Snowy 2.0 will not produce nearly as much electricity as claimed. We must hit the pause button

snowy hydro

The federal government’s much-vaunted Snowy Hydro expansion is supposed to smooth out the bumps in electricity supply as Australia transitions to renewables. But not only is the project a bad deal for taxpayers, our analysis suggests it will deliver a fraction of the energy benefits promised.

Fossil-fuel power generators store coal or gas at the point of production. This means electricity can mostly be created on demand when homes and businesses need it. Renewable energy cannot do this. If wind or sun is not abundant, solar panels and wind turbines may not produce enough electricity to meet demand. At other times they might produce more than required.

The Snowy 2.0 project is supposed to provide a solution to this problem – storing renewable energy for when it is needed.

The project’s cost and time estimates have blown out massively. It would now be surprising if Snowy 2.0, including the transmission upgrades it relies on, comes in at less than A$10 billion or is finished before 2027.

But there is another serious problem. Our analysis has revealed that of the extra pumped hydro capacity promised by the project, less than half can be delivered. There is now overwhelming evidence the project should be put on hold.

The problems we know about: cost and time blowouts

The list of possible alternatives to Snowy 2.0 is long. Aside from other pumped hydro projects, it includes chemical batteries, encouraging demand to follow supply, gas or diesel generators, and re-orienting renewable generators to capture the wind or sun when it is less plentiful.

But despite this plethora of options, the federal government announced the Snowy 2.0 project without a market assessment, cost-benefit analysis or indeed even a feasibility study.

When former Prime Minister Malcolm Turnbull announced the expansion project in March 2017 he said it would cost A$2 billion and be commissioned by 2021. This was revised upwards several times and in April this year, a A$5.1 billion contract for partial construction was awarded. This excludes the costs of transmission and other considerable expenses.

The main contractor says the project will take eight years to build – bringing us to 2027 before the full scheme is completed. We will happily wager that more delays and cost increases will be announced.

Then prime minister Malcolm Turnbull during a tour of Tumut 3 power station when announcing the expansion in 2017. Lucas Cochairs/AAP

Snowy Hydro has not costed the transmission upgrades upon which the project depends. TransGrid, owner of the grid in New South Wales, has identified options including extensions to Sydney with indicative costs up to A$1.9 billion. Massive extensions south to Melbourne will also be required.

Snowy Hydro contends it should not pay for the new transmission lines because the benefits would flow to the entire grid, not just its venture. In other words Snowy Hydro argues, conveniently, that we should count the benefits but ignore the costs when thinking about their project.

The numbers simply do not add up

The Snowy 2.0 project grandly claims it could generate at its full 2,000 megawatt capacity for 175 hours – or about a week. This capacity can also be expressed as 350 gigawatt hours (GWh).

Energy Minister Angus Taylor has talked up the project’s superiority to smaller-capacity alternatives such as batteries.

But the maximum additional pumped hydro capacity Snowy 2.0 can create, in theory, is less than half this. The reasons are technical, but worth taking the time to understand.

The figure below outlines the main physical features that define Snowy 2.0. It includes four dams: Tantangara, Talbingo, Jounama and Blowering. For simplicity, we have numbered these from 1-4 in the following explanation.

The Conversation, CC BY-ND

When Snowy 2.0 generates electricity, water will be released from Dam 1 at the top of the system. It will flow through a long tunnel to the smaller Dam 2. The flow of water drives turbines which generate energy. When the turbines are reversed, the water is pumped back to the top to continue the cycle.

For Snowy 2.0 to produce the 350 GWh of electricity claimed, the top dam must be full and all that water released through the system. But replenishing the top dam after this event would take many months of pumping water from elsewhere in the system, and use up 40% more electricity than was originally generated. So the 350 GWh would never be achieved because it is extremely inefficient and inflexible.

In reality, the pumped hydro capacity of Snowy 2.0 is defined by the amount of water that the smaller Dam 2 can hold. If the scheme was a closed system, with no other water flowing in or out, it could produce around 230 gigawatt hours (GWh) of electricity.

But the system does not exist in isolation. Part of the existing Snowy Hydro scheme, known as Tumut 3, also uses Dam 2. It creates pumped hydro electricity by cycling water between that dam and the even smaller Dam 3 below it.

For Snowy 2.0 to operate at full cyclical capacity, Dam 2 must be empty to receive the water. That would entail emptying Dam 2 into the smaller Dam 3 and from there to Dam 4 at the bottom of the system. This water could not be used again to generate electricity. This “lost” water would have generated 60 GWh worth of electricity in the Tumut 3 scheme.

Khancoban Dam, part of the soon-to-be expanded Snowy Hydro scheme. Snowy Hydro Ltd
This means that as a cyclical pumped hydro system, Snowy 2.0 does not add 230 GWh of capacity. When you subtract the 60 GWh from the 230 GWh, Snowy 2.0 adds just 170 GWh of recyclable pumped hydro. This is less than half the claimed storage capacity.

And this is the maximum cyclical capacity in theory only. Snowy 2.0 would never produce continuously for the time needed to generate and then pump 230 GWh because it would never be economically viable to run it this way.

In practice if Snowy 2.0’s lower dam is operated in future as it is now – almost always close to full – the cycling capacity of Snowy 2.0 may be as low as 40 GWh – around one tenth of the promised number.

What does all this mean?

These facts put Snowy 2.0 in a completely different light. There are many competing alternatives that can provide storage far more flexibly for a fraction of Snowy 2.0’s price tag. These alternatives would also have far fewer environmental impacts or development risks, in most cases none of the transmission costs and could be built much more quickly.

It is always difficult to press the pause button on a major project once it has begun. But the evidence for doing this is overwhelming. In pursuit of the public interest, the federal government should put the project on hold and ask a reputable investment bank to publicly advise, perhaps through the Productivity Commission, what Snowy 2.0 would be worth if built.

A credible independent valuation would establish with some confidence how deeply Snowy Hydro will have its hands in the public’s pockets. A panel of independent experts should then be asked to publicly advise whether taxpayer money is needed to meet the demands of a renewables-dominated power system, and if so, the best way it should be spent.The Conversation

Bruce Mountain, Director, Victoria Energy Policy Centre, Victoria University

This article is republished from The Conversation under a Creative Commons license. Read the original article.

Comments

8 responses to “Snowy 2.0 will not produce nearly as much electricity as claimed. We must hit the pause button”

  1. Geoff S Avatar
    Geoff S

    Are there really “many competing alternatives that can provide storage far more flexibly for a fraction of Snowy 2.0’s price tag” as claimed ?

    The author states options include “other pumped hydro projects, …chemical batteries, encouraging demand to follow supply, gas or diesel generators, and re-orienting renewable generators to capture the wind or sun when it is less plentiful.
    – Are there any other pumped hydro schemes close to the capacity of Snowy 2.0 ?
    – Chemical batteries are hideously expensive by comparison. Take the cost of the Tesla SA battery: $90 million for 129 MWh. Scale that up to even the worst-case numbers suggested above (ie 40 GWh) and it is many times more expensive than the worst-case estimates for Snowy 2. For 170 GWh it is frightening.
    – Demand management will only make marginal improvements over short time periods
    – Reorienting wind turbines: ahh, isn’t that what they all do already ?

    Sure, the cost might be higher than the politicians want to believe, the capacity less than initially claimed, it may take a few years longer to build, and there are undoubtedly environmental costs. But what are the realistic alternatives that can provide a significant amount of storage (100’s of GWh’s) at an acceptable price sometime in the next five years or so ?

  2. Chris Baker Avatar
    Chris Baker

    The main point of the article is that Snowy 2 will not produce nearly as much electricity as claimed.
    The author goes to some trouble to show that the 175 hours of continuous operation as claimed is not really achievable as a cycled pumped hydro unit. This claim stacks up reasonably but is this really how it would work? And would this overstatement of cyclable capacity have any real impact on its usability or its value to the grid? How much electricity did Snowy actually claim it would make?
    In the Marsden Jacobs report, under the heading How Snowy 2 Would Operate are listed three modes of operation:
    ⁃ Operating in a repeatable way each day pumping 10.5 hours and generating 8
    ⁃ Operating in high output continuous days in a row
    ⁃ Pumping for continuous days in a row
    The questions raised in this article about the actual energy available doesn’t really impact on the main modes of operation. Snowy 2 could still operate as described by Marsden Jacobs, and it is these modes of operation that would describe the amount of electricity that the project would produce. This wouldn’t change in any practical manner so I think the claim doesn’t hold water. The original headline claim about its ability to generate for 175 hours is really just an attention grabbing headline and not a serious measure of the electricity that this project would provide.
    There’s another way to look at it and that is that it will substitute quite well for a coal fired power station, probably more than one. At 2000 MW it should be able to substitute for Yallourn 1500 MW and Vales Point 1320 MW. At present these power stations may produce their nameplate capacity occasionally, but by the time Snowy 2 is producing these two power stations will be lucky to be reliably producing 2000 MW between them.
    When combined with a suitable mix of wind and solar, Snowy 2 could replace those coal stations. On those occasions when there are multiple days of low wind, you’d have the capacity to generate at up to 2000 MW for a few days. This was always a reasonable objective for snowy 2, and would remain so even with a derated capacity.
    The chance of needing to discharge the whole 175 hours continuously would hardly ever arise — it just happens to be its absolute maximum capacity. If it was ever really absolutely positively needed, and the upper dam was full, it remains true that it could generate at 2000 MW for the 175 hours claimed.
    And the actual useful task that it could do, to replace some coal fired power stations, is still achievable.
    And isn’t this what we want?
    No other pumped hydro project on the horizon can do this.
    This task of filling in for those multiple days of low wind is the hardest task of all, and the most expensive, and the thing that might easily have been put off until later. Why not do it now? I think it is a good use of my money.
    The low hanging fruit of demand response and integrating EVs into the grid will still be done anyway. Wouldn’t it be great to spend government money on the most difficult part of getting to 100% renewable, and leave it to the market to bring in the low hanging fruit?

  3. Ian Avatar
    Ian

    The idea behind the Snowy2 scheme is that renewables are intermittent, and backup storage is needed , and that renewables may be unavailable for periods as long as a week.

    We have nearly 10 years of data regarding the generating patterns for wind, solar and hydroelectricity. We know precisely how much day to day, weekly and seasonal fluctuations to expect. The data is nicely displayed in the Opennem graphs .

    The data is remarkable in that the daily fluctuations are very consistent . In the last few years where rooftop and utility solar has been part of the generator mix, there has not been a day without this resource and for the most part, the quantity of solar has not varied much at all. The total wind output has similarly been very consistent from day to day.

    Enough storage to cover daily renewables energy fluctuations is probably all that is really needed. Even an overbuild of wind and solar may not need to be that large to meet weekly and seasonal changes in demand and supply.

  4. Phil NSW Avatar
    Phil NSW

    Now we have the biggest potential player in the reconstruction of our future grid support infrastructure saying their project helps the whole grid and so they should not be liable for any grid upgrade. This allows any project past or present to have the same allowance. All grid upgrades are now the grid operators domain and not the individual projects. Fast tracking the grid expansion therefore should be undertaken as a matter of priority. I can see this idea going along way. Go Angus.

    1. RobertO Avatar
      RobertO

      Hi Phil NSW,

      See, I told Australians it cheaper to keep Liddell going and it cheaper to keep Vales Point going in NSW. There is far too much Wind and Solar and we need to keep coal going, it “Baseload, dispatchable, and reliable power” (and my job is secure until I retire to the Cayman Islands, Climate Change is not real and sea levels will never rise.
      Regards Angus T!

      We should never scare the kids. Lossing a few bugs from the windscreen is not and never will be any sort of issue, you’re just frightening the kids.
      Regards Scott M!

  5. RobertO Avatar
    RobertO

    Hi All,

    Here is a simple question?

    Are the National Grid numbers going to rise?

    For about the last 10 years they have been on a slow but steady decline, so if they are going to rise when will it start?

    Every forecast I have seen is saying “Yes there is an upturn coming and most show it starting today!”

    Am I the only person that thinks the National Grid numbers are going downhill and will continue for about the next 10 years or so, slow and steady decline, even BEV’s (and FCEV’s) will not stop the decline (in fact I suspect in 2024 onwards for the decline to increase slightly as V2G and V2H become acceptable, it a trickle discharge at peak times, and a trickle recharge of BEV’s during high solar times 9.30 am to 3.30 pm weekdays and this is for midwinter. Summertime will be longer available period but not much of an increase in charging and most will BTM.

    There is one possible reason for a rise in the National Grid and I think it is at least 5 or more years away. If we start a program to export VRE (any of the many ways possible). It will need to be very big as most of this is not on the National Grid and most will never be added to the National Grid.

  6. johannes Avatar
    johannes

    “But despite this plethora of options, the federal government announced the Snowy 2.0 project without a market assessment, cost-benefit analysis or indeed even a feasibility study.”
    You’ve watched Utopia, right?

    1. Joe Avatar
      Joe

      But, but, but our Anxious Failure was on ABC 7.30 show this week in an interview with 7.30 show presenter Leigh Sales. All good according to Anxious Failure with the cost-benefit and feasibility study getting tickles of approval.

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