S.A. could dump gas plans if batteries, demand response deliver

heat wave demand response

South Australia may well have killed two birds with one stone with its choice of GE mobile diesel units for its temporary and long-term back-up generator plans, but it may have also given itself an option to kill the more expensive bird altogether.

RenewEconomy understands that the contract signed between US power firm APR Energy and South Australia Power Networks only extends to the short-term leasing of the mobile diesel generators.

The state government, meanwhile, has only committed to an “option” to translate this into a long-term contract using the same units in permanent locations, meaning it is not locked in.

It suggests that if demand response (DR) initiatives proposed by the Australian Energy Market Operator and the new Tesla battery storage installation prove their mettle this summer and next, and yet more DR and storage is brought to the market, then South Australia may decide that no government-owned gas generator is needed.

It certainly has that flexibility, and if the response to the expression of interest conducted by the Australian Renewable Energy Agency is any guide, there are plenty of demand response opportunities – where energy users are paid not to use electricity at certain times – ready to be exploited.

ARENA, in conjunction with AEMO and the relevant state governments, is seeking 100MW of demand response in Victoria and South Australia, and another 70MW in NSW.

But when it sought expressions of interest in May, ARENA was overwhelmed by the response, which came from 90 different organisations, and was as great as when South Australia and then Victoria pitched for battery storage proposals.

“We had a fantastic response from the market,” ARENA said in a letter to stakeholders, noting that a combined total of 693MW of demand response (excluding diesel) could be delivered by December 1 this year and a “whopping” 1,938MW (also excluding diesel) by December 1 next year.

Around one-third would be sourced from residential customers, another 30 per cent from industrial and 20 per cent from commercial. The technologies included industrial load curtailment (pumps, motors and other processing equipment), batteries, HVAC systems, distributed generation and residential appliances.

The proposals came from a mixture of energy retailers, networks, technology vendors, specialist aggregators, local councils and large energy users.

The response underlines the potential of a smarter way to deal with demand and pricing peaks than simply shovelling new fossil fuel generators into the system, and much of this could be brought to market if the Australian Energy Market Commission, the main rule maker, gets its act together.

The AEMC has been invited by the COAG energy council to join the 21st Century and produce a policy of demand management by next summer. It has been sitting on such rules for several years, arguing that demand management already exists.

That may be true, but that capacity is mostly in private contracts and all but invisible, and therefore all but useless for a market operator to deploy when demand is high, wind or solar output is low, and one of the fossil fuel generators crashes in the heat, as they are prone to do.

Reneweconomy has long argued that building diesel generator or adding new gas plants is not the smartest option for South Australia, because there are faster, cheaper and cleaner alternatives.

You would imagine that AEMO’s new boss, Audrey Zibelman, would also like to prove her point, oft-repeated, that demand-side alternatives such as storage, energy efficiency and demand management, are a lot smarter and cheaper than building new plants that are expensive to operate.

They are also quicker. And this is a crucial point given the creaking nature of Australia’s ageing fossil fuel fleet, and its inability to respond quickly to major events. That issue was one of the major, but unsung, features of the Finkel Review.

As Zibelman said way back in March, before a meeting with SA Premier Jay Weatherill:  “You don’t have to invest in generation that you are only going to use a few hours a year, because you can use the load itself as a balancing resource.

“It is that signal that says (to peaking power plants): ‘Hey there, we don’t really need you,” that’s going to help moderate prices. It’s pure economics applying to them and making demand a much more active portion of the grid.”

Still, it is not surprising that South Australia has chosen diesel to try to ensure the lights don’t go out again this summer, particularly with an election due in March. It has been mocked by its need to turn to fossil fuels, but let’s also remember that fossil fuel grids need huge back-up as well to address their inherent inflexibility.

In Western Australia, an 82MW diesel plant in Merredin, costing $95 million and paid for entirely by subsidies, built in 2012 and never switched on, stands as testimony to the folly of the fossil fuel mentality.

The load-shedding in South Australia in February was not due to renewables or a lack of capacity, but a misreading of the weather by AEMO and a decision by Engie to keep its 250MW Pelican Point gas station idle while its customers had their power switched off.

AEMO has since revised its processes – promising among other things to read updated weather bulletins – and under Zibelman (who took up her role in March) will have a new approach, and new technologies to deploy during the inevitable future heatwaves.

Apart from the demand response that could be in place, there will be the Tesla battery, able to inject fast response in the case of network faults, or fill in some of the gaps in the case of inadequate generation (read our explainer here to sort through the nonsense propagated by the naysers).

It will have 276MW of emergency diesel back-up contracted to APR by the local network operator (it will be interesting to see how that goes in the heat), and over the longer term it will have more dispatchable renewables, and even a solar thermal plant, contracted to supply the government’s electricity consumption.

AEMO’s safety-first decision to require four gas plants to run when there is more than 1200MW of wind power underlines what’s at risk if the system fails again, even if it’s not the fault of renewables.

The price rises that decision has caused, or rather the price falls that it inhibits, increases the economic argument for a solar tower with storage plant like the one proposed by SolarReserve for Port Augusta, or if anyone can produce similar technology at the same price.

There is also the 5MW virtual power plant being put in place by AGL, another version implemented by SAPN, and numerous other large-scale solar and battery storage options that could also see fruition – from the likes of Carnegie, Reach Energy, RES, Zen Energy and Lyon Group.

Other projects include a possible pumped hydro facility discussed by the new owners of the Whyalla Steel Works, and a silicon-based storage option proposed by Adelaide outfit 1414, for a wind farm in the state.

In the end, it is hard to imagine why the South Australian government would then need to intervene over the longer term. The temptation may be great, because right now they trust neither the market, nor the private players.

But, assuming they win the election, and AEMO keeps the lights on (the former depends on the latter), it will be hard to imagine why, in the cold light of day, they would possibly need it. They might even say so before the polling day.

Comments

27 responses to “S.A. could dump gas plans if batteries, demand response deliver”

  1. Ren Stimpy Avatar
    Ren Stimpy

    Not to mention that residential, industrial and commercial solar and storage is also demand reduction.Attacking the problem from the other end.

    1. Alastair Leith Avatar
      Alastair Leith

      yes although as we start seeing fuel switching from fossils in industry to grid sourced electricity demand could rise too. depends on the resources available to them on location and if they sign exclusive PPAs with RE developers I suppose. If wholesale prices stay high they will sign PPAs more and more I expect, is much more common in USA, many major tech brands and others shooting for 100% RE via the grid transmission on PPAs.

  2. George Darroch Avatar
    George Darroch

    Here’s hoping. I still see idiotic references to the batteries only supplying “7 minutes” of supply, not realizing that they’re like the starter motor on your car.

    1. Joe Avatar
      Joe

      George, please pick up a copy of today’s Australian Financial Review Newspaper. Now that Italiano Matty is no longer the self proclaimed Minister for The Mining Industry we had Peter Freyberg from Glencore ( Coal Mining & Aluminium giant here in Australia ) step right up as the ready made replacer for Italiano Matty. Peter F. was full of gems commenting on the importance of Coal, delaying meeting Paris emissions targets and his warning about the rhetoric on battery storage…..on which the words of wisdom from Peter F.. and you’ll love this as I quote……”The much heralded proposed renewable lithium battery storage in South Australia of 129 megawatt-hours would power our aluminium smelter here in New South Wales for a grand total of 7.7 minutes.”
      Oh dear, another dude in a position of high responsibility that doesn’t even know the purpose of Elon’s Big battery that is coming to SA. We had stupid ScotMo the other day with his Big Banana, Big Prawn slag off at Premier Jay and Elon’s Big Battery. Now we have another dumbass in Peter F. Why does no one get in front of the National TV and take these dopes down once and for all. They lie and spread mis information without being called out.

      1. Mike Westerman Avatar
        Mike Westerman

        And in particular remind the halfwit of how much longer than 7min it takes to restart a smelter after a large fossil fueled power station trips on a hot day, and takes hours to restart.

      2. Stephen Avatar
        Stephen

        Methinks stupid like Goebbels!

  3. Tom Avatar
    Tom

    I’ve been thinking about solar thermal and storage.

    These power stations (as I understand) are limited to about 150-200MW and are limited by the distance between the mirrors and the tower. There’s only so much sunlight that can be reflected before the mirrors get very far away.

    A large amount of the capital and probably ongoing cost would be in the construction and maintenance of highly accurate dual axis trackers for each of the mirrors.

    Which makes me think – what if there were the molten salt tanks coupled to a steam generator coupled to a wind or PV farm but heated by a simple electric element powered by surplus energy from wind or PV? You could make the tanks and the steam generator as big as you like, and not worry about the mirrors, towers, or even the pumps.

    It would not work at the moment, as the wind/ PV farms get the same price as every other generator, and the prices are invariably high, so there is no incentive for them to store (with its capital costs and efficiency losses) rather than to just sell it.

    But in future if there were parallel “base-cost” and “dispatchable” markets with dispatchable energy being paid a premium, or if the extreme price troughs become more frequent on sunny and windy days, then this is something which might begin to make sense. Especially if it is cheaper to store energy in the form of hot molten salt than electrochemically in a battery.

    1. Mike Westerman Avatar
      Mike Westerman

      Tom – this was looked at by Lloyd Energy when they were developing their graphite solar thermal facility – in fact they actually proposed to build one on Yorke Peninsular to be used when wind down that way was constrained off. Graphite has advantages over molten salt – more inert, not aggressively corrosive and higher temperatures, offsetting the lower cost of the thermal storage medium. Both suffer from Carnot efficiency limits. But way cheaper and probably more suitable for SA is chilled water storage or ground effect storage systems for heating and cooling, with the advantage as well of being able to be installed at the load.

      1. solarguy Avatar
        solarguy

        Yes but only at the load.

    2. solarguy Avatar
      solarguy

      CST can be any bloody size it needs to be, you would build multi 200 MW plants to get the size you need.

      As far as resistive element goes for molten salt storage that can be done, but heat pumps are more efficient but more costly.

      1. Alastair Leith Avatar
        Alastair Leith

        Can be any size — if there are multiple receiver towers. There’s a limit to how far they can reflect a beam of sunlight and have it hit the target (and not wobble off it either I’d expect). It’s perhaps harder than it sounds to shoot light a kilometer and hit a bullseye; they need an even spread on the receiver or could damage it or reduce performance, near enough is not good enough.

        One big cost saving the Spanish builders of Gemasolar were looking at was a different way to track and focus the rays onto the target, in Gemasolar they had some kind of optical device on every single heliostat IIRC, but were looking to some kind of video processing software to use off-heliostat cameras somehow (heck knows how they tell one beam from another with 1000s of beams!).

        1. solarguy Avatar
          solarguy

          Well of course there would be multiple towers Alastair, that why I said multiple plants. i.e. each plant has a tower. Perhaps your more tired than I am.

          1. Alastair Leith Avatar
            Alastair Leith

            yeah okay, note Ivanpah is three towers for 377 MW (Net) / 392 MW (Gross) capacity => ~125 MW per tower.

    3. Alastair Leith Avatar
      Alastair Leith

      THere’s no reason you can’t but remember the big efficiency loss is in the conversion of heat to mechanical energy in the turbines. As per the Rankin Cycle, so batteries, while much more expensive, also much more efficient round trip. Also the conversion of solar radiation in to heat is more efficient with a mirror and thermal receiver and a PV panel (~20-25% for mass produced modules) then power circuit losses.

      One problem it could solve is working around thin transmission by filling local storage near a demand centre for use at times of peak load in a capacity much greater than what the transmission has capacity for itself. The thin pipe concept.

      Pumped Hydro Energy Storage makes more sense for pairing with wind as there are less energy losses because no electrical to heat to kinetic energy conversions.

      1. Gary Rowbottom Avatar
        Gary Rowbottom

        One man’s Rankin Cycle efficiency loss is another two other people’s jobs! Not a bad thing I say. the base unit for CST will be mostly in the 100 – 150 MW range, and replicate plants for more capacity – a la Solar Reserve’s proposed Sandstone project – 10 towers for 2 GW I think. Essentially CST could be one of “coarse tuning knobs” in the renewable energy system – relatively big blocks of energy, relatively big storage. Other technologies (like batteries), more the fine tuning knobs. PHES perhaps has both modes, to a fair extent but needs to be charged using grid power – so in itself it is a power time of use shifter. Remains to be seen how much CST may be optimum, a moot point as we wait, still, for Australia’s first utility CST/storage build. Based in SA Government’s court on that one. The 30 month build time (and required additional preparatory time), won’t see MW from one of these get any closer until that long term power purchaser link is secured.

  4. Rob G Avatar
    Rob G

    The gas build talk in SA was a bluff to ward off the dinosaurs. Another clever move by Jay.

    1. solarguy Avatar
      solarguy

      Jay’s the man!

    2. Alastair Leith Avatar
      Alastair Leith

      Electoral insurance — at a pretty cheap price in terms of what it costs to win an election.

  5. Mike Westerman Avatar
    Mike Westerman

    Giles you would have seen the announcement by Hillgrove of the PHES to follow the end of mining at their Kanmantoo site – I believe it was in the Advertiser, but also on their most recent announcement to the market. If this comes on line before 2020, along with the other measures above I would think it completely obviates the gas reserve alternative.

    1. Giles Avatar

      No, that one passed me by. thanks for heads up

      1. Mike Westerman Avatar
        Mike Westerman

        http://www.asx.com.au/asxpdf/20170731/pdf/43l0bxtvldrz61.pdf or search ASX Announcements for HGO in the last week.

        2nd last page

    2. Alastair Leith Avatar
      Alastair Leith

      2.4 years is a quick build for PHES, but if there’s little remediation to do it’s possible I suppose. 🙂 Can’t wait to see dedicated PHES storage go in to shut up the RE critics once and for all (well never the Bolters — but hopefully some of them).

  6. Mark Diesendorf Avatar
    Mark Diesendorf

    Giles, in your eagerness to dump open-cycle gas turbines (OCGTs), I think you may have overlooked two relevant factors: (i) the timescales during which technologies can be operated and (ii) timescales for disseminating (i.e. planning, ordering, building and commissioning) the technologies. (Under ‘technologies’ I include other measures such as energy efficiency.)

    Batteries and contracted demand reduction can respond to an imbalance between supply and demand within milliseconds. They are both a vital part of future large-scale 100% renewable electricity systems. They can be disseminated widely within 1-2 years. But they can only be operated during events of duration up to 1-2 hours. Battery storage (in MWh) is expensive and hence limited, and so is the time that refrigerators, air conditioners and aluminium smelters can be offloaded.

    So we also need technologies that can operate for several days at a time when requiired. Those that can be planned and installed within 1-2 years include OCGTs and diesels. These can burn either fossil or renewable fuels, when the latter become available in quantity. Major energy efficiency programs can also be implemented on this timescale.

    Pumped hydro (including off-river) and CST with thermal storage will likely take longer to be disseminated, possibly 3-5 years depending upon rated capacity.

    It seems to me that South Australia needs to implement most, if not all, of these options. OCGTs and diesels can do essential things that batteries and contracted demand management can’t, and vice versa. For the longer term, once pumped hydro and CST are built, we can remove OCGTs and diesels from large grids.

    1. BushAxe Avatar
      BushAxe

      I’ll be honest I find your opening paragraph quite patronising, I think it’s quite obvious that there is alot in the planning pipeline for SA energy in the next two years. So for other than those who know firsthand what is in being planned the rest of us will just have to speculate, because the SA government has been keeping everyone guessing so far.
      Alinta has proposed a 300MW OCGT north of Adelaide so why would the SA govt bother building their peaking plant as well?
      Why are batteries only good for 1-2hrs? And you don’t think the price of batteries are going to fall anytime soon even though it has been widely discussed?

      1. Alastair Leith Avatar
        Alastair Leith

        No news still on the Tender for ~100MW of SA government power purchase which Solar Reserve and others tendered on (and the TOR were changed from Renewables to “low carbon” which fossil gas with FE>3% isn’t but somehow people convince themselves it is “low carbon” by pretending FE’s are only 1% b/c industry assures us so).

        Mark is correct that fossil or bio-fuel OCGT are essential to power a grid when there’s a high penetration of RE and a thin connection to other parts of Australia, as is the case with SA (and the island WA SWIS grid). Overcapacity of Wind, Solar and PHES/CST/Battery storage has to be so great to cover the last 5-15% that its *way* more economical to build OCGT and then ideally fuel them exclusively with biofuel. Hopefully the biofuel is coming from well managed plantation timbers like the Mallie Gum plantations already growing in WA wheatbelt that also lower salinity levels by pumping out surface water for the ag region. There’s work by Curtin Uni engineers on using biochar to filter the biogas to get the impurities out that tend to deteriorate OCGT and then convert it to liquid fuel for bulk storage. I’m sure others are also looking at this kind of tech, there’s a significant market opportunity as soon as (global) C pricing comes about.

  7. Andrew Coffey Avatar
    Andrew Coffey

    Giles and readers, you might find this podcast interesting: http://www.bbc.co.uk/programmes/p057xsl0 and the associated papers. It considers the time it took to realise benefits of the invention of the dynamo and electric motors for industrial purposes.

    Originally the dynamo was used to directly replace steam powered infrastructure (swap out and in electric motors). It was only after years of application that industry realised the efficiency gains of the electric motor by utilising smaller and more distributed units.

    Where am I going with this?

    I see many analogies with the current state of play with respect to batteries and demand response. We won’t need permanent units installed if we can quickly figure out how to optimise batteries and demand response.

  8. michael nolan Avatar
    michael nolan

    Giles and Mark D, I thoroughly agree with your comments.
    Demand Mgt will provide much of the solution from the demand end, short term OCGTs and diesels and batteries the supply end, and medium long term Pumped Hydro and CST.
    Economics for PHES is very compelling on the duration side.
    Consider SA’s ‘Tesla’ batteries are $130 m for 100 MW/129 MWh system ….$1.3 m per 1 MW and per 1.3 MWh,
    Whereas Snowy 2.0 Pumped Hydro is reportedly $2Bn for 2 GW and with a duration of almost 1 week. By my calcs (7 days = 168 hours supply at 2 GW) Snowy 2.0 equates to $1 m for 1 MW, but a cracking 168 MWh.
    In other words, when comparing price per MWh, battery will last just over an hour, and PHES is duration in the order of 1 week…….for the similar installed cost per MW output.

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