at substitutes. So batteries are one substitute which is rapidly becoming cheaper than a grid connection.

    At $2500/kW (over 15 years), suppose you are only interested in backup power. A home generator is just a fraction of this cost on a capacity basis. The value of PV and batteries is that it minimizes the amount of fuel one consumes in the backup generator. They both minimize the annual cost per kWh of an off grid system, while the generator provides cheap emergency backup.

    So the battery enables cheap solar to be used more widely. But the battery also enables the generator to be used quite efficiently. If you need less than 30kWh/day and have sufficient kW in your batteries for peak power use, then you only need a generator with about 1.5kW max power. That is, on a day with no sun, you need a generator just large enough to meet your daily energy need, not your peak demand. So this is a small generator with little front end cost.

    So in the final analysis once solar+storage brings the average energy cost well below the utility price, the grid connection is only worth a small backup generator that is used sparingly.

    Utilities will need to come up with a more compelling value proposition than “you might need us on a rainy day.” I believe the utilities will need to offer energy trading as an economic incentive for PV owners not to leave. If the solar or battery owner can trade power with the utility at competitive prices then value is being created for both DER owner and utility. But if utilities are not willing to capture value through trade, then their backup service will be worth no more than a cheap generator.

    1. Phil Avatar
      Phil

      That’s so true , imagine if the genset technology were improved with an affordable quiet auto start high efficiency LPG run generator that could augment the peak supply as well as recharge the batteries for the same money as JUST the peak electricity spend (and including all maintenance costs) .Even using todays less than ideal generators not designed for the task off grid stacks up. As the fixed charges rise more and more everyday it’s only going to make the network business model redundant.

  5. Reality Bites Avatar
    Reality Bites

    Maybe that new Director General will come to the conclusion that offering a uniform service to all at a guaranteed price is not sustainable and expanding the network or indeed maintaining reliability is a loss leader that should be discontinued. As it currently stands the NER says that supply has to be provided, however with alternative options available that could change.

  6. David Leitch Avatar
    David Leitch

    I don’t believe the argument is as simple as comparing the $KWh of storage with the $KW of network costs. Specifically the network provides redundancy whereas storage has to provide its own redundancy. You probaly need 3KWh of storage per KWh of consumption to allow for cloudy days etc. I haven’t done the exact arithmetic.

    Outside of which its crazy for society not to take advantage of the sunk costs of the grid. The way to go is the hybrid model for most communities. Rural QLD may be an exception. If you look at towns that have no grid (remote power) the costs are still high. Energy Developments proposal for Cooper Pedy is an interesting case study.

    Also the networks likely have the lowest cost of capital and battery economics are very capital cost and cost of capital dependent.
    .

    1. Terry Leach Avatar
      Terry Leach

      The problem with cloudy days is they can continue for a week or two, and you’ll run out unless you have a ridiculously expensive amount of storage. You only need storage for overnight and the morning peak. The solution to cloudy days is to increase the size of your solar array. The rule of thumb is 1KW of PV for each KWH of consumption in cold, cloudy weather, to avoid needing significant backup generation. But this means that over a year you will produce about 4 times your consumption.

      Here is the real issue. The network is such a significant component of the cost of electricity, that the economics become unstuck if there is widespread solar with storage. A solar PV owner with an array that provides 100% of annual usage on average, might go from 20% self-sufficiency to 80% just by adding storage. Who then pays for the network? Make the PV owner pay a large fixed charge and they’ll upsize their array if they can and defect from the grid. Who then pays for the grid?

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