Snowy 2.0 cost blowouts might be OK if the scheme stored power more cheaply than batteries. But it won’t

Two years ago, Snowy Hydro announced a reset for its troubled Snowy 2.0 giant pumped hydro project amid cost blowouts. The supposed final cost was A$12 billion.

Last week, Snowy Hydro acknowledged this figure was no longer viable after a cost reassessment.

I estimate the final cost will be well over $20 billion, excluding new transmission lines – more than ten times higher than the original estimate of $2 billion.

As costs have climbed, Snowy 2.0 has lost supporters. The remaining defenders include former prime minister Malcolm Turnbull, who signed off on the project, the federal Energy Minister Chris Bowen and we know of one academic engineer but there may be more.

The question has always been whether this scheme is worthwhile. Despite repeated cost increases, its few defenders continue to argue that Snowy 2.0 offers much cheaper storage per kilowatt-hour than a battery would deliver. In this argument, the cost of Snowy 2.0 is set against the energy storage potential and then compared to the cost of installing chemical batteries to deliver the same storage.

This defence is simple but wrong.

Moving water

Storing energy such as in pumped hydro schemes and electro-chemical batteries is necessary to decarbonise electricity supply, as they make it possible for surplus renewable energy to be stored and used later.

Snowy 2.0 is a major new pumped hydro project that will become part of the existing Snowy Hydro scheme. It can be thought of as a “water battery”.

In the Snowy 2.0 scheme, water is pumped uphill from the Talbingo lower reservoir to the Tantangara upper reservoir when energy is cheap, and then runs back downhill through turbines to produce power when prices rise and more power is needed.

Talbingo in turn gets most of its water from Eucumbene Dam via the existing Tumut 1 and Tumut 2 generators.

So far, so good. But there are three practical complications:

• Talbingo is the upper reservoir for the 1,800 megawatt Tumut 3 pumped-hydro station, which means it needs to be kept near full so Tumut 3 is available to produce at maximum capacity and efficiency

• Talbingo is only two-thirds the capacity of Tantangara and hence can’t accommodate all its water as is the case for a typical pumped hydro system

• the downstream pondage for Tumut 3 (Jounama) is just one-sixth the capacity of Talbingo. So, depending on the water level in Jounama, Snowy 2.0 and Tumut 3 power generation has to be limited so as not to unintentionally lose water.

The end result is that if water is to be kept within the Talbingo/Jounama system and not lost be being released down the Blowering dam, filling Tantagara and then releasing it is heavily constrained by other elements of the system.

That’s not all. If Tantangara was full and Snowy 2.0 generated flat out for seven days, virtually all the water emptied from Tantangara would be lost downstream of Jounama and would then need to be replenished.

Whenever Snowy 2.0 is generating flat out, the Tumut 3 generator would also need to be generating to make use of the flowing water. But this would flood the power market, driving prices down and hence reducing the income needed to recover Snowy 2.0’s investment.

As a result, Snowy Hydro has no incentive to operate Snowy 2.0 in this way, and will almost certainly withhold its full capacity from the market just as it does now with Tumut 3.

Pumping water uphill

Snowy 2.0 faces economic constraints as well.

It takes energy to pump water uphill from Talbingo to Tantangara. Pumping will only be done when electricity prices are cheap, which will usually be for a few hours each sunny day when price are low. And it will only make sense to fill Talbingo from Eucumbene Dam by releasing water through Tumut 1 and 2 into Talbingo when prices are high.

The result: cost-effectively filling Tantangara will take many months.

Now let’s look at the demand for Snowy 2.0’s service. Defenders claim its ability to discharge power for a week is an advantage. But since Australia’s National Electricity Market began in 1998, there’s never been a period when the extremely high prices needed to make Snowy 2.0 worthwhile have been sustained for more than a few hours continuously.

If the energy market ever sees sustained, multi-day periods of extremely high prices, the market response will be to quickly build gas or diesel generators and add more batteries. Both are inexpensive, representing a tiny fraction of Snowy 2.0’s cost per kilowatt of added capacity. Greenhouse gas impacts would be inconsequential, given the generators would be very rarely used.

As a result, the vast bulk of Snowy 2.0’s storage capacity will sit unused in Tantangara because it is so difficult to cost-effectively fill Tantangara and there’s unlikely ever to be the demand to fully discharge it.

Chemical batteries are outcompeting water batteries

Now compare Snowy 2.0’s operational and technical constraints with those of electro-chemical batteries. These batteries go from charging to discharging in a fraction of a second. They do not have any of the operational and economic complexities of situating a new pumped hydro generator in an extremely complex cascade hydro system.

As a result, a kWh of battery storage capacity is likely to be used much more frequently than a kWh of Snowy 2.0 capacity. Grid batteries typically discharge their full capacity at least once per day and often many times a day. Snowy 2.0 is unlikely to ever discharge its full capacity.

So, while batteries may cost more to install upfront, they will be used much more intensively and so their higher costs absorbed over much higher volumes, so that their average costs are lower. It’s the same economic logic seen in the choice between trains versus buses versus cars – trains are usually cheaper per passenger-kilometre when heavily used, but much more expensive if near empty.

This is why battery storage is booming in Australia and many other countries. Private investors are piling in, typically with little or no public subsidy.