Here in SA every new project for “renewable energy” has to obtain approval from the State Planning Commission. The proposals are advertised in the paper and are available to the public via a website. The other day I noticed a proposal for a 100 MW/100 MWh battery to be installed at the major sub-station that served the now demolished Northern Power station at Port Augusta.
It will be called the “Playford Utility Battery Storage Facility” in recognition of the power stations developed by the Playford government and now demolished. The project cost is given as $100 million.
It’s quite a beast. There will be 27 containers full of batteries, each 45 feet long and 27 containers with inverters and transformers as well as a control room and the necessary substation facilities to connect it all up to the grid. It is sited adjacent to the Davenport substation where the HV transmission lines to Adelaide and other areas originate. Incidentally the HV transmission organisation also has a project in for approval at the same site, two “synchronous condensers” worth around $45 million, with another six to be sited in other areas later. The battery facility has a design life of 30 years but the batteries will need to be replaced after 15 years. All battery and other components will be imported as they are not available in Australia.
With all the energy madness that we see every day I thought I would try and compute the battery cost to “firm up” a wind farm. I have chosen the Hornsdale wind farm near Jamestown, (host of Mr. Musk’s big battery) for the analysis. Hornsdale comprises 105 3 MW turbines, (total rated capacity 315 MW, over an area of 7,500 hectares). The promotional blurb says it will produce 1,050,000 MWh annually giving it a capacity factor of 38%, giving a “real” capacity of 119 MW. This is probably a bit optimistic but I will use it anyway.
The tricky thing with trying to get a quantity of storage nailed down is the intermittency on an hourly and daily basis. Just because it is rated at an “actual” 119 MW does not mean it will generate at this rate every day, or hour for that matter. I will use an average output knowing that it really does not represent reality but allows a battery cost to be calculated.
At 38% CF an average days output is 119 X 24 = 2,856 MWh. For one days storage there would need to be 29 100 MWh battery storage systems at $100 million each. This would cost $2.9 billion.
After 15 years, or maybe sooner, the batteries would have to be replaced. Say the batteries comprise 80% of the system cost, then the replacement cost after 15 years is another $2.3 billion. If a weeks’ worth of storage was needed multiply those numbers by 7 and get $20.3 billion and $16.2 billion for replacement. The battery systems would need to be re-charged every day. Some days might not be a problem but on others the wind might not cooperate and they would not fully charge. On “good” days there might be some excess to serve demand but it will always be unpredictable.
A chart showing the output from Hornsdale for the past 24 hours indicates the maximum 220 MW at around 7 p.m. yesterday and the minimum at 14 MW at the time of writing, 2 p.m. on 8 May 2019. About half the day was above 100 but over a third was below 50. The official PR uses an “average” of 119 MW over the whole day.
These calculations are for one 315 MW wind farm. Currently we have 6,106 MW Australia wide “installed capacity” in wind. Using the 38% CF this gives a daily output of 55,687 MWh, so one days storage requires 557 100 MWh battery systems costing $100 million each, a total of $58 billion. In 15 years there would be another $46 billion to replace the batteries giving a 30 year cost of $104 billion and it still would not provide “dispatchable” electricity.
How many coal fired or nuclear power stations could we get for that?