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.

For the last two conclusions, I assumed 2016-17 AEMO 'efficiency' ratings (from SA) of 29% for wind power and 15% for solar power. Elsewhere in Australia, on average, these 'efficiency' estimates could be lower.

I've not allowed for 'unused' or 'curtailed supply' power (see my previous opinion piece). I've not allowed for costs of insuring against renewables uncertainty given their intermittent supply. Dealing with the first of these (adding more batteries) can sometimes lower renewables power costs (net). But adding more batteries delivers such savings at higher power costs. At reduced costs, savings are smaller or negative. Insuring against uncertainty (adding generation and storage capacity) raises costs both for wind and solar.

What unsubsidised renewables unit costs would make them cost-competitive against fossil fuels? Do a 'what if', using the $625,000/MWh cost of Victorian batteries and the 2016-17 AEMO SA results for wind and solar 'efficiency' (29% and 15%, respectively). There are many scenarios. Assume a very wide range of fossil fuel costs, from, say, $60/MWh to $200/MWh, for comparison. In summary (also see chart below):

• Renewables cop multiplied power storage. This 'fixed $ cost', as a %, declines as power costs rise.
• Intermittency also multiplies required renewables generation capacity to deliver the same power, 24/7.
• Competitiveness requires wind generation unit costs to be around 20%-27% of fossil fuels, or less.
• Competitiveness requires solar generation unit costs to be around 5%-12% of fossil fuels, or less.

• Even with $200/MWh fossil fuels, wind and solar must be below $53/MWh and $24/MWh, respectively.
• If fossil fuel costs are in the range $60-$200/MWh, 'new' coal (including HELE) is cheaper, unless the matching wind cost range is less than $12-$53/MWh, or solar generation costs less than $3-$24/MWh.
• If 'old coal' costs $40/MWh, solar generation, after battery costs, must be free to compete. If batteries are free, competitive renewables must cost under 29% (wind) or 15% (solar) of the cost of fossil fuels
• Shifting further to reliance on renewables-fuelled power increases average power costs.

The Australian power user has been assaulted by battery fever. This fever is a complication from cost and reliability consequences of an earlier renewables fever contracted by politicians and passed on to the grid. These maladies now feed on each other. I gather the justification for one NT battery project is as follows:

'he government-owned utility Territory Generation wants the battery to provide contingency frequency control ancillary services (FCAS), reduce the required spinning reserve from its various gas and diesel generators, provide peak shaving and ultimately allow for more solar PV in the local grid.'[Emphasis added]

'Autumnal Cannibalism' is a 1936 painting by Salvador Dali. I think about it in Australia today. In energy policy, and other areas, we seem to be feeding on each other.

As we chew away our efficiency, productivity, and international competitiveness, we might consider policy 'own goals' gnawing at our living standards. Our energy policy's a ball we've blithely kicked into our own net. Affordable power's a receding vision and fond memory, political assertions to the contrary notwithstanding.

Put all this in context. Australia's 1.3% global emissions share is falling as others increase theirs. Alone, can our renewables (or other emissions reduction) efforts deliver even a zero reduction globally? No.

In the Tower of Babel that is our energy policy debate (sic) today, it seems that's an inconvenient answer.