Snowy Hydro 2.0: More Expensive Than Battery Storage
In the Snowy Mountains, a monumental machine has begun to carve its way into the earth creating a giant tunnel for what will be one of the world’s largest pumped hydro energy storage systems.
The Snowy Hydro 2.0 scheme is so ambitious politicians may be legally required to wear two hard hats when posing for photo ops.
The project will build giant turbines able to supply 2 gigawatts of power deep underground. Only three existing Australian power stations can provide more. Snowy Hydro 2.0 will store so much energy that — technically at least — it could supply two gigawatts of power for days.
Two ways the project will benefit the environment are:
By storing surplus renewable energy, it will improve solar and wind generation economics and aid their integration into the grid.
Its ability to supply dispatchable power will allow the closure of two gigawatts of coal capacity — or potentially more thanks to its higher reliability.
With these great benefits, it’s a pity it looks like it will be a massive waste of money.
That’s because by the time it comes online, using batteries plus additional solar, and wind generating capacity will be cheaper.
This article will look at how Snowy Hydro 2.0 stacks up against large scale battery storage — both existing and soon to be built. I’ll tell you now it doesn’t look good. The best we can hope for is the project will be a cream elephant rather than a completely white one.
Despite looking like it will be a giant waste of money, some people argue that without Snowy Hydro 2.0 the Coalition would be doing nothing at all for the environment, apart from expressing aspirations, and it’s better for them to waste money on an expensive project than do bugger all.
That may be true, but there would be a greater economic and environmental benefit if the government spent the money more wisely in ways that would make us all better off.
I Refuse To Call It Snowy 2.0
Before going any further, I’m putting my foot down and refusing to tack the stupid point zero on the end of its name. From now on, as far as I’m concerned, it’s just Snowy Hydro 2.
It’s not a computer program. If it doesn’t work, we can’t patch it or put out a new version. It’s serious real-world construction with serious real-world consequences.
I’ll only agree to call it 2.0 if the number after the decimal is the number of people who die in accidents on the project. This would cause the original Snowy Mountain Scheme to become Snowy Hydro 1.121+.
Snowy Hydro 2 Outline
Snowy Hydro 2 doesn’t involve building new dams. Instead, two existing reservoirs will be connected by a massive tunnel 24km long. Electricity will be generated when water flows from the upper to the lower reservoir and energy stored when the water gets pumped back up again.
The distance between reservoirs is well under 27km. They aren’t going to add unnecessary zig-zags or make a detour to avoid Hell.
The 27km figure is the total length of all tunnels, including a 2.6 km exploratory and construction tunnel. You can see the massive machine that has begun drilling it in the latest monthly video report for the project:
Snowy Hydro Company Strangely Incompetent
You would think Snowy Hydro, the government-owned company in charge of the Snowy Mountain Scheme, would be able to get the tunnel length correct.
The CEO’s remuneration is four times greater than the Prime Minister’s, so you’d expect that would pay for someone at least four times as dedicated to avoiding embarrassing mistakes. Since that’s clearly not the case, the next obvious step is to try paying him five times as much.
An Underground Power Station
The generating and pumping will occur 850m below the upper reservoir level in an underground power station inside a giant artificial cavern.
This may be carved out of the natural rock, but if the rock turns out to be unstable — a definite possibility — it will instead be housed in an underground chamber reinforced with massive amounts of steel and concrete at an incredible cost to prevent the mountain from crushing the puny humans for their humongous hubris.
Completion Date Unclear
According to the Snowy Hydro Company, testing will begin in 2025, and the scheme will be complete in 2026. But this may be optimistic, and some dam people who understand these damn things say completion in 2027 is more likely. Accidents or lousy geology may push commissioning even further into the future.
Power & Energy
The Snowy Hydro Company says the scheme will:
Provide 2 gigawatts of power
Store 350 gigawatt-hours of energy.
In my opinion, I’d say they got one of these correct. Maybe I should say they got 1.0 correct.
Before I explain more about the power and energy Snowy Hydro 2 can supply, if you don’t know why a kilowatt and a kilowatt-hour are different.
Snowy 2 Energy
The Snowy Hydro page says the scheme will be able to store 350 gigawatt-hours of energy:
Snowy 2 Power
The planned generating capacity of Snowy Hydro 2 is 2 gigawatts. Provided 2 gigawatts of generator capacity are installed, and the water tunnel is wide enough, that’s how much power it will provide.
There are only 3 power stations in Australia that can provide more power, and they are all giant coal-burning monsters.
But if all the hydroelectric dams of the original Snowy Mountains Scheme have their power output combined, it comes to just under four gigawatts. So Snowy Hydro 2 increases the potential power output of the entire scheme by just over 50%.
Snowy 2 Energy
The Snowy Hydro page says the scheme will be able to store 350 gigawatt-hours of energy:
But this is not the full story. If you are going to say the scheme can provide 350,000 megawatt-hours — 350 gigawatt-hours — you have a responsibility to point out it can’t provide that much in normal operation. While there is more than enough water in the upper reservoir when it’s full to provide that much energy, I’ll point out the usable volumes of the two reservoirs:
Upper Tantangara reservoir: 215 gigalitres.
Lower Talbingo reservoir: 116 gigalitres.
The upper reservoir is considerably larger.
This means we can’t use all the available water in the upper reservoir without losing water from the lower one. You can’t pour a schooner into a teacup without spillage.
The lower reservoir will never actually be empty. If you tried to empty it, you’d get too much mud and yabbies in the final flow. This means the normal working amount of energy storage will be much less.
This paper goes into detail and puts it at perhaps 40 gigawatt-hours. While I don’t have time to check the paper’s dam conclusions, I’m willing to accept its figure. But I will note that water can be allowed to spill out of the lower reservoir, allowing Snowy Hydro 2 to supply much more than 40 gigawatt-hours if necessary, which is a useful capability to have.
I’ll also note that 40 gigawatt-hours is fine. It represents 20 hours at its full 2-gigawatt power output. Even in an extreme and prolonged heatwave, there should be opportunities to replenish the upper reservoir in the middle of the day when solar output is high and late at night when both demand and temperatures are lower.
One figure I’ve seen given for the efficiency of Snowy Hydro 2 is 80%. The actual figure is 76%1. This is typical for new pumped hydro.2 Since the water for Snowy Hydro 2 will have to travel around 23 km more than usual, I’m a little surprised its efficiency is expected to be that high.
But Snowy Hydro 2 will be new and huge and undoubtedly have very high generator efficiency. The water tunnel is also going to be extremely wide, which will reduce flow resistance.
You may have seen much worse efficiency figures given for Snowy Hydro 2 than 76%.
These include estimates for transmission losses going in and out. All energy storage will have these losses, although the more conveniently located the energy storage, the less they will be. For example, they’ll be less if it’s located in your home or business rather than the Snowy Mountains.
One advantage of hydroelectric and pumped hydro schemes is their long lifespan. The “design life” of Snowy Hydro 2 is apparently 100 years. All I’ll say is I hope whoever designed it is very young, so we can still give them a hard time if the tunnels collapse in 70 years.
Snowy Hydro 2 Cost
When Prime Minister Malcolm Turnbull announced Snowy Hydro 2, we were told it would cost $2 billion. It will actually cost a lot more, but what’s a few billion between friends? The government has signed a construction contract for up to $5.1 billion, but it allows for additional costs to be added on, so that may not be a limit.
It also doesn’t include transmission upgrades estimated at $1.9 billion for NSW and a lesser but unclear amount for Victoria. My guess is, at the moment, a reasonable cost estimate is around $10 billion. I think we would have to be very fortunate for it to come in at only $7 billion, including transmission. History shows large scale prestige projects are far more likely to come in over budget than below.
Record low-interest rates should have lowered the cost of Snowy Hydro 2. But since no one has been crowing about this, I suspect all it has done is prevent the estimated cost from climbing even higher. Note lower interest rates also improve the economics of battery storage and renewable generation.
Cost Per Kilowatt Of Power
If Snowy Hydro 2 comes in at $10 billion, its cost per kilowatt of power output comes to…
$5,000 per kilowatt
That’s not cheap. Gas generation normally comes to under $1,000 per kilowatt, while Kogan Creek Power Station — the last coal power station built in Eastern Australia — cost $2,100 per kilowatt in today’s money.
Of course, these aren’t examples of energy storage. Fossil fuels have to go in if you want electricity to come out.
Even if we’re very fortunate and Snowy Hydro 2 comes in at only $7 billion, that still comes to…
$3,500 per kilowatt
Cost Per Kilowatt-Hour Of Energy Storage Capacity
Working out the cost per kilowatt-hour of energy storage capacity for Snowy Hydro 2 is not easy. If we use Snowy Hydro 2’s unrealistic figure of 350 gigawatt-hours of energy storage and a $10 billion total cost, then it comes to…
$29 per kilowatt-hour.
That’s really cheap. But if we instead use the lower figure of 40 gigawatt-hours, it comes to…
$250 per kilowatt-hour
The great thing about upper and lower figures like this is, it allows proponents and opponents of the scheme to look at the same information and enjoy hours of unproductive argument.
For Snowy Hydro 2 to make financial sense, it will have to provide energy at a cost that, all up, is lower than the cost of either:
Battery storage, or…
Some combination of battery storage plus additional renewable generation capacity.
This article — which mostly goes over my head — argues that Snowy Hydro 2 won’t be financially viable because wholesale electricity prices have been falling. I’m not going to do any complex accounting like that to prove which is better.
I’ll just point out the obvious and then tell you what I think. If you don’t like my opinion, please feel free to ignore me like an internet date who spots me first.
The Alternative: Batteries + Renewables
Battery storage is the main alternative to pumped hydro. It has the following advantages:
Much cheaper per kilowatt of power output.
Unlike pumped hydro, battery storage is rapidly falling in cost.
The main drawbacks are:
At the moment, it has a much higher cost per kilowatt-hour of storage capacity.
Much shorter lifespan.
Extra Renewable Capacity Cheaper Than Extra Storage
At the moment, large, utility-scale batteries cost over $300 per kilowatt-hour of storage capacity.
If we consider Snowy Hydro 2’s maximum possible energy storage capacity, it comes to only $20 to $29 per kilowatt-hour for a total project cost ranging from $7 to $10 billion.
This makes it look impossible for batteries to compete when it comes to very large scale energy storage, but they don’t need to. This is because the low and falling cost of solar and renewable generation means it’s cheaper to instead build some battery storage plus extra solar and wind generating capacity to.
Reduce the total amount of energy storage required, and…
Ensure the built energy storage is charged as needed despite periods of low sunshine or low winds.
At the moment, the most cost-effective approach is to use a combination of wind and solar energy generation dispersed over a wide area. But as the cost of solar is falling faster than the cost of wind, in the future, most of this additional capacity will be solar power.
Note SnowyHydro won’t just use this extra generation to charge batteries or pump water uphill. They will sell most of it to the grid, resulting in very low wholesale electricity prices on average. Don’t believe any fossil fools who may try to tell you differently, provided it’s clean, cheaper energy is a good thing.
The cost of battery storage has been falling fast. Here’s a graph showing how its price has fallen for electric vehicles over the past eight years:
The dark blue shows the cost of battery cells in US dollars per kilowatt-hour, while the number above the columns shows the cost of cells wired together into battery packs.
Because it’s a bad idea to run an energy storage facility by passing a current through a tub of loose battery cells, we should consider the pack price. In 2020 at today’s exchange rate, it was $178 Australian per kilowatt-hour. In US dollars, they’ve declined 80% from 2013, which is an average fall of 20% per year. While there’s no guarantee their price will fall as fast in the future, they will get cheaper.
Big Battery Costs — Hornsdale Power Reserve
A large utility-scale battery is a lot more than just battery packs, and this makes it more expensive than the cost per kilowatt-hour of battery pack capacity on the graph above.
The best example of a large, utility-scale battery storage facility is the Hornsdale Power Reserve in South Australia. — also known as the “Tesla Big Battery”. It was built within 100 days and came into operation three years ago on the 1st of December 2017.
It has been expanded since then, but when first built could provide 100 megawatts of power and cost around $89 million. This makes its cost per kilowatt of power…
Hornsdale Power Reserve: $890 per kilowatt of power
As it had 129 megawatt-hours of storage capacity, its cost per kilowatt-hour was a little better at…
Hornsdale Power Reserve: $690 per kilowatt-hour of storage capacity
If Snowy Hydro 2 comes in at $10 billion, its cost per kilowatt of power will be $5,000, which is 5.6 times more than Hornsdale. But its cost per kilowatt-hour of storage capacity is much lower no matter if you think $20, $29, or $250 is appropriate.
Snowy 2 Vs. Hornsdale
Assuming Snowy Hydro 2 costs $5,000 per kilowatt of power output, for that cost, we could instead buy:
4 kilowatts of Hornsdale power output.
5.16 kilowatt-hours of Hornsdale storage.
1.44 kilowatts of solar capacity at $1,000 per kilowatt.
If we stack this up against $5,000 worth of Snowy Hydro 2, then the Hornsdale setup has some interesting advantages:
Four times the power output for the same price
Solar generation that can either charge the battery or be sold on the wholesale market
Higher efficiency — likely to be over 90%
The disadvantage is the low kilowatt-hours of storage capacity, which is only 3% of Snowy Hydro 2’s maximum energy storage capacity:
$5,000 of Snowy 2 = 20kWh – 172 kWh of storage (depending on who you believe)
$5,000 of Hornsdale + solar = 5.16 kWh of storage.
But this is not as large a drawback as it appears.
Snowy Hydro 2’s Operating Capacity Will Be 17%
Snowy Hydro 2’s operating capacity is expected to be 17%. This means it will provide 17% of the energy it would if it was magically able to operate non-stop at full power. This 17% figure represents an average of four hours a day at full output, although it won’t always be going flat out.
Its output will be variable, and on some days, it will provide no energy to the grid at all, while on others it would supply much more than average.
The 5.16 kilowatt-hours of Hornsdale storage capacity would be able to come close to providing energy in a similar way. While its total energy storage capacity is much less, it can supply more than the average amount of energy $5,000 worth of Snowy Hydro 2 is expected to.
It will also be assisted by its solar capacity that will reduce the amount of energy the batteries would need to supply during the day and which can also be used to help keep the battery charged. The solar capacity could be expected to supply over one kilowatt-hour on a heavily overcast winter day and over eight kilowatt-hours a day in the middle of a summer heatwave.
If Snowy Hydro used the Hornsdale Power Reserve battery at their expected capacity factor of 17%, then I would expect it to last for well over 10 years.
Its capacity would slowly decline, and after 15 years, it may be at 70-80% of its original capacity. But because it also includes one kilowatt of solar panels, they would use the batteries at less than 17% capacity factor.
This would extend their life, but it would probably make more economic sense to instead provide more energy in total to the grid and put the additional income towards paying for battery replacements that will be needed long before Snowy Hydro 2 starts wearing out.
Note that if part of a solar farm, solar generating capacity can last 30 years or potentially more.
Batteries Aid Transmission
Because batteries can be located where it’s convenient in whatever quantity is desired, they can lower transmission costs. They are often used to avoid the need to upgrade transmission capacity.
For this reason, I am not making the $5,000 worth of Hornsdale batteries pay for additional transmission while it is included in the Snowy Hydro 2’s $5,000 worth of capacity.
The $5,000 of Hornsdale batteries and solar would be able to bring in additional revenue that $5,000 of Snowy Hydro 2 can’t:
It can take advantage of brief periods of high electricity prices by supplying up to four times more power.
It can take advantage of periods of low electricity prices by charging up to four times faster.
Lower losses due to higher efficiency Revenue from its solar component selling energy directly to the grid.
Mainly due to its greater power output, I’d expect this set up to bring in far more in payments per year than $5,000 worth of Snowy 2.
But because Snowy Hydro 2 may come in at less than the $10 billion or so I expect and because I can’t be certain the additional return from the battery setup will be enough to replace them when they fail, I can’t pick a winner.
Feel free to decide which one wins if you like. But note this is only a financial judgement. The Hornsdale battery example comes out ahead environmentally because it includes solar capacity that provides clean energy.
While I can’t declare a winner, it is interesting that Snowy Hydro 2, which probably won’t come online for at least another five years, couldn’t wipe the floor with batteries installed over three years ago.
The Victorian Big Battery
A battery bigger than the one in Hornsdale will be completed outside Geelong in November this year. It’s called the Victorian Big Battery or VBB3 for short. This beast will supply 300 megawatts of power and will have 450 megawatt-hours of energy storage capacity.
How much it will cost is unclear, but a loan of $160 million has been taken out to pay for it. Just to be on the safe side, I’m going to assume it will cost $180 million. This gives the following costs for power and storage capacity…
VBB cost per kilowatt of power output: $600
VBB cost per kilowatt-hour of storage capacity: $400
At these amounts, which are a reasonable estimate of what large scale battery storage will cost this year, batteries beat the mud and yabbies out of Snowy Hydro 2.
The $5,000 price of one kilowatt of Snowy Hydro 2 power output could pay for:
4.17 kilowatts of VBB power output.
6.25 kilowatt-hours of VBB energy storage capacity.
Plus $2,500 left over to put towards future, even cheaper, replacements.
Alternatively, it could pay for:
8.3 kilowatts of power output
12.5 kilowatt-hours of energy storage capacity
Another option would be:
5 kilowatts of power output
7.5 kilowatt-hours of energy storage capacity
2 kilowatts of solar at $1,000 per kilowatt
With this final option, it could supply 5 times the power output of $5,000 worth of Snowy Hydro 2 and close to twice its average daily output from battery storage alone. It would also average around 10 kilowatt-hours of generation a day from solar.
I would expect this setup to have no problem bringing in enough additional income to replace its batteries in the future when needed. These replacements will, no doubt, cost much less than they do now.
Big Battery Costs Will Fall
If my estimate of the cost of the Victorian big battery is correct, it represents a one-third fall in the cost per kilowatt of power output over four years. It’s also a 40% fall in the cost per kilowatt-hour of storage capacity. While I can’t declare a trend from just two examples, if this price fall continues, the cost of large scale batteries in 2026 would be:
2026 cost per kilowatt of power output: $370
2026 cost per kilowatt-hour of storage capacity: $210
There is no guarantee that battery costs will fall this quickly, but I am confident by the time Snowy Hydro 2 comes online, the cost of big batteries will be under $400 per kilowatt of power output and under $300 per kilowatt-hour of storage capacity.
Little Batteries May Be Much Cheaper
So far, I’ve only compared Snowy Hydro 2 to big batteries. But I strongly suspect by 2026 — the earliest Snowy Hydro 2 could be completed — distributed batteries will be providing storage for the grid at what will effectively be very low cost. This will include business and home battery storage and electric vehicles.
If 1% of Australia’s vehicles are electric by then and half of them are connected to the grid at any one time and able to supply 7 kilowatts of power if needed, that would total to around 700 megawatts or around one-third the power of Snowy Hydro 2.
While I’m not saying we will have electric vehicles able to supply that much power to the grid in five years time, we could achieve it with effort. It might take a lot of effort, but it would still be less than what’s required to bore giant tunnels through 27 km of rock.
Keeping assumptions the same, once 50% of vehicles are electric, they could supply 35 gigawatts of power. That’s the maximum output of over 17 Snowy Hydro 2s.
A Transparent Elephant
The Australian government has a clear choice before it:
They can take the easy path of proceeding as planned and allowing Snowy Hydro 2 to continue, or…
Make the difficult choice of admitting they were wrong to support the project and call it off.
So I guess this means our future will have lots of low-cost battery storage and the fourth or fifth largest pumped hydro facility in the world.
If I’m right about the falling cost of battery storage, then Snowy Hydro 2 will become the whitest of white elephants. My guess is it will turn so pale it will become invisible. That is, the Coalition will pretend it doesn’t exist when the topic of wasting money comes up.