First thoughts: Snowy 2.0 will lift emissions without more renewables

Print Friendly, PDF & Email

The project data

The key data we focus on for Snowy 2 are summarised in the following table.

Figure 1 Snowy market data. Source: Snowy, ITKe

Figure 1: Snowy market data. Source: Snowy, ITK

The distinguishing feature of Snowy compared to all the other pumped hydro projects in our sample is the amount of storage.

Snowy has between 10 and 20x more storage at maximum power output relative to 11 other recent and proposed projects in our sample. That’s assuming we haven’t made any mistakes in looking at these projects.

Figure 2: Pumped hydro projects Australia and world wide sample. Source:Company, Cannacord Genuity, ITK

Figure 2: Pumped hydro projects Australia and world wide sample. Source:Company, Cannacord Genuity, ITK


Figure 3: Pumped hydro projects, storage hours at max power. Source : Company ITK

Figure 3: Pumped hydro projects, storage hours at max power. Source: Company ITK

We note:

  • The capex per MWh of the max deliverable energy ie, $4.1 billion divided by 350GWh @ 365 days a year is just 10 per cent of that for a more typical project at around $35/MWh
  • The project has zero chance of making money solely as merchant plant – that is, buying electricity out of the market at off-peak, pumping it 27km uphill at 70 per cent round trip efficiency, and then selling it at peak prices. The project will be able to earn some revenue doing that, but 2000MW is a lot of power and obviously only a small fraction of the storage could be used on any one day. Just to cover the round trip efficiency requires off-peak prices to be, say, 30 per cent less than peak prices. Snowy will certainly have a strong upwards influence on off-peak prices and negative downwards push on peak prices if it was operating on a regular basis.
Figure 4 Average of daily max and minimum demand and price Source: NEM Review

Figure 4: Average of daily max and minimum demand and price. Source: NEM Review

  • As the NEM is current configured, Snowy 2 will increase electricity demand, and increase CO2 output. In a sense it will make breakdowns of the other plant more likely due the probably coal-fired power required to do the pumping. Of course this would could change in a largely renewable world, but at first blush when renewables penetration is 20-30 per cent there is a 60-70 per cent chance of pumping being powered by coal. If you think the lowest prices are when renewable output across the NEM is high, then the pumping might have a higher renewables share. At the moment, though, NSW – one of the key markets for Snowy – has a relatively low renewables share, anyway.
  • The private sector would not contemplate this project without a PPA. Yet at the moment there is no PPA. Perhaps one will be guaranteed by the NEG. The most cynical way of looking at the whole thing is that the NEG is just a mechanism to make Snowy 2 viable. We don’t necessarily go that far. But what we would say is that for the time being we don’t see AGL or Origin having much need for Snowy’s reserve capacity, even in 2024.
  • So more broadly, excuse the pun, we can only see Snowy making sense if Australia has a much higher share of renewables than contemplated under the NEG.
  • Snowy states it can finance the project off its own bat. Clearly this will mean no dividends to its owners for the next seven years. Even so, it will require, say, $3 billion of likely on balance sheet debt. Snowy had $0.8 billion debt at 30 June 2017 on $3.5 billion of assets. Although profits were well up last year, if generation output falls the REC income would drop sharply.
  • Snowy will likely crowd out various other pumped hydro projects, particularly those proposed for South Australia. It looks as if more transmission from South Australia to the eastern states will be built making Snowy 2 pumping requirements and output sort of accessible to the South Australian market.
  • However we doubt it will crowd out all the battery stuff. Snowy talks about its inertia but, increasingly, digital inertia is more easily and swiftly provided by batteries which can be installed in three months. No one is going to spend $4 billion to get some inertia into the system that can be provided by batteries for much less. Secondly, behind the meter batteries, which have entirely different economics, are also likely to proceed provided the forecast supply and price reductions become a reality.
  • Finally, we note that the costs and the time line are in a sense still best-case estimates, in that despite contingencies they exist only on a piece of paper. Whether it’s reverse osmosis water plants, peaking gas in Victoria, or LNG plants around Australia, experience suggests that any project budgeted to take seven years and cost $4 billion is at risk of taking longer and costing more.

David Leitch is principal of ITK. He was formerly a Utility Analyst for leading investment banks over the past 30 years. The views expressed are his own. Please note our new section, Energy Markets, which will include analysis from Leitch on the energy markets and broader energy issues. And also note our live generation widget, and the APVI solar contribution.


  • Dee Vee

    How many hours can we get out of the existing Snowy 1 without the extra $6b capacity increase for Snowy 2?

    Seems this existing capacity has been surreptitiously added into the total capacity to make Snowy 2 look far more appealing than it is.

    • Mark Fowler

      I don’t think you quite “get” the concept of pumped hydro. Currently all the potential generation from Snowy 1 is realised within the constraints of water supply (rain and snow), environmental and irrigation requirements. Pumped hydro adds to the output by effectively reusing the water already in the system.

    • RobertO

      Hi Dee Vee Snowy 1 has Tumut 3 as a PHES system, one of its problems is that it has to run overnight to recharge, however Snowy 2 has some 9 days of storage so part of the business case may be that Snowy 2 may run to recharge Snowy 1 and Snowy 2 may play the market more when prices are very low or even negative. Until we see the business case we are all guessing answers.

      • Mike Westerman

        Robert T3 has 1500MW of generation but only half the units are pumps, so it doesn’t have to recharge at any time other than when it is cheap to do so ie it can vary the generation thru T1,T2 and T3 to cover high priced events pretty much indefinitely as long as it stays within the water allocations downstream. They are making minimal use of the pumps now so it makes you wonder what they really want a Sherman tank for – certainly not for something benign.

  • Philip WT

    It can’t do 350 GWh/day. At 2 GW at most it can do is 24 GWh/day assuming 50% pumping time and 50% running time and ignoring efficiency. Do your capex/$MWh figures need adjusting?

    • David leitch

      Philip. You are completely correct. I did the calculation that way to be consistent with all the other pumped hydro projects i look at.

      • Philip WT

        Thanks David, so at 24 GWh per day its capex costs must be very similar in $/MWh of delivered energy to the others (seeing as the article mentions that at 350 GWh/day it is around 10% of the capex of other projects?

        • Mike Westerman

          And at a more credible upper limit of 12GWh/day, the economics are a dog.

  • Tim Forcey

    “we can only see Snowy making sense if Australia has a much higher share of renewables…”


    • David leitch

      Tim: Compliments of the season to you. Where do you see this project getting its revenue from in say the 2025-2035 time frame? I can see some arbitrage revenue, but most of it will have to come from selling firming contracts and right now I am struggling to see who will need to buy that quantity. To me it looks like a mismatch between power and storage. I am trying to see how most of that storage could have a value.

      • Mike Westerman

        Especially David if you read in SnowHydro’s recent annual reports! Last year admitted pressure in that market, this year shows a decline in income from insurance, with the outlook being zero growth.

        When you look at international data (I’ll dig out the paper), the majority of schemes operate for <8h in a cycle – it's difficult to see how more could be sold profitably. If that's the case then the cost of $400k/MWh of storage is 2-3x many other available schemes.

        • David leitch

          The potential use is as a strategic reserve. Say a big coal generator like Bayswater broke down. Snowy could replace if tor say a week. That’s useful, buts good as it is its still only a week.

          Maybe the wind output over the NEM is down over Summer, Snowy could operate to supplement that and refill the dam in Winter when its windier.

          There are possibilities and they are all stuff that’s interesting but the private sector wouldn’t race to spend $4 bn in the vague hope that it might make a return.

          Its a shame the outline of the business case wasn’t released. The political statements today reported in the SMH about replacing diesel are hardly reassuring. There seems to be this attitude of “we know what’s best for you” don’t worry if Paul Broad says it will get an 8% return then that’s all you need to know.

          At the same time I have some sympathy for people who have worked very hard to make this project go as far as it has. I’m sure there is more to it than is revealed so far. Its just as I say, its not obvious how it will make a return.

          • Mike Westerman

            If Bayswater broke down for a week or a year, a robust network of solar, wind and distributed PHES would step in and cover it. And who pays for strategic reserves? When did the RFP’s for that come out?

          • WR

            The surest way to get excess supply for arbitrage purposes is to have a lot of utility-scale PV. If you have that, then every sunny day will give you power at low cost for recharging your storage. That might be 200+ days per year where you can recharge during the day to suppl the evening peak.

            But you need a LOT of PV to ensure that happens. For NSW, you would probably want in excess of 10 GW of installed PV capacity to achieve that. Is that going to happen by 2024?

          • Tom

            @WR – The other problem with this is: who’s going to install the PV?

            If NSW already had 9GW of installed PV creating a huge duck-curve in the wholesale energy price every sunny day, would you be the company to install the 10th GW?

            Just as DL correctly cited that Snowy 2.0’s demand will raise the trough prices and its supply will drop the peak prices, therefore reducing its viability, the same applies to PV and wind installation. The more there is, the more it will depress prices when it is sunny or windy respectively.

          • Mike Westerman

            Hopefully the other 75% of rooftops without solar will jack the current 6GW to 20GW before too long. Too bad for Malcolm if the existing owners throw in a Powerwall each – 19GWh of storage would put Snowy 2 out of business!

          • Tom

            Partly – Snowy 2 is talking about 350GWh of storage. Less useful storage than battery storage, but a truckload of storage nonetheless.

          • Mike Westerman

            But who pays for it? If only 8h/day is saleable, then only 16GWh is an economic asset, and the rest is going to have to be paid for by some other means. If that is a levy on all power sold to provide “reliability” shouldn’t it have to pass a RIT-T?

          • Tom

            @MikeWesterman – I totally agree (despite my very basic knowledge on RIT-Ts). It’s the same for open cycle methane generators or diesel generators that are on standby for 8720 hours per year and are only used for 40 hours per year. Terrible capacity factor, huge LCOE, but necessary nonetheless. Even if they only generate during wholesale price spikes they’ll never make their money back, so who pays?

          • WR

            Yes. The PV won’t be installed until it’s cheap enough to make a profit from a PPA of $30/MWh.

      • RobertO

        Hi David, I believe that if Snowy 2 goes ahead (as 2000 MW) it will remove some 1400 MW of active coal power (Not Name Plate). Coal is finding hard to compete with solar duck curve, with wind at any time and if you add PHES it becomes the nail it its coffin. It need not run 24 hr a day as solar takes coal on 8 am to 5 pm, wind does it bit at any time during the day or night and PHES does its damage any time it running (often only peak time only). PHES can buy at any time day or night (and will do) making this amount of coal worthless. As for the 9 days of storage that has very little value at this time as we only have a small amount of RE in the system. It maybe that Snowy 3 and Snowy 4 are the value add that make this storage worth something.

        • Mike Westerman

          Robert the only place likely to see a duck curve in the next few years is SA, and Snowy 2 is transmission constrained from doing anything about that. So coal will be untroubled – they will have a market for power at night in the rest of the NEM.

          If you believe that this project will work, but just not yet, then why is money being spent on it now, rather than the projects we need now?

  • Jack Gilding

    Thanks David. Hydro Tas says of its Battery of the Nation project “Work is already underway on these projects and we expect that our initial studies will be completed towards the end of 2017.” so hopefully we will soon have two big pumped hydro projects to compare.

  • Ray Miller

    Before anyone gets excited just get hold of and watch the 2017 ‘Utopa’ series this year for a reality check on how big government projects go…it the answer is still SH2.0 watch it again.
    I’m sorry but a lot more convincing needs to take place before I’m a believer, with Australia’s carbon emissions so high and increasing any energy project answer needs to dramatically reduce our emissions to have any chance of viability.

    • Joe

      Oh, The Utopia…classic watching this stuff…spotlighting Govt. stuffups.

  • Tom

    Hi David,
    Check out the operation of Shoalhaven pumped hydro. It has the highest capacity factor of the Australian pumped plant @ 20%.

    • Mike Westerman

      This is probably more to do with the water transfer and constraints than that driven by the market.

      • Tom

        It’s hard to find enough cheap hours to charge and expensive hours to discharge. 20% CF is reasonable, 70% is impossible.

  • howardpatr

    Note that the report for Turnbull makes no mention of the alternative of developing some of the pumped-hydro site identified by the UNU’s Prof Andrew Blakers.

    Snowy 2 will be seen as Turnbull spin if comparisons are not made between it and say the top ten sites from the 22,000 identified by Blakers.

    Why hasn’t appropriate funding been provided for studies to be carried to determine comparative cost/benefits of distributed pumped- hydro? Twenty nine million dollars were found, without a blink of an eye, to promote Turnbull’s proposal.

    No wonder Turnbull’s determination that he should be guided by the likes of Abbott, Joyce and Canavan on the inevitable path to a renewable energy future has resulted in yet another nickname for him; TRUMPBULL.

  • David Osmond

    Hi David,
    any comment about whether Snowy Hydro will be incentivised to underutilize Snowy 2.0 to avoid reducing the profits from their other generating assets? Similar to your observation of CS Energy underutilizing Wivenhoe.

    • David leitch

      No-one is going to spend $4 bn and not try and get a return on the money. Snowy, at the moment, doesn’t own any thermal generation and this will be its largest asset. It will be incentivised to get a return on the capex. In other words the company and market structure are quite different to that of CS Energy.

      • David Osmond

        Hope you’re right, but is it really all that different? I believe Snowy has 4.1 GW of other hydro and 1.3 GW of gas, so in terms of power the other assets are well over double Snowy 2.0’s 2 GW.

      • wideEyedPupil

        Doesn’t Visy burn a lot of paper up (and probably claims RECs for it!) there and put that into the same transmission lines?

  • 67% not very efficient, no doubt due to the long (27 km ) tunnels.
    Yes it will ‘crowd out’ other lower cost projects in dispersed locations. Marginal cost of pumped hydro increases greatly after the ‘low hanging fruit’ first several hours of storage that is used frequently. At some point fueled generation (OCGT’s) is cheaper for backup than building more storage.
    I wonder how come there hasn’t been costing of a range of projects

    • Mike Westerman

      Don’t wonder why – just think of Malcolm trying to shore up his votes!

    • wideEyedPupil

      Dams produce CO2, but not like the CO2 and methane F.E.s OCGTs and their fossil gas transmission lines and well heads produce. For backup (and winter wind droughts) we need either biofuels or power2gas and gas/liquid biofuel storage capacity.

      Let’s get the NEM up to 50% RE before we fret about additional backup though.

  • Tom

    Brilliant article. I’ve been screaming exactly these same points from the rooftops to anybody who’ll listen – it’s hard to get the message out.

    Its sole function is to convert non-dispatchable energy into dispatchable energy, at a cost of 35-50% of the regenerated energy lost in inefficiencies. And at present the major source of non-dispatchable energy is coal.

    I read the “summary” on Snowy Hydro’s page. The modelling is all hidden from the public, but the summary stated that this scheme was compared with batteries and open cycle gas turbines, and came out cheaper. So no comparison with smart demand management, or smart utilisation of privately owned and funded assets (home batteries, EV batteries, rooftop PV), or other forms of storage (thermal-steam, hydrogen etc). And they didn’t reveal any of their assumptions – they possibly assumed that battery storage will remain the cost of a Tesla Powerwall forever more.

    1414 Degrees (developing molten silicon thermal storage with steam regeneration) claims that their electricity regeneration efficiency is around 50%. That’s almost as good as Snowy 2.0’s 67% at full capacity. Thermal storage could potentially be placed by every wind and solar farm, the wind and solar farms tripled in size, and no transmission upgrades would be required.

    We talk about capacity factor of generators. Capacity factor of transmission and distribution lines is possibly even more important. Think about it – transmission and distribution comprises half of your energy bill after all.

  • Ken Dyer

    I have been following Tony Seba for some time now, and it seems to me that Snowy 2 flies directly in the face of his predictions, many of which are being realised today.

    Snowy 2 is a political straw man.

  • Ian Cutten

    Apart from the will to do it, why is it not possible by 2024 to have sufficient intermittent energy (PV & wind) to power the vast majority of pumping. If it is a proposed requirement that intermittent production must be made dispatchable with balancing storage then it should also be a requirement that pumped hydro be made dispatchable by balancing with the use of intermittent energy.

    • Ray Miller

      The use of ‘intermittent’ in this case is a poor choice, Dr Finkel was using ‘variable’ renewables (like wind and solar) as this is a more accurate description, the variable is predictable within a close margin, as for the large machines which tip randomly, regularly are ‘intermittent’ which is a more accurate terminology.

  • Mike Westerman

    Bear in mind they haven’t done the EIA yet. The lower levels of Tantangarra are anoxic, with frequent algal blooms affecting the lake. Those toxic waters will mix with Talbingo. Warmer Talbingo water will tend to float on the top and release a significant amount of heat which is going to impact on the alpine environment around the lake. Redfin Perch will get thru despite the magic of electronic zappers in Chapter 17 of the FS and will end up destroying the trout there and in Eucumbene. The stirred up poor quality water is going to be released down into Eucumbene and hence into both Murray and Tumut schemes. And so it goes on…all in our most important national park.

    • RobertO

      Hi Mike Westerman, When the water is released into the bottom dam it will both mix and airate (via movement of water). It will be about a 2 degree (or 3
      at the most) difference which will depending on useage will not add up to large change in temperature (its something like 0.001 Degrees per hr and then your may pump some of this water back up. This site will not be used for 9 day straight at a time.

      • Mike Westerman

        If it doesn’t get regular hours of use, it won’t pay for itself (meaning we all will pay for it). If it does get regular use, inflows into Tantangarra will stir up anoxic water which is then released into Eucumbene or returned to Talbingo. 2-3deg change will alter the winter environment around this lake which sits in a sensitive alpine environment. And that’s quite apart from the issue of spreading redfin perch into Eucumbene so all the trout their die of EHN. Dinorwig, which gets a mention in the Snowy 2 FS report, killed all its endangered Arctic Char as a result of warm water and stirring up anoxic water.

        Wrong size, wrong place, wrong balance of power and energy, wrong reasons for promotion.

        • RobertO

          Hi Mike Westerman The temperture difference between the two lakes is 2 or 3 degrees and when you move water between the two site the change in temperture will only be about 0.001 per hr. Given that this system will be used about 4 to 6 hr per day your are raising the temperture by about 0.06 of a degree, then it it gets pumped back up you are removing part of that water so the effective change is about 0.03 degrees per day. Most of that heat will evaporate from the lake anyway. Heat is not an issue at this site. Climate Change is going to do more damage that that.

          • Mike Westerman

            Your understanding of the stratification of alpine lakes is letting your down! Most of the 235GL in Tantangarra is at the long term annual average temperature for that location, probably around 7-8deg (but you wouldn’t believe how hard it is to find data on that). So water coming in from Talbingo will initially float on the surface releasing all that heat into the environment. Not big deal in summer, but in winter when it is -10 or -15 the impact is dramatic. In winter that warm water then cools to ambient, and then falls down thru the warmer water below, stirring up the anoxic water from below and killing fish and other aquatic creatures. That stirred up anoxic water is then released into Talbingo, where it also does damage, before the whole cycle starts again. Apart from that, when redfin perch infected with EHN arrive in Eucumbene it’s all over red rover for the trout fishing industry.

          • Ian

            Good points you make Mike, far more to consider than just a big water battery

          • Brad Sherman

            In winter, neither dam is likely to be appreciably stratified. When full, Tantangara is only 12 m deep on average (max 18 m). In winter, it is extremely unlikely that either Tantangara or Talbingo will not be well-mixed, nearly saturated with oxygen, and close to isothermal at the 7-8 C you mention. The seasonal stratification cycle will see the development of a surface mixed layer around 4-6 m thick which will be slightly warmer than the local mean air temperature. The water at the bottom will remain close to it’s temperature determined by overturn in winter – the minimum temperature is typically reached during the first half of July. There is hardly a reservoir anywhere in Australia that doesn’t follow this pattern.

            The response of the stratification of both reservoirs will depend strongly on exactly where the outlets are placed: I’m guessing close to the bottom of both reservoirs so as to maximise the useful storage available to the power station. If Talbingo is 2-3 C warmer than Tantangara, then this water if discharged at depth in Tantangara and will probably serve to substantially or completely remove the thermal stratification. After it enters the bottom of Tantangara, it would rise and entrain ambient water from Tantangara as it approaches the surface. How far it rises, depends on how much entrainment occurs and that is sensitive to outlet dimensions and flow rates. Shear-driven resuspension of bottom sediments could be a major issue depending on what is in the sediments.

            It is theoretically possible (in the absence of design data and modelling – I’m just speculating here) that pumped storage operation would reduce the level of anoxia and potentially internal nutrient loading (caused by anoxia) in Tantangara which in turn could reduce the severity of algal blooms.

            I would expect potentially profound changes in the annual stratification cycle and, through it, the aquatic ecology of both reservoirs. How desirable or undesirable the ecological changes might be is hard to speculate without performing some thorough reservoir hydrodynamic modelling. I am sure such modelling will be undertaken as part of the EIS.

            Reservoir hydrodynamics and water quality is what I do for a living – you pushed a button!

          • Mike Westerman

            Brad all my reading of water releases in the Upper Murimbidgee show highly anoxic water from stratification in the 40m odd depth of water nearer the wall, fed by eutrophication from organics feed in during summer regrowth due to draw down of water. This is apparent from photos of the reservoir. It may well be there is good mixing in the upper several meters but it certainly seems to not lead to wholesale “turning over of the reservoir. I suspect that a strong mechanism in winter is near freezing water in the broad shallows leading to subsurface flows down into the depths of the reservoir, taking enriched sediments with them. This water would be cooler than the body of the reservoir at depth sitting at the long term annual average temperature. Obviously the stratification that is observed suggests mixing is only occurring in winter in the uppermost layers.

            Warmer inflowing Talbingo water will raise some anoxic water to the surface, entraining sediment with it as you point out. Each cycle is less than 5% of the Talbingo volume, but once on the surface that water will cool rapidly in winter and lead to little mixing beyond the intake/outlet. Talbingo water is likely to also be drawn from depth and so be equally anoxic, so I doubt the health of either reservoir is going to be improved.

            It’s good that you are a professional in this subject but I suggest you do more research on this potential issue and bring it to forums such as this. I am a professional hydro engineer and do not want poster-child failures to smear the PHES industry due to a single politically motivated misfire!

          • Brad Sherman

            Can you provide the data (or a link to it) you refer to? I’m happy to provide interpretation. I’ve spent decades specialising in reservoir management (oxygenation, reservoir destratification, algal bloom control, CH4 emission measurement) and my comments reflect experience across a wide range of reservoirs, but none in the Alps.

            I’m very familiar (i.e. instrumented them personally) with the dynamics of Cotter, Googong, Hume, Chaffey, Hinze, Little Nerang.I’ve been asked to make sense of data from Burrinjuck and a number of others including Dartmouth, Eildon and Burrendong. Most reservoirs lack high quality data that allow unequivocal interpretation – especially post 2000 or 2005ish. DLWC did a very good job monitoring (from my hydrodynamic biased perspective) back in the 1990s during the surge in algal bloom research in Australia.

            Differential cooling (relatively colder water generated in shallow areas that underflows and accumulates at the bottom) is a definite possibility for Tantangara. We saw quite a bit of it in Chaffey Dam. However, the gravity currents seldom have sufficient energy to resuspend sediments to any appreciable extent and penetrative convection will mix to the bottom of all storages in Australia during winter. I would be astonished if Tantangara was an exception given its location unless temperatures do fall below 4 C and then the equation of state complicates things considerably.

            Again, this will be straightforward to model and I hope someone is doing this because it could bring a quick end to speculation one way or the other. Or at least identify the need for some superior data acquisition for a year or two to suss out exactly what is happening.

          • Mike Westerman

            Brad I searched fruitlessly for temperature profiles but found quite a few reports on water quality in relation to the various studies done on supply of water from the Upper Murimbidgee to Canberra. I’ll have to see if I have the links. All showed poorly oxygenated water and frequent algal blooms in summer. It seems that it does suffer stratification and I can only think of the gravity flows from snow melt. You are no doubt correct that the velocities would be too low to resuspend bottom sediment but these flows would bring in eutrophic water from the alpine grasslands around the lake. Given the natural lakes in the park that don’t seem to suffer the problems this one does I can only conclude it is because it is drawn down over autumn and winter to provide capacity for the spring snow melt, and so it fluctuates very much more than a natural lake and hence has badly different dynamics.

            What appalled me however was the cursory treatment of issues like these in the preliminary EIA. When you consider how Dinorwig has fared so poorly, with the complete destruction of its native Arctic Char population and ongoing turbidity, this is inexcusable. But then, the entire process is opaque and politically driven, and doesn’t not augur well for well founded PHES elsewhere.

          • Mike Westerman

            PS – the preliminary EIS in the FS was the most pathetically inadequate effort I have seen for a long time, given that this is a mega project in Australia’s most important land based national park.

  • RobertO

    Hi All, Some of the comments below miss the point of Snowy 2. It oversized storage (some 9 days’ worth) is not useful today, but given that Hydro can last 50 years or more and if we are going down the pathway of RE then we will need good storage like this for when we have a run of poor RE days. When people say it’s going to cost us 30% or 40% of the energy in the cycle, they are forgetting that if you have to turn a wind or solar generator off you lose 100% of that energy. If the Snowy Board can convince its three shareholders (Vic, NSW and Fed Gov) by only forgoing profit payments (that we as consumers are paying anyway and I do not trust Mr Paul Broad) and they release both details and plans related to Snowy 2 and Snowy 3 and Snowy 4 then maybe I can support it. If Transmission is built through to CRE zones then maybe yes rather than no. I do support the idea that at the start if our NEM is only 10% average RE then it’s a poor supporter of RE, however if 1500 MW of active coal is withdrawn from the NEM because of Snowy 2 (it a numbers game of selling at the highest point and buying at the lowest which as we add solar may become the hrs from 10 am to 3 pm and if wind adds at the same time they may be able to buy at negative prices) the RE % in the NEM increases. 1 post below says “whom is going to add this next GW of solar, I would not want to be the one!’ misses the point of both business and households add solar because of the grid prices. We may finish with 6 million generators in this country (single solar systems on roof of many different sizes). In a RE system we will still need some gas as backup. And as for the prices I am of the opinion that they will not drop in the next few years (I have told my bosses that I expect 10% rises year on year for 5 years). Our Gentailers have too much market power take Stockyard Hill and the said price of “Below $60 MWhr” to which you need to add in the cost of buying the LGC (selling at $80 approx.) making an effective price -$20 and we a consumers are paying $240.00 for that same MWhr (and yes some of that money goes to the Transmission (ruled by AEMC whom want their share of household solar).

    • George Darroch

      It’s an insurance policy for a high penetration NEM.

    • wideEyedPupil

      9 days powering the entire NEM or 9 days delivering a very modest 200 MW export?

  • PaulC

    I can understand the arguments about project cost, size, and input supply here, but isn’t there also an important impact of reducing price spikes during peaks and outages?

    The 2Gw is an extra peak supply as well as supplementing any RE shortfalls. So arbitrage in a well-functioning market might prove difficult but in the current dysfunctional market, the logic for Snowy2 is broadly similar to the Hornsdale reserve isn’t it?

    And does this also crowd out peaking plants? If so, this might push down peak prices quite a bit vs. gas peak generation ?

    I suspect that MT is much more worried about high prices and system black events than the pure economics (which can be fudged) and carbon reduction (which we know isn’t a consideration!). Most importantly, he can be seen to be doing something, much like the Howard era model of pursuing some ineffectual action as a sop. If that’s the model I could understand the logic – deeply flawed as it is.

    • Mike Westerman

      It may well eliminate some of the gas but the remainder will set the price (and Snowy 2 will be careful not to dispatch so much that the price drops too much – in fact it will have more tools to manipulate prices). When it is pumping, it adds significantly to demand which will pull up the price of power during that period. Since we are a long way short of saturation with solar, that for sure means it will buy coal fired power at night, and occasionally wind. Meanwhile, in SA, which effectively is a separate market because of transmission constraints, will arrive at solar saturation in the near future, while having to rely on expensive gas for reserve. So it is obvious to any objective observer that the first PHES for >30y should have been built in SA. But that would have sent the message that this government was genuine about RE and PHES. Instead Frydenberg/Turnbull rubbish SA and announce this albatross.

      • PaulC

        Like I said: “ineffectual action as a sop”.

        They could of course not worry about profits and dispatch more power to lower prices though. So the big question is how will they choose to act as an economic agent – rationally or altruisticly? There is a recurring theme in the grid where public assets have been extracting rents for governments short of cash. So I can’t argue for altruism here.

        Mostly though, the 5+ years (estimated) to build buys time. If we’re optimistic, there could be new interconnects and it could deliver better outcomes. For now though, we should all see it as nothing more than window dressing – a fig leaf to cover the embarassing lack of coherent policy.

        Politics 101: whenever possible, kick the can down the road!

        • Mike Westerman

          It certainly has all the wrinkles to produce delays: the Franklin was stopped by Greenies who were hated by the woodcutters and other who supported the HEC. In this one, I suspect it’s only a matter of time before the Greenies fire up, for good reason given the idea of dumping spoil in the existing lakes, clearing 80km of tracks and cutting a new very wide multi-circuit transmission line thru the park, but maybe in a strange alliance with Fishers who won’t want trout fishing in the upper lakes destroyed. When Greenies and Farmers got together they stopped coal seam gas in its tracks in NSW and Vic, so we will see!

        • Greg Hudson

          Remember, we are talking about Snowy Hydro here… One of the biggest ‘gaming’ culprits in the power business (IMO). You think this will change with SH2? Not bloody likely.

  • Malcolm M

    If Snowy 2 is at least 6 years from completion, PPAs for large solar could then be down to $35/MWh. So if Snowy arranges contracts for 1000 MW of large solar for pumping, it can then sell into the peak and firming markets at $75/MWh and above. There would be no need for batteries or firming on the solar, because if a cloud affects output the pumps just stop.

    • Mike Westerman

      Yep – in principle. So it makes a margin of $40 on 16,000MW per day, and say that happened everyday even weekends, it generates $234M/a. Even without counting any O&M costs it still hasn’t made a 0.2% return over 30y. So clearly when Turnbull says its economically viable he is thinking of a subsidy that Snowy 2 will get but that others wont.

      • Ian

        What is snowy 2.0 supposed to represent? If it’s to be a backup generator when wind and solar fail completely right across the NEM then it is too small: 2 GW trying to supply 35GW , just ain’t gonna cut it. If it’s a daily time shifting battery for coal or renewables then it’s not much different to other batteries, the storage component is fantastic, but it’s too large for time shifting capability. Most of the time it would cycle for about 8 hours to 24 hrs: ie 16 GWh to 48GWh frequently used capacity divide that into $4 billion: $250/kWh to $83/kWH. Not bad at all, but considering the efficiency is about 70% round trip over time lithium batteries with a round trip efficiency of about 90% may be more economical.

        This project comprises tunnels bored through the ground joining two existing dams, why not double or triple the capacity say 6GW : $12 billion, 24×6 = 144GWh still $83/kWh

        Hopefully when the tunnels are made they will be large enough for an up grade, if the piping is made large enough for 6GW generation/pumping, but used for 2GW turbine/pump sets initially then the round trip efficiency may improve – less velocity of flow = less resistance. The engineers optimising this system are probably very smart, hopefully they will be allowed to run the numbers for a much bigger capacity as I have just suggested.

        Another consideration is that this Snowy 2.0 system could be used for once through hydro – why bother returning the water up hill there is 350GWh in the top tank @ 2GW this will last 7 days ,sell this water for 15c/kWh: $52million a year Not very good. Mike’s figures for pumped hydro of $40/MWh x 16000MWh/day = $234 million a year is 5.85% per year.

        PaulC’s idea would probably get this project across the line at least initially. He is saying that with 2GW of dispatchable power there will never be a price spike again: it might reduce the average wholesale price by $10/MWh, 190TWh sold each year would save $1.9 billion a year! ie that amount of money would not be available to other generators, that’s going to put a big damper on coal and gas. MT might just shoot his supporters in the foot with his Snowy 2.0 how funny.

  • epicycler

    the headline for this article is Snowy 2.0 only makes sense if we have more renewables.

    That is precisely why we need it. To facilitate more renewables. Chicken or egg. We need both.

    Whether a very big centralised repository makes sense in an increasingly distributed network is just another perspective on it. I guess its whether we want big central selfish storage or little distributed selfish storage.

    Snowy 2.0 will be big enough to be a price maker. All the way until 100% renewables at peak.

    • Mike Westerman

      I would have thought we firstly need honest government and rule maker with less of an ideological bent to give the RE market some confidence. But secondly we need storage where the level of penetration of RE is approaching critical and where options for reserve are limited. SA is the only location at present in that situation. But also if RE generators own their own storage they can leverage more out of it.

      • epicycler

        second point first. Build Snowy 2.0 before the RE penetration becomes critical. A strategic move. Assume RE is cheaper. It takes longer to construct Snowy 2.0 than the renewables.

        yup, we have not just ideological rules (that’s a recent addition), the rules have generally been written at too low a level of abstraction, written around technology at the time of writing rather than round the underlying requirement. Huge impediment to change. Common systems problem.

        “you want to connect a what to our network?”. (pun intended).

        Part of why some functions of the big battery are contracted to SA Government rather than AEMO. Part of why its “paired”. A way of demonstrating where rule changes are required. Common approach to changing a system.

        Storage doesn’t really get a mention in the rules. In the management and control systems its treated as generation for peak capacity planning. Its always full when required? Energy constraints are vaguely considered as an afterthought.

        There’s a difference between the store we put stuff in and the stuff we put in the store. A difference between storage and inventory. Somewhere in the future the market will allow them to be separated. I’ll pay you to store my energy. Self-storage for energy.

        Then the RE Generators won’t need to own the storage, or can share collectively owned storage, to leverage it. Without the economic analysis of Snowy 2.0 its difficult to know but it could be partly founded on “rent seeking” (income in excess of what is required to keep it operating). One of the problems of markets.

        • Mike Westerman

          If we were going to build preemptively but strategically that still wouldn’t point to Snowy 2 – I would do a first round of 1GW/6GWh of PHES across 5 or so projects in SA plus FIT incentives for evening peak export from rooftop solar (to encourage a further 1GW/2GWh of storage) – SA is then in the clear and a net exporter 90% of the time. Second round would be North Qld (which is already underway) and similar FIT incentives everywhere in NEM (would give u potentially 10GW/20GWh storage). Third round is 5-6GW/40GWh along east coast of NSW plus 1GW/10GWh of PHES outside NP using Snowy ponds but not owned by Snowy.By that stage you are at 20GW/78GWh of storage and 40GW+ of solar and wind so I’d reckon the jobs done for now. All by 2030. And hopefully around 2025 we start phasing out FF transport.

          • epicycler

            for both Paul and Mike. Nice to talk.

            when I look at where we are now I see market failure. I would have started with market repair. That’s if I even believed that a market could be designed to optimise the supply of electrical energy for network and financial efficiency through two disruptive transitions. Markets are messy and the results sub-optimal.

            In many circumstances economies of scale don’t benefit me as a consumer.

            Unfortunately the pragmatist in me suggests that Snowy 2.0 will do the job. That the transmission system will evolve around it and the AEMO’s Renewable Energy Zones.

            If Snowy Hydro tries to game the market with Snowy 2.0 then either other entrants will arrive with different storage models or the renewable generator companies will demand market rule changes. Probably both. I hope Snowy Hydro doesn’t believe because its State Government owned its immune to change.

            I suspect that electricity networks will evolve to look like any other industrial supply chain. With inventory to balance supply and demand, and relieve distribution bottlenecks, and cover planned outages, and insure against unplanned outages, and ….. its as if we are building from both ends at the moment, big storage in Snowy and little battery storage at the home, the need for the middle bit hasn’t yet been seen clearly – apart from by you and I. 🙂

            My guess is its not been seen because all the planning models are capacity (power) models. Not a sign of inventory (energy) apart from an obscure reference to energy constraints in the system adequacy reports (ESOO, EAAP, MTPASA, and STPASA). The answer to distribution bottlenecks has been to increase distribution capacity instead of spread the load over time …. how silly is that?

            As well as the physical introduction of storage there is a very big paradigm shift that has to occur.

            My preferred approach would have been to undertake a strategic review of market rules and mechanisms and network management systems. Redefine the purpose around the future not the past. Redefine around the need, and a sense of direction, not the technology.

            I’m guessing that won’t happen any time soon. Based on announcements in the last few days it looks like Australia may be on a course to do a bit better than the business as usual “shuffling deckchairs on the Titanic” approach we’ve had. The result will be imperfect, but I’m guessing it will work. I guess in that respect its light years better than the Titanic model of improvement epitomised by NEG.

          • Mike Westerman

            It’s certainly one of those points in history! The struggle between lawyers who talk about everything but believe nothing, and engineers who talk about nothing and believe something. We engineers are generally a very concrete lot, not given to picking up our tools and moving to the new thing. But we do like prototypes, so the success or failure, and ability to influence the lawyers, will come from SA succeeding with a solution that works despite all it’s up against. Snowy 2 is irrelevant to that challenge because of the transmission constrain and lack of economic local generation, and in fact probably works against it for political and practical reasons.

            But on the bright side, the likely delays to Snowy 2 and a change in Fed government means SA mostly likely achieve it’s goal of a robust RE dominant and exporting local network. It will involve too much state intervention to be called a market, but that shouldn’t mean realistic investment tests are avoided.

          • PaulC

            Completely agree that fixing the market would be logical as a first step along with the planning processes. [Current planning is mostly only via participant actions taken in response to price signals.]

            Trial and error will likely get to a workable grid eventually, but every failed experiment is both a lost opportunity and an increase to costs (especially transmission). Unproductive assets get built but not written down for decades and who pays all the while?

            So why then – when there is now an announced AEMO planning activity – do we still insist on more trial & error – announcing major discretionary projects with questionable business cases? Answer… because there are too many cooks and because AEMCs repeated failure to reform the system invites political intervention. Camels and committees spring to mind!

            So absolutely… fix the structure, processes and rules and then build the right assets. And fix the leadership and the aforementioned will just happen imo. How hard can it be?

          • epicycler

            might have to settle for world peace first.

          • wideEyedPupil

            Also you can build a few more CST plants in Australia for $4.1B (which as David says in the article is likely to become 6 or 8 billion given the experience of other engineering projects in Australia of this scale). With 6 CST plants in Australia from Solar Reserve their price tag for a 260 MW CST (matching Snowy 2’s generation capacity but not storage, which is arguably enough because even Snowy wont see out a week of becalmed NEM in winter wind drought with low PV generation) with 13 hours storage would be well down on the LCOE that Aurora is (which we don’t know but Solar Reserve say their future plants in Australia will demonstrate a considerable learning curve and supply chain efficiencies.

        • PaulC

          I think most would agree per Finkel that much of the energy problem can be traced to current grid structure and (lack of) planning together with archane market rules designed for a different age (central generation).

          So that’s the big question here… why build a single large PHES project which concentrates market power [even in a storage market in future] and goes against the notions of CREZ in that it implies central rather than distributed generation?

          Is it really cheaper because of scale? And does a large scheme present better consumer outcomes? Can’t say I’ve seen evidence for either of those assertions thus far. More data definitely required on both economics and distribution.

          • Mike Westerman

            A few generalisations can be made: schemes with long tunnels generally lose out to schemes with short ones, schemes with existing pondages win over those needing new ones unless the environmental impact constraints defeat you, economies of scale depend as much on transmission constraints as pure size. Snowy 2 fails on all of these.

        • wideEyedPupil

          Do you have some direct interest in getting Snowy 2.0 up?

          • epicycler

            just the natural inclination of an individual not inclined to take risks with the only planet we have.

          • wideEyedPupil

            Then you’d back small scale pumped hydro energy storage that can be built on previously clear land at a quarter of the cost per MWh stored and distributed around the NEM where it is needed, not all behind one interconnector. And if Snowy 2.0 is to act in the FCAS market (which will be dominated by batteries before Snowy 2.0 is even connected to the NEM if it gets up) it would be much better located near demand centres, not as far away as you can get on the NEM from them.

          • epicycler

            I suspect the value of Snowy 2.0 is it will take 7 days to empty it. And a bit longer to fill it. I think of it as insurance. While pumped storage was historically added to networks for stability the advent of renewables has changed the need. Smaller scale pumped storage, nearer to a diurnal fill/empty cycle, sits somewhere between batteries and Snowy 2.0. The FCAS market will change radically with more batteries and other storage. Once storage exists the marginal cost of providing FCAS is about zilch and prices will reflect that.

          • Mike Westerman

            So what do you think the value of 7days x 2GW is, weighted for the rarity of it happening? Especially when AEMO has contracted for 6GW of curtailment over 2017-2018 summer, at no where near the interest costs on $4.5B!

          • wideEyedPupil

            Prices will reflect cap-ex and op-ex not just a zlich marginal cost! Certainly times are a changin’. 🙂

          • epicycler

            for Mike and Pupil,
            insurance is like that, for rare events. Just looking far enough ahead to when the proportion of renewables approaches 100%.
            One of the problems of the existing wholesale markets is “missing money”. Driven by marginal costs. Why renewables will always undercut fossil in the market.
            At present storage is tied to generation (unintended consequence of market rules written before more storage was envisaged). Unless its changed small scale storage will probably be built around the diurnal demand cycle and tied to specific generation. That seems to be how its planned at present. Which almost certainly excludes insurance.
            It really depends how much we are willing to pay for insurance. Or excess capacity if you like (another interesting market mechanism).
            I tend to think of Snowy Hydro 2.0 as “strategic”.

          • Alastair Leith

            I think fossil gas is our “insurance”. Certainly on the SWIS we already have enough capacity to get well past 50% RE (from 8%) and have improved security and reliability.

          • Mike Westerman

            Absolutely Alastair, the insurance will come from the lowest capex options, since their low utilisation will make high marginal costs irrelevant. Thus OCGT’s running on biodiesel would yield low cost insurance and no net carbon emissions, if curtailable loads (ie zero capex insurance) is not available to cover the contingency.

          • wideEyedPupil

            If you pay $X for a storage system and use that storage to it’s full depth and cycle it once a day 300 times in the year, then that storage system is going to 100 times cheaper than spending the same $X for storage that only gets used 3 times a year cycled to full depth. (assuming full depth discharging only for sake of an extreme example that illustrates the value in using storage not just sitting on an “insurance” policy)

    • Ken Fabian

      Epicycler – I think it will make a good argument for supporting more ambitious large scale renewable projects, including those without their own storage component, even whilst it’s under construction – it can potentially cut through the “but intermittent, needs fossil fuels backup” objections. Hasn’t Turnbull been claiming that is it’s function? It may not be enabling RE by intent; I suspect that this was always just one more way for the LNP to kick the climate/emissions/energy policy decision making down the road past it’s current term of office and past another election, after which it’s luminaries can be more comfortable engaging in opposition as The Opposition.

      It seems to me it will be our job to remind them that Renewable Energy is what Snowy 2.0 is FOR – and to continue to campaign for the end of the giant de-facto subsidy on Fossil Fuels – the perpetual amnesty on externalised costs – as the alternative to continuing up front subsidies for low emissions technologies.

      I think RE will grow regardless but if it is to grow sufficiently to be a real brake on global warming it needs to grow faster and bigger, at unprecedented rates and scales – scales appropriate to that of a world of 8 billion plus people who want and deserve some enduring prosperity.

      • epicycler

        yup, its certainly a very mixed signal from Turnbull. Either a brilliant strategic political move to renewables or a random desperate thought bubble. I’d love to have been a fly on the wall when Snowy proposed it to him.

        Regardless, combined with AEMO plan, I can see an Australia in 10 years time having a very high level of renewables, on a trajectory for more, and fossil generators complaining about having to close early.

        The third leg of the stool is AEMC rule changes to remove barriers.

        Hopefully it adds to other examples around the world of “this is how it can be done”. A very important part for those countries that doubt.

        And yes, RE will grow, it makes sense, just that now it makes economic sense. I’m hopeful Snowy will accelerate it. The alternative of business as usual is too horrible to contemplate.

        Not there yet, keep up the pressure.

  • RobertO

    Hi All, Snowy CEO announced that this project would have an 8% ROI. Most of the comments below are saying that it cannot make any ROI given their knowledge on Snowy 2. Given that we do not have all the information (Business Case) most are a waste of time arguing their case. We know only a couple of things at this time for certain, 1 it being rated at 2000 MW 2 it will have some tunnelling (27 km long) and an underground power station, a penstock 200 meters high and diameter of 20 meters. I was sitting on the fence (50% -50%) for this project. We need PHES (but is this the right one to do?). What transmission upgrades will it need? Does the Transmission upgrade go through a CRE zone? One comment below says this project will smooth out the highs and lows of the wholesale market (is that good or bad?) yet others say this will only be powered by coal only at night time (what about transmission through CRE zone going to SA wind farms? What about day time solar?). What about the environmental aspects others ask? I ask “When did you forget that man always destroys the environment for his own benefits?” If this PHES retires coal are we not better off (Gas because it handles stop start better than coal is likely to remain). For these reasons I am now at 55% for, 45% against, and another reason I have moved is that Snowy 2 has 9 days storage and I ask “What impact will Snowy 2 have on the PHES of Snowy, most PHES have to buy at cheapest time generally over night, but will Snowy need to do the buy each night (and what effect does that have on the business case, or will we see Snowy only move when wholesale market goes negative or very low prices.

    • Mike Westerman

      Robert I think we can accept physics: there are only 24h in a day, and the scheme will need to pump 1.5x as often as it generates because it’s 67% round trip efficient. So on average it generates 40% of the time. That means a limit of 19GWh per day but you’ve paid for 365GWh – a bit like a Ferrari engine on a bicycle: fantastic on paper, bit of a dud in practice. Worse if you’ve stolen that Ferrari from the Ferrari F1 team.

      • RobertO

        Hi Mike Westerman, Why will it need to pump all the time to return to full. What stops it from running down the total storage when prices are high and then spending a couple of weeks when it pumps more than it needs to get the levels back to near full. It becomes a maths equation of balancing useage verses pumping times 1.5 all to equal the 365 GWhr storage volumn. I do not think it would run for 7 days continuously (but it could) and then spend the next 10 – 11 days pumping (maybe if they got heavy rain only on the top dam, it would be less). I suspect that it will do something like 2 hr a day less pumping (say 4 Generating and 4 hr pumping or losing 2 hr a day storage. If Electricty prices are low or even negative they may choose to pump more, and it could be driven by how much wind/solar is hitting the system. As for paying for the storage these dams are already there and if we were at 95% RE then Snowy 2 would be a fantasic idea. If Snowy 2 cost you energy then some people claim that is a total loss, we should not build it, where as I say if you have a wind farm, or solar farm which is a better loss 100% at the farm or 37% on PHES. When you can get the COALition to support you to do a PHES system in SA then I will support you totally, but do we have time to wait for that, or is it better to start on something now, when in the longer term we will need some of these bigger storage (days worth). Also where or what is Snowy 3 and where and what is Snowy 4.

        • Mike Westerman

          Robert regardless of the pattern of pumping and generation, the facts remain that it can’t generate more than 40% of the time, and that means on average 19,4GW/day and that means it can never generate enough revenue to provide an 8% ROI unless there is a hidden agenda to provide it with alternative and unique revenue streams just for being there. If that is the case, then why aren’t they being made available to all comers? The money being proposed here could reduce the cost by almost half for Powerwalls of equivalent storage for 1.4M households with solar, which would give them a payback of about 8y.

          SA has about 1GW of PHES projects, some already with ARENA support, that will go ahead regardless of Turnbull’s support, and beat his project to market – that will hurt I’ll bet!

          Destroying the environment to save the planet is never a good idea. If Turnbull was more than an amoral motor mouth, he would have asked Snowy for projects that weren’t in the National Park but used Talbingo for example, but that would’ve raised the question as to why you’d give a state owned company that special consideration and the market power to develop them.

          • Brad Sherman

            Destroying the environment? The proposal uses two existing storages, Tantangara and Talbingo, and just shifts water between the two. I’m not aware of any proposed increase in storage capacity of the reservoirs, just the removal of the current bottleneck for transferring water from Tantangara through Eucumbene and onwards to Talbingo.

            The principal environmental risk of pumped hydro (speaking as a water quality person) could well be upstream transmission of invasive species of disease or invasive species. Tantangara is the headwater to the Murrumbidgee so there is, theoretically, a risk of transmission to that portion of the ‘bidgee upstream of Blowering Dam. Below Blowering one can reasonably expect whatever is in Talbingo to pass downstream. I don’t expect any major issue in this regard.

            Regarding the operation of pumped hydro, I feel the discussion here may be conflating energy efficiency with hydraulic efficiency. The round-trip efficiency of of67

          • Mike Westerman

            Brad merely using existing reservoirs does not absolve the project of rigour in examining the impacts. The very nature of pumped hydro makes it a very different proposition. The issue of redfin perch bearing EHN has already been raised. Given the connection of Tantangarrra to Eucumbene and other important trout habitats, this is an issue.

            You point on the operations overlooks the way this project is being financially justified: what is promoted is an enormous ability to supply power for a prolonged period, ignoring the fact that there are other much cheaper ways of providing for standby, the most obvious being overbuild of solar and construction of offstream, localised PHES. That makes sense, Snowy 2 does not.

            All that a very large increase in installed capacity and storage at the Snowy does is increase its market power in the peaking and caps market. Quite apart from wrongful development in a very significant national park.

          • Brad Sherman

            I don’t think I ever suggested absolving SH2 from rigorous assessment. BY THE WAY, I know plenty of fishos who’d love to see trout eradicated, by the way, although I doubt they’d see their replacement by another invasive species as a good deal. These people all prefer native Australian warm water fish! Me, I prefer to eat trout.

            Without all the financial info on the table, I’m not prepared to pass judgement one way or the other. It certainly feels expensive to me. You could deploy 400,000 14 kWh Tesla Powerwalls for the same money (very approximately!) but only get a fraction of the nominal storage.

            I am deeply suspicious of anything proposed by the current government given their determination to undermine renewable energy at every turn.

            I certainly agree that clearing land for another transmission line and constructing new roads in a Natl Park should be heavily scrutinised. It’s hard to imagine circumstances that justify it, especially if there are numerous alternative sites that might provide a similar benefit without the environmental consequences.

          • Mike Westerman

            Brad perhaps I miss interpreted your response! My point re Snowy 2 is that a) the timing and manner of the announcement was suspicious b) it ignored the most obvious need for storage in general and PHES in particular at the time, which is SA c) it blithely assumed a mega project in the most significant national park in Australia was ok with comments and the preliminary EIA reinforcing this attitude d) it seems intent on misstating the financial viability by referring to the total storage rather than the marketable storage e) it seems to be grossly understating the technical risks given the length of the tunnels and size of the underground cavern.

            As you say, coming from this government the project is tainted at the outset. My hope is that other serious developers of PHES will find appropriate projects, build them expeditiously and within budget, and market them profitably and show up this one for what it is.

          • wideEyedPupil

            Turnbull and Frydenberg got in a pissing contest with Weatherill and this was the best comeback they had, ours is bigger than yours (and we get to trash some National Park while we’re at it and that will wedge greenies — that’s got to be a win in itself).

            This doesn’t stack up and the NEM isn’t at the penetration of RE where this is required yet. Existing gas can and does balance existing and approved wind and solar easily ATM. Build PHES in SA, the only place that needs it so far.

      • wideEyedPupil

        Yeah, nine days of storage but not nine days of storage of NEM’s demand, nine days of storage to output 200 MW. Seems a total mismatch to me, and is all about proving coal some place to sell and to cover the risk on unscheduled outages from plant explosions etc (but even 200 MW isn’t a huge amount of coverage given multi-unit 1.8 GW coal plants in Australia)

  • Ian

    This project appears to use two existing dams without destroying more habitat, that’s got to be a good thing.

    • Mike Westerman

      Not really – Tantangarra is very thermally stratified and anoxic at depth. It is also in a sensitive alpine national park of enormous significance – an environment not able to be “offset” as there is so little fairly inaccessible and unspoilt land like it. If we wish to exploit existing reservoirs in the Snowy for some reason tho that reason escapes me, then do so outside the park.

      • Ian

        Your points are noted. Hopefully those designing this scheme will debate this properly so that we don’t damage the alpine area further.

        What should be studied carefully is exactly how much percentage solar and wind can be installed before storage is needed. Given that the NEM is a huge geographical area and given that there are lots of opportunities for demand management.

    • Mike Westerman

      There is high ground to the west of Talbingo for example, requiring only 4km of tunnels and allowing one dam and intake/outlet to be bored in the dry. A 120ha reservoir could provide 2GW for 6h with only a low wall. If you really thought you could sell 8h of generation then the wall would be a few meters higher. Of course it’s not a NP so there would have to be negotiations to acquire the land, whereas in the NP they just pass an Act to extend the lease for the Snowy.

    • wideEyedPupil

      I would not assume 27 km of tunnels, access roads, removal of 27 km of tunnel fill, new transmissions lines and construction crew access means “without destroying habitat”, quite the opposite I’d suggest!

  • Gerg

    Blakers’ generic stand-alone pump storage price is not $0.8m/MW for 6 hours storage (Figure 2), it’s $0.8m/MW for *zero* hours storage. You need to add $0.07m/MWh for the storage component in his pricing model, so 6 hours would be $1.22M/MW. Also, your European scheme prices are old and don’t appear to include escalation (e.g. Goldisthal would be nearer 2017$1.4B). Changes the picture a little.

    • Mike Westerman

      And bear in mind Blaker assumes that out of 22,000 candidate sites there are 200 first rate sites that will be optimal. I think $1.22M/MW means very optimal! They wouldn’t be having 24km tunnels or management in an alpine region inaccessible in winter. But it does underline why these mega projects should be well down the priority list.

    • David leitch

      Hmm I may have mis transcribed the Blakers numbers, I’ve certainly allowed for the storage costs in the spreadsheet.

      Thanks for the update on Goldisthal.

      Really appreciate those corrections. Thanks Gerg.

  • Rodent

    Dear all
    I am a new poster, my first time, although i have followed the comments made by everybody with great interest.
    I have some observations/questions re Snowy 2.0, but first I have noticed the Generation of Snowy Hydro does not every seem to be very much, often between 87mw and several hundred daily? I then read The Annual Report to gain a better insight into the existing Generation. Report says Hydro Generation was 5437GWh (with 4694GWh for 2016) Further down in the Report it says that Snowy Hydro 2.0 will- increase the Generation of the iconic Snowy scheme by approx 50% making up to 2000 Megawatts to the NEM. BUT then it says -the 300GWh of Energy Storage would be used to ease pressure- .Given that 300GWHh is only about a 5.5% increase in energy storage thaten paying Billions for that amount of increased capacity seems poor value for the Dollar? can anybody explain this outcome, or am i just confused?

    • Jonathan Prendergast

      The 300GWh is the storage capacity. It can use this many times over the year to generate 300GWh many times over. At 2,000 MW capacity, it could supply say 10 hours per day (other 14 pumping/charging) seeing 20GWh per day generation, or 7,300 GWh per annum. This is more than doubling current generation. In practice it is likely to provide a fraction of this daily, and a higher proportion in times of low renewables.