Batteries cannot make renewables reliable
By Paul Homewood
An interesting analysis from David Wojick, originally published by CFACT:
Utilities are starting to experiment with adding batteries to wind and solar projects. These storage projects are feeding the mistaken belief that batteries can cure the intermittency that makes wind and solar unworkable as a reliable source of power.
The reality is that these battery projects are trivial in size compared to what would actually be needed to make wind or solar reliable. The cost of battery based reliability would actually be stupendous, far more than we could ever afford.
Here are some simple numbers to make the point. The reality would be far more complex, but the magnitude would not change much.
First comes the cost of utility scale battery facilities. This is much more than just the cost of the batteries. At utility scale these are large, complex facilities. Connecting all of the batteries involved and getting them to work properly together is a big challenge in itself. AC-DC-AC conversions are also a big deal, plus there are buildings, transmission stuff, etc.
In many cases these costs are proprietary, but the U.S. Energy Information Administration has surveyed a number of these facilities. See their “U.S. Battery Storage Market Trends,” May 2018.
The reported costs are pretty wide ranging, but the average is close to $1,500 per KWh, so let’s use that round number.
Note that this is the cost per KWh of storage capacity, not the KWh cost of energy from the batteries taken over time, which is a very different matter. There is a lot of confusion on this point. The KWh cost of juice goes down as the batteries are cycled more often, but the cost of the battery facilities themselves does not change. In fact the cost may go up because batteries that can be cycled faster cost more.
At utility scale we are talking about megawatts, not kilowatts, so the battery cost is $1.5 million per MWh. By coincidence, $1.5 million per MW is also roughly the cost of a wind farm. Much follows from this.
A smallish wind farm might have generating capacity of 100 MW, so costing around $150 million. The cost of the batteries to make this farm a reliable power generator turns out to be much, much greater.
Suppose we want to store enough juice to back up the wind farm for just one day, when the wind speed is too low to generate any power. Let’s say we simply need 100 MW for 24 hours, or 2,400 MWh.
At $1.5 million per MWh that is a whopping $3,600 million or $3.6 billion. In short, the batteries cost 24 times more than the “backed up” wind farm costs. In fact in this case the battery cost will be the number of hours times the wind farm cost.
This huge cost certainly makes the wind farm unaffordable, but it gets much worse. Under standard conditions a wind farm produces no power around 25% of the time, due to low wind conditions. Low wind periods of up to a week are fairly common, created by stagnant huge high pressure systems. The power battery system has to be big enough to accommodate these long periods of no wind power.
A week has 168 hours so we need 16,800 MWh of battery storage capacity, at the enormous cost of $25.2 billion, just to make a $150 million wind farm reliable. This would obviously be absurd, which makes the whole idea of battery backup absurd. Even if the cost of batteries were to come way down, say by 90%, the cost would still be wildly prohibitive.
The battery systems that are being announced by major utilities are nothing like real backup. They seldom store even an hour’s worth of generated power (at great price). But they are often touted as being a big step toward making renewables reliable. This is either deep ignorance or pure deception.
Batteries simply cannot make renewables reliable. They cost too much.
https://www.cfact.org/2019/04/26/batteries-cannot-make-renewables-reliable/
Comments are closed.
NOT A LOT OF PEOPLE KNOW THAT 
I don’t think I’ve seen any comment that utilising batteries for back up means that the generation capacity must also be increased. When a battery is discharged, then the generators have both to recharge the battery and supply the load.
Ian, it’s a comment I have often made wherever I see Greens advocating battery back-up. I’m sure Paul has said as much in his many good posts on the subject.
I’m amazed that this simple fact is overlooked when it is probably the biggest flaw of using batteries for power leveling…you need to produce excess power (assuming from renewables) to charge the batteries! If a 100MW wind farm only produces 25% of the time, it is only a 25MW hour base load supplier if you have 75MW hours of batteries.
Mike, it’s something Paul has mentioned before: Too many people think that batteries are sources of power – ie: generators – rather than just stores of it.
It is not unlike Joanna Haigh, on yesterday’s R4 PM, telling Evan Davies that wind-farms should be used to make hydrogen which could be pumped through the gas grid and used as a fuel for H Cars or houses. She hasn’t done the sums and Evan Davies takes it all at face value. (Face-palm).
Using storage consumes power in the round trip losses, which appear to amount to about 20% in Musk’s South Australian battery – a figure that will rise as the battery and inverters age. However, the idea is that if a wind farm produces say 30% of its capacity on average, you charge the battery when the wind blows strongly and there is a surplus to current demand, which allows you to discharge when the wind falls back. However, grid batteries are far too puny to do this as anything other than adding a small stream of arbitrage income from charging when the grid is in surplus, and discharging at peak demand prices as a little pimple on the overall operation. They make most of their income from grid stabilisation operations, with fairly high frequency switching between charging and discharging to match fluctuations on the grid that wind farms make much worse.
Every study I have made suggests that it is cheaper to overbuild capacity and throw away surplus energy it produces in windy conditions than to attempt to store it. But the amount of overbuild required to avoid storage being required is simply massive if you have no dispatchable backup generation. You end up throwing away something like 80% of the energy generated, which makes it five times as costly.
In addition to the ridiculous cost, while the windfarm is charging the batteries it cannot supply that power to the National Grid.
Mind you, the batteries could be charged using conventional fuel sources. 🙂
“Mind you, the batteries could be charged using conventional fuel sources.”
Cut out the ’round trip’ efficiency losses of 10% – 20%, and use conventional fuel sources to feed directly into the grid! 😉
Correct. What are hydrocarbons if not a chemical battery? Lets continue pouring the high efficiency, high octane liquid battery called petrol into our engines.
Germany has too much wind power – when the wind blows – that they have to export that power somewhere. Their neighbors are installing costly devices to prevent an undesired dumping of German surplus. Any storage would be good in this situation. That batteries are currently impractical is a separate issue.
Storage only works when it is economical. The main parameters are capital cost per unit of storage, round trip loss rate, rate at which storage can be filled and emptied, and market opportunity. Storage makes more money the more frequently it can be filled and emptied, assuming that there is a round trip margin to be had. The reality is that even pumped hydro – by far the cheapest kind of storage – struggles to make a margin in Germany.
In fact, grid scale batteries trade much more on their ability to switch their output or charge rates very rapidly to balance the unstable output from wind farms. Here’s the South Australian battery operation over the last quarter resolved at 5 minute intervals:
http://nemlog.com.au/nem/unit/HPRG1:HPRL1/20190201/20190501
You can download the data by clicking on the csv link and calculate roughly the battery charge state (you will have to assume a round trip loss rate, which unfortunately in reality is a variable that depends on things like ambient temperatures and how much use it’s getting). You can calculate the money it makes from charging cheap and discharging expensive. One day with spike price can be quite valuable. But sometimes it appears to be doing things that make no economic sense unless you realise it is been paid for its rapid switching stabilisation capability.
The main reason Germany has ‘too much’ wind power is the delay in building transmission lines to the south, due to objections about how and where to site them (e.g. underground when no overground option is acceptable).
Germany operates almost continuously in net export mode except when the sun and wind fail.
https://www.energy-charts.de/price.htm?year=2019&auction=15m&month=4
They are using neighbouring countries to absorb the surplus and to wheel power South. When it all gets too much, prices go negative and curtailment occurs. See Easter Monday in the chart – when demand was of course much lower than normal.
German expansion of wind capacity has now ground almost to a halt. Nominal capacity is now approach 60GW, which is more than demand most of the time. Further expansion can only see rising curtailment rates, partly for reasons of energy surplus, and partly for reasons of grid stability – which underlie the investment in new lignite fired capacity.
The other factor that needs consideration is the fact that I don’t know of one battery type that does not have a finite life. If the real reason for preventing climate change was to save the planet as opposed to impoverishing the West the solution would be Thorium nuclear reactors. The dangers of radiation leaks from this 80 year old technology is minimal and the risk of meltdown, non-existent. The problem for the alarmists is that solar panels an wind turbines perpetuate the propaganda due to their ‘in your face’ visibility.
“80 year old technology”
Wut? It has never produced a kilowattsecond of electricity.
The myth that grid-scale battery storage is going to make intermittent renewables viable stubbornly refuses to die. Hopefully this article is another nail in its coffin.
The Manhattan Institute recently issued a similarly scathing article on the “magical thinking” that goes with renewables, see https://www.manhattan-institute.org/green-energy-revolution-near-impossible. The author uses Tesla’s Gigafactory, the world’s largest battery factory, as a benchmark. According to Google its total battery production is about 20 GWh per year. The 2017 UK wind electricity production was 50 TWh so the battery capacity to storage a week’s worth of that supply would be about 1 TWh which would need about 50 year’s worth of Gigafactory total production, in other words a totally infeasible amount.
In their latest “Net Zero Emissions” report the Committee on Climate Change waffle on about the “rapid cost reductions” of batteries. They want to increase UK offshore wind capacity from 8 GW at present to 75 GW in 2050 which will make the required battery backup capacity about 8 times even more infeasible than now.
A year or so ago Tesla installed the “world’s biggest battery” in South Australia to fix regular blackouts triggered by short lived transients (nothing to do with long-term becalming backup measured in days) which had a capacity of just 129 MWh and cost a reported $US100 million.
Any calculations about the amount of battery storage needed and the cost needs to take into account the increase in electricity demand that would result from the electrification of transport and especially heating. Given the massive difference in demand for heating between summer and winter we would need to store months worth of power, not just enough for a week or so of calm weather.
I would love to know what the carbon footprint of manufacturing batteries is. If it is true that we need to reduce global emissions by 50% by 2030 then it makes no sense to increase emissions by making billions of batteries even if it was possible and affordable.
The use of batteries for grid-scale storage of electricity in a decarbonised economy is totally infeasible, end of story. The Manhattan Institute paper referred to above notes that about 50–100 pounds of raw materials have to be energy-intensively mined, moved and processed for every pound of battery produced.
As well as the impracticability and costs, even more fundamental is whether there would be sufficient resources available to produce these immense quantities of batteries, even if there was enough processing capacity to refine them. And what would be the environmental cost of extraction?
“And what would be the environmental cost of extraction?”
It won’t take place in any of the virtue signalling “developed” countries, so it won’t matter. Just like claiming that you have reduced your CO2 emissions, when in reality all you’ve done is export manufacturing (and the associated emissions) elsewhere…
I was under the impression that the batteries real function is to buy time for conventional generating plant to ramp up and provide reliable power and so only need to last as long as that takes. That’s also the function of the STOR diesels. And this is all good except that the proper generating capacity is slowly being reduced.
The panacea of batteries as storage can be ‘sold’ to the public.
The primary benefit of grid batteries nowadays is for voltage control & grid stabilisation. The causes of which are too much highly-intermittents and reducing ‘spinning reserve’ from conventional plant.
But batteries can’t be ‘sold’ to the public on that basis, because it means admitting what causes the need for them!
Correct. Batteries with rapid load switching capabilities help to allow a higher proportion of wind generation to be grid connected without the grid falling over into blackout.
The grid has to keep supply and demand very closely balanced at all times. When they are not balanced, then the imbalance primarily results in variations of grid frequency, but can also result in voltage fluctuations and fluctuations of the phase difference between voltage and current. Normally the grid handles the shorter term fluctuations by relying on the spinning inertial flywheel energy stored in generators and motors, which automatically is stored and released in response to frequency changes and power imbalances, with more prolonged imbalances over a few seconds producing a change in fuel input to compensate (but that takes time to act – a few more seconds if the change is not too great, and rather more if you have to bring on or shut down additional capacity).
Wind farms are notorious for introducing various forms of instability to the grid. There is a useful discussion of some of them here:
http://drømstørre.dk/wp-content/wind/miller/windpower%20web/en/tour/grid/rein.htm
and more discussion of some others here:
https://www.academia.edu/1299837/Analysis_of_Voltage_Flicker_due_to_Fixed_Speed_Wind_Turbines
Hmm.. astonishing numbers indeed, Paul. Quite clearly if they are correct, then the invocation of battery storage as an answer, or even an ammelioration, of intermittency problems with renewables, is at least an order of magnitude, and likely two or three, out of sight of any practical application.
What is the data for pumped storage (maybe like Dinorwic?). I know that there are big power losses by this route, but would seem to be more likely useful, as the reservoir can accumulate slowly, and can sit there for months until needed. Do you have the numbers on this? The capital costs are straightforward civil engineering, and well known..
Hydro generation can supply “some” energy needed, but not likely anywhere near enough for a long period of wind “out of the working range.” . Also, there is the obvious need of a correct physical circumstance – a steep walled gorge with a fairly narrow exit area for the dam to be placed. After all, you are using a lot of acreage to contain that water, and most rivers don’t flow through deep gorges, they flow through rather wide flood plains, and the greater the head the greater the output.
Then there is the need of the water itself being impounded, thus not necessarily available for other uses such as agriculture, and of course, there will be the loss of water to evaporation. I am not at all sure what the output of the Hoover dam is, but I doubt if it is greater – or as great as – a nuclear plant. Then look at the acre-feet of water that it has to store in order to supply that power. In the case of that dam, the river is replacing at least part of the amount flowing through the generators. Now consider that you would have to take at least half that power just to run the down stream pumps to pump that water back into the reservoir, if not more than all of it.
Final thought on that would be considering the “downstream” damage that suddenly being forced to release all that water to make up for an intermittent supply would do to the downstream landscape. When all is considered, “power storage” to smooth out the intermittency of wind is not possible, and the only real solution is continuous, reliable generation.
I don’t know how much it would cost to build it now.
Last year we got 2.5 Twh from Pumped Storage, so a bit under 1%
It would probably make more sense to use Norway’s storage, once the interconnector is built. But they obviously won’t do it for nothing.
Here is an analysis of using Norway’s storage:
https://euanmearns.com/how-much-wind-and-solar-can-norways-reservoirs-balance/
the conclusion:
But where will this immense amount of storage be found? One thing is for sure; it won’t be Norway.
Dinorwig can store about 9GWh. In the past 24 hours, pumped storage (there are a couple of smaller sites as well) provided 4.1GWh, and presumably pumped perhaps 5GWh or so during the night (the round trip loss is about 25%). It has a peak capacity of 1.7GW, and its peak use today was at around 8:30 a.m. at 1.243GW, and its fastest ramp rate over 5 minutes was 356MW at 8:15 a.m.. Its main use is peak lopping and grid stabilisation, and it depends on being able to cycle a large fraction of its storage daily to make money. The longer you store, the less economic it becomes – you make money by increasing turnover.
Dinorwig cost £425m in the 1980s. The problem is that there simply aren’t many sites that can be used in the UK (unless you are prepared to flood Scottish glens on a grand scale). One that has been in planning but never made an economic case is Coire Glas:
http://euanmearns.com/coire-glas-the-raging-best-of-pumped-hydro-storage/
It is the problem everywhere. Maybe a PR campaign to use the lakes recreation-ally as well, but there are probably few if any, except maybe in Scotland??
Even here to make it worthwhile it would need to be enourmas with a good drop………oh well
Battery backup means people will die an hour later.
According to the fantasy report from the CCC this week, batteries cost $200 / kWh and will fall to $73/kWh by 2025. This drivel is part of the UK planning for the future!!
The only “batteries that do actually work look like 2 lakes at different elevations. Water is pumped uphill when there is excess capacity and run downhill when power is needed. The lakes take a lot of space!! We have a “battery” like this near Fresno Californis. Most folks do not think of the lakes as energy storage…….
Train to nowhere: run a train full of rocks up a hill with surplus electricity, then back down when there’s high demand.
ARES energy storage technology employs a fleet of electric traction drive shuttle-trains, operating on a closed low-friction automated steel rail network to transport a field of heavy masses between two storage yards at different elevations.
https://www.aresnorthamerica.com/grid-scale-energy-storage
Done that one too:
http://euanmearns.com/is-ares-the-solution-to-the-energy-storage-problem/
Reblogged this on ajmarciniak.