Assault and batteries

Big batteries' are the rage du jour in SA, Victoria, the NT, and Queensland. SA has the Tesla 'big battery' (a bunch of much smaller batteries linked together), and competition to see who can (at taxpayers' expense?) install the 'biggest next thing'. The Andrews Government has announced two 'big batteries' for regional Victoria. There's talk of one or more in the NT, and in Queensland, as well.

These are 'big'. What they do isn't. The SA battery can power 30,000 homes for one hour. The Victorian batteries can power 39,200 homes for one hour. Ignoring industry, for homes this is a drop in the power demand bucket. These batteries apparently have very fast response times, supplying power quickly on demand. That's good for grid stabilisation under demand or supply pressure. We'll need many more if volatile renewables generation increases its power share.

Don't get charged-up. Some physical and cost reality checks about what batteries can and can't do follow.

First, batteries are power receptacles. They don't generate any power. They can time-shift power to periods when it can be used better from periods when it can't. Time-shifted battery power can stabilise the grid (in short bursts). It can augment power supply, but again for short periods, and for a few.

Second, because they don't generate power, batteries must be charged by generators, whether fossil-fuelled or renewables. That power can be used once. If used to charge batteries, it can't be for other purposes. We need extra generation capacity to provide power directly to users, plus capacity to charge batteries when needed. For fossil fuels, it's mainly peak or off-peak supply: less so batteries. For renewables, it's charging batteries. Charging and discharging batteries adds efficiency losses and costs on top of direct supply costs.

Third, batteries aren't free. Their puny output costs. The announced Victorian batteries reportedly will cost $50 million in total, based on initial estimates. The full cost of SA's 'big battery' is unknown. These are large, multi-million dollar projects. Power from any generation source, routed through batteries, must cost more than the same power supplied directly. Batteries cost middlemen margins.

Fourth, the debate about affordability and reliability, versus lower emissions, is all over the place. Governments (now) talk reliability. Others, (eg, ACCC and Grattan 'Institute'), propose less reliability and more affordability (and lower emissions). The National Energy 'Guarantee' is silent on reliability standards. If we must, dealing with the trade-off between affordability, reliability and lower emissions is difficult. Those wanting cheaper, less reliable, power assert not all risk should be eliminated. This is a straw man. The rising share of intermittent renewables is increasing the risk of outages a lot. If we want reliable renewables, we must offset that increased risk. That will be costly. Politicians should say so. They don't. They deny it.

What does power discharged from batteries, charged by various energy sources, cost? Take the Victorian regional 'big battery' project. The $50 million project has a claimed total capacity of 80MWh. Assume a ten-year battery life (it could be less). We can vary depreciation, discount, and general inflation rates, and assumed generator power costs for charging, using fossil fuel, wind power and solar power. Keep it really simple. Spread the implied $625,000/MWh cost equally over every hour of a ten-year period, and add it as a 'fixed charge' to user power bills. This is a simplified 'levelised' cost of storage (LCOS).

While unlikely, assume the $50 million estimate is an all-up cost. Maintenance, plus arbitrage earnings from selling power discharged for more than costs of re-charging, plus subsidies, net out to zero. In reality, maintenance costs will be additional. Net arbitrage revenue with 100% renewables might not be large. With 100% renewables, we don't control supply in peak- and off-peak demand periods. For example, with solar, there'll be no re-charging in the small hours of the morning, as now with coal! Net costs could be higher.

Assume power generation costs per MWh are the same for fossil fuel, wind or solar charging of batteries. I don't think that's true, of course, but I want to isolate costs of using batteries as middlemen between generators and end-users. This assumption is dropped later.

Using differentunsubsidised energy sources:

• Direct power supply from fossil fuels to users is cheapest. No batteries are needed, supply is reliable, and per MWh fuel costs are assumed to be the same as for other sources.
• Battery power charged from fossil fuels is more expensive because of battery costs.
• Battery power charged by intermittent wind power is much more expensive. Even assuming equal generation unit costs, battery/grid stabilisation costs, the need for much more wind power generation capacity, plus additional battery storage, multiply the cost of power (4-5 times that for direct fossil fuel power). The capacity multiplier, all-up, is nearly 6 times fossil fuel capacity.
• Battery power charged by intermittent solar power is more expensive again. With the same generation unit costs, battery/grid stabilisation costs, the need for much more solar power generation capacity, plus additional battery storage, multiply the cost of power more (up to 10 or more times that of direct fossil fuel power). The capacity multiplier, all-up, is over 12 times fossil fuel capacity.