Batteries not configured to remove demand peaks, network says | RenewEconomy

Batteries not configured to remove demand peaks, network says

Print Friendly, PDF & Email

SAPN says new algorithms needed to ensure that residential battery storage can address large swings in demand, not make them worse.

Print Friendly, PDF & Email

SA Power Networks, currently running the largest residential battery storage trial in the country, says its early finding suggest that battery storage devices are not configured to help reduce network peaks. In fact, in some ways they may be making the situation worse.

SAPN last year installed 100 batteries in customer premises in the city of Salisbury, in what is the largest virtual power plant installed to date, and is now getting some early results from the three-year trial.

The most dramatic finding is represented in this graph below. It shows how solar affects grid demand and what happens when battery storage is added. Rather than smoothing out the peaks, it can actually make the “ramp up” periods more abrupt.

Screen Shot 2017-03-27 at 8.58.21 AM

According to Mark Vincent, SAPN’s head of network investment strategy and planning, this is not a good outcome.

Vincent told RenewEconomy during a recent visit to SAPN’s innovation centre in Adelaide that it underlines the need for new algorithms to be put in place to change the behaviour of battery storage devices so it takes the peaks – both bottom and low.

This is how we would like it to look (bottom graph below). And he repeats SAPN’s push for a demand tariffs to provide the market signals to help that happen. Without a properly calibrated demand tariff, the peak might be clipped, but the bottom won’t be (top graph below). Both are important to address.

Screen Shot 2017-03-27 at 11.21.26 AM copy

Battery storage will be crucial for the SA network, because the state has traditionally had the highest volatility, and with the introduction of more wind and solar, will reduce its dependence on traditional fossil fuel plants.

That leaves battery storage to play a crucial role in meeting peak demand and providing grid stability, and SAPN hopes that it will help offset further investment in new poles and wires or equipment upgrades.

Already, the state has 650MW of rooftop solar, accounting for nearly 6 per cent of its demand in 2015/16, and within a decade the output of rooftop solar is expected to be more than minimum demand in the state.

“We expect installation of solar to keep progressing and an accelerating uptake of battery storage as prices fall in coming years,” Vincent said in a recent staff newsletter than revealed the findings.

“Integrating batteries and solar PV is very attractive for us because, on hot summer days, peak demand on the network in residential areas occurs during the 6 PM to 9 PM period when solar generation has reduced.

“Charging batteries during periods when energy would otherwise be exported into the grid, should mean that energy is available for discharge during periods of peak demand or high supply cost.”

Vincent says that the way battery management software is currently configured, a solar PV/battery combination “doesn’t help us sufficiently in managing the peak, nor does it help with the forecast trend of PV exports exceeding demand in SA in the early afternoon.”

Screen Shot 2017-03-27 at 11.21.26 AM copy

“To maximise the benefits of solar PV/battery installations, smarter algorithms in battery management software are needed to slow down the rate of charging of the batteries and their rate of energy discharge so we can lop off the demand and generation peaks.

“In turn, we need to make sure that our tariffs are designed to encourage battery vendors to configure their systems in this way, and so that customers will also see a benefit.

“Without those changes to the configuration of batteries so that they charge and discharge in smarter ways, widespread uptake of batteries has the potential to lead to inefficiencies that will require a significant response from us as distribution network managers.”

SAPN says that while it “doesn’t make financial sense” for most customers to invest in batteries just yet – contrary to some private estimates – it admits that prices are reducing rapidly.

“We think it’s inevitable that customers will invest more and more in battery systems. Our challenge is to make sure that they operate these systems in ways that reduce and don’t increase network costs to all customers.”





Print Friendly, PDF & Email

  1. Miles Harding 3 years ago

    I don’t think the this story started out very well.

    The first graph depicts the issue with a sudden shutdown when a battery reaches minimum state of charge. OK except that there are a 100 batteries in the trial, so the shutdown edge will be in the form of a chi squared distribution — no big sharp edges.

    We are probably stuck with this sort of behaviour while home batteries are too small to smooth an entire day’s cycle. Tesla’s Powerwall 2 is likely large enough to avoid most of the effect described, certainly the third generation product will be.

    I would be very susprised if the Tesla 100MW battery couldn’t interact with the grid management system to produce an effective peak shaving and stabilisation outcome.

    Considering the battery charge management software in a Tesla car, I would expect that it would be feasible to use the Powerwall 2 products in a similar grid support role, provided that the tariff and remuneration structure is able to convince consumers to participate.

    It may be more correct to say that the power network isn’t configured to help itself.

  2. Ray Miller 3 years ago

    Is this the same SAPN who each day at the same time switch all the hot water storage units on at the same time and that frequently increase in wholesale price?
    Note the same shaped spike in load just before midnight each day.

    While I agree controlling the time and amount of discharge and charge is important but would it not be better to also integrate the battery with the bigger picture as well with a focus on emissions reduction, efficiency and weather forecasting?

    • Miles Harding 3 years ago

      Would it not also make sense to replace all storage hot water with heat pump types (typical 1kW input) that can be run directly from solar output during the day? These spikes could be all but eliminated within a 10 year period.

      I have long wondered why such a changeover couldn’t be done with a revenue-neutral subsidy system where resistive hot water is penalised and heat pump or solar hot water encouraged. While penetration is low, subsidies are high, reducing to zero as penetration reaches full. Penalties on resistive hws remain after the completion.
      The state will need 100,000(?), so it would make sense to use the deceased car industry to make the units locally. It would have good export potential.
      Heat pumps may also use a CO2 cycle that is significantly, more efficient than CFC/HFC, particulary near freezing.

      • nakedChimp 3 years ago

        Better use propane (R290) like the Indians or essentially everybody else, CO2 cycles are higher pressure than current systems and thus need more copper.

    • Rod 3 years ago

      We’ve still got dumb meters here so not much scope for short term action on the HWS spike
      But they should definitely be looking at it

      • Just_Chris 3 years ago

        oh dear, that’s really a big problem. There are 3 things you require to manage the grid 1 – you need to be able to measure the load 2 – you need to be able to alter supply 3 – you need to be able to pay the person who altered their supply or demand.

        Installing 100 batteries haphazardly and expecting it to help is really not very rational. It’s also not a virtual power station, if you can’t control it remotely. It’s not management, it’s 100 batteries programed to shift the load which really isn’t all that helpful. I think that is what is being explained in the article but I can’t understand the logic if it is.

        • Ren Stimpy 3 years ago

          It’s at least a courageous step into the domain, you ****** ****.

          (Where ***** ***** means retarded negative fuck!)

          • Rod 3 years ago

            Have you been on the sauce again Ren?
            Read the post again. It is positive but suggesting better timing/use of the storage.
            At least that’s the way I read it.

          • Ren Stimpy 3 years ago

            Yeah but that hasn’t clouded my judgment, thanks.

            The more batteries we have in the system the more stable the grid supply!

            The more batteries supplying storage and energy to the grid, the less we will have to rely on shonky opinions!

          • Ren Stimpy 3 years ago

            I’m the very model of sobriety

          • Rod 3 years ago

            In my humbling experience, computers and alcohol don’t mix.
            Especially for internet shopping and if you spill some.

          • Ren Stimpy 3 years ago

            It’s all down to the sweet spot.

            On either side of that…
            … too pissed, too enthusiastic to comment (with accompanying insults)
            … too sobre – too reluctant to comment with brutal honesty.

            It’s a dilemma. And I know which way it will fall.

      • Rod 3 years ago

        Having said that, they are rolling out free smart meters (I stupidly took one hoping to get good monitoring)
        I would need to check if mine still boosts at 11pm.
        I did ring and ask to have mine reset to about 4am but was told no from memory, or was it at my cost? Anyway decided all too hard.
        But theoretically they should be able to change all those with smart meters without a site visit.
        But hey, when you are on a good thing…..

        • Greg Hudson 3 years ago

          ”free smart meters”
          You’re a lucky man. Us Victorians were lumbered with a mandatory installation, at OUR expense ($1000+) per household.

          • Rod 3 years ago

            Libs or Labor? I’ll bet the touted “savings” didn’t go to the households.
            Yes, I was lucky, I got a free import export meter in 2008 after my PV upgrade and another one last year with the wi-fi dongle.

          • Ren Stimpy 3 years ago

            Are you guys just feeling up each other’s balls?

            Smart meters are this generation’s goddamn futures, you old idiots!

          • Rod 3 years ago

            If used correctly and for the benefit of the customer they could be fantastic.
            But I haven’t seen a time of use plan yet I would sign up for.
            And if I was forced to pay $1000 for mine I would certainly be against them.
            Great for the retailer to dump meter readers though.

          • Ren Stimpy 3 years ago

            Sorry Rod you have always been good to me… but the greater good and all that!

          • DogzOwn 3 years ago

            And wholesale price in VIC for smart meters is twice as much as in any other country because local entrepreneurship bought up L&G to get biggest share, then sold it all to Toshiba. Clever Country or what?

  3. Rod 3 years ago

    They are pushing the TOU barrow any chance they get.
    So peak prices only 6pm – 9pm? Pull the other one.
    I’ve read it a couple of times and am struggling to understand the graphs but my logic circuits say “does not compute”

    • Finn Peacock 3 years ago

      It looks like they are pushing the ‘demand tariff’ barrow to me – very different (and much nastier) than ToU. Especially for solar owners:

      “And he repeats SAPN’s push for a demand tariffs to provide the market signals to help that happen. Without a properly calibrated demand tariff…. ”

      • Rod 3 years ago

        Thanks Finn,
        In what universe do they think this is peak? (7am-11pm Mon-Fri)
        I’ll be avoiding ToU and demand tariffs.

    • Ray Miller 3 years ago

      The proposed (Queensland Competition Authority) Peak time in Queensland is daily 3pm – 9:30 pm Dec to March. So the various distribution companies are trying on a rang of times and strategies.
      I agree it would make a lot of sense for the small battery systems to “trade” on the NEM, I’m in for $14 kWh, surely this is an ideal outcome.

      • Rod 3 years ago

        $14 kWh. Nice. I’d part with my electrons for that.
        With high solar penetration like SA and QLD the actual peak has been pushed out to about 6pm or later (you don’t have daylight saving, AFAIK)
        6 1/2 hours of “peak” is just gravy for them.
        I haven’t seen a ToU regime yet that would get me to switch.

  4. CaresAboutHealth 3 years ago

    If the graph is correct, and the minimum network demand is now at 12 noon, why doesn’t SAPN switch on a proportion of controlled load systems to make better use of available power? Why doesn’t it also encourage time of use tariffs so that better use can be made of power when it’s is plentiful?

    Instead, we have the silly situation where: “According to the Australian Energy Regulator, a price spike to $14,000/MWh occurred just after 11.35pm on Monday, September 5, 2016 when demand jumped 212MW as the grid operator switched on all the electric hot water systems under its controlled load operations.” [1]

    An intelligent time-of-use tariff introduced by an intelligent network manager would encourage households with batteries to keep some energy in reserve for when it’s most needed. Hope we can look forward to that happening.


  5. Geoff Bragg - SEIA 3 years ago

    So put in a bigger battery and have a bigger sponge to absorb energy beyond 12:00pm, and discharge through the night. That might fix the graph (& network loading)?

  6. Brunel 3 years ago

    There is no such thing as NEM time.

    And what about the hour that disappears?

    They should use UTC instead.

  7. Cooma Doug 3 years ago

    The load side residential systems could measure local frequency and respond in milli seconds. This would effectively create a despersed response acting to reduce power swings.

  8. MaxG 3 years ago

    It’s a scam… detect potential and exaggerated problems early to ring fence the battery issue before it hurts the traditional incumbents.

    As said elsewhere, the minute people put a proper battery up — which IMHO caters for non-sunshine hours, the raised issue no longer becomes a problem.

    • Rod 3 years ago

      I’ve had my suspicions about this one from the get go.
      I expressed interest just to get some info but got knocked back because I’m on PFiT. Couldn’t figure that one.
      I would suggest people get their own and sign up to reposit and let the market decide when you sell.

      • Greg Hudson 3 years ago

        Except Reposit does not appear to have any transparency as to ‘when, and how’ grid credits are applied. Earning $1 when the price spikes to $140/kWh does not seem to be logical to me (because of that lack of transparency). Similar to what Tom (above) said, the Reposit USER is the one who should be setting buy or sell trigger points, not relying on Reposit to ‘manage’ the system, and keeping the user in the dark like a mushroom… And you know what mushrooms are fed…

        • Rod 3 years ago

          Thanks for the heads up.
          The way things are going, I’ll probably just cut the cord when my PFiT runs out.

          • Greg Hudson 3 years ago

            I know the feeling, having just sold my house with PFIT, and moved to a brand new house with no solar. Mind you, I’m no longer paying through the nose for my power either. Same supplier (Red Energy) same area, same distributor. Price has dropped from 33c/kWh to 19.719c/kwh inc GST on both of those.

  9. Tom 3 years ago

    The solution is so bloody simple it’s ridiculous!

    Allow battery owners to bid to sell electricity on the wholesale market (via remotely controlled inverters), receiving wholesale prices, in competition with the generators. Also allow them to buy power at wholesale prices plus a “transmission margin”.

    Battery owners could “set and forget”, for example “sell whenever price > 60cents/kWh”, and “buy whenever price < 3 cents/kWh plus transmission margin of 6 cents/kWh". Or, if owners choose, they can check the AEMO forecasts and change their buy and sell bids several times every day.

    Then, if the price spiked from $400/MWh to $14,000/MWh, the battery owner would be paid $14/kWh (not 60 cents/kWh) for the energy that they sell at this price.

    If you can do it with an online share trading broker, surely you can also do it with your inverter.

    What would this achieve?
    – Battery owners would "enter" the sales market at multiple price points, hence there would not be the sudden spike in demand as batteries run out.
    – Similarly battery owners would buy at multiple price points, and so there would not be the sudden "bottoming out" in demand as batteries become full.
    – People may choose to buy 40 or 60 kWh of batteries instead of 7 or 14kWh.
    – People would be less likely to leave the grid – the "electricity death spiral" would be avoided.
    – Batteries could also be drained (at bid prices even if the wholesale price is lower) or charged (for free) for FCAS.
    – Private investment (rather than subsidies or government investment) will stump up the capital and the risk to smooth the intermittent generation of wind and solar, and they'd do it with a smile.
    – Large generators would lose their ability to manipulate the market and price gouge.

    Everyone (except the incumbent generators) is a winner! What's the problem?

    • Martin Sevior 3 years ago

      Totally agree Tom! All it needs is a small Government intervention, (to get past the incumbents crying about losing their rents) to change the rules and it would all Just Work.

  10. Ian 3 years ago

    Presumably the graph represents just those 100 houses.If so, then quite clearly the battery storage is too small, mismatched for the rooftop solar arrays or the control settings for charging and discharging the batteries poorly set. The study is less than one year into the trial so of course teething troubles could be expected.

    If the aim of the illustration is to justify decision making preferences then the whole project is a political exercise. – ie Choosing facts to justify a policy or decision. If that’s the case then their “study” has achieved its aims. Hopefully people will not be so easily conned in this way.

    If the study is to optimise the use of battery storage then these preliminary results should lead to a modification in the study.

    Others commentating have shown exactly what needs to be done.

    1. Double the battery size this will use most of the solar energy generated and release this to cover most of the evening peak.
    2. Set water heaters and other discretionary loads to operate in the day
    3. Adjust battery management system to improve ramping times and reflect FiTs to better compensate time of feed-in
    4. Install grid battery storage to compensate for inadequate domestic storage

    • Ian 3 years ago

      Subsidies can be a powerful incentive to adoption of a new technology and clearly storage is the missing element or weakest link in the energy transformation change. Early adopters are an essential part in establishing a supply chain for a new technology so why not subsidise them heavily and rapidly reduce the subsidy as installations increase or the technology drops in price.

      It seems that the preliminary findings of this study suggest two ways to gain most benefit from distributed storage 1. Build sufficient storage to match solar energy export and then match the evening peak demand. 2. Make best use of insufficient storage capacity to exactly shave peak energy export with peak energy consumption.

      If households are treated with punitive tariffs they will resent this and go off grid as quick as battery storage becomes financially viable – sufficient storage capacity could mean atrophy of grid use. If proper compensation and subsidy is given for households with storage to trade and commune with the grid in mutually beneficial ways then the grid may retain its relevance .

      This study could be an inflection point. Distributed storage will occur, the networks’ response could make or break them. Just look at the graph , the amount of storage only needs to be doubled to give a house day to day energy independence.

    • Greg Hudson 3 years ago

      ”2. Set water heaters and other discretionary loads to operate in the day”
      This is such a no brainer, but one needs to look back as to WHY water heaters are set to turn on at midnight… because when it was invented, there was little power being used after midnight (and probably still is) except, now there is no coal spinning turbines and heating that water. Instead, everything has been turned on its head, and the sun is providing the major pile of power at a point 12 hours inverse to the coal. The solution is so simple… As you pointed out above.

  11. Pfitzy 3 years ago

    I believe these batteries are all installed with Reposit? Therefore the Virtual Power Plant (VPP) would be available immediately, and management of the flows could be tweaked with the appropriate profile settings.

  12. Gnällgubben 3 years ago

    Batteries should be installed at the substation level instead of in the garage. That way several people can share battery capacity, dynamically allocated according to needs.

    • Rod 3 years ago

      Yes they should, for grid purposes.
      But the residential customer is then subject to the various prices set by generator, distributor and retailer.
      Power to the people!

    • Greg Hudson 3 years ago

      Id rather have my own, so I have control over when (or if), how much I export, and at what minimum price. Someone needs to set up an open source software program that we could all use ! And cheap(er) hardware too for that matter.

Comments are closed.

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