“Cheap” batteries to the rescue?

The conclusion of the report titled “Fast Erosion of Coal Plant Profits in the National Electricity Market” (see previous post) is that a surge of cheap renewable power capacity could force some coal fired power plants to close earlier. The authors attribute this to “low bids on the wholesale spot market” due to “lower operational cost”. Solar and wind are bidding at low or even at negative prices and coal fired power plants are now competing for those moments when solar and wind can’t produce (much) power, pushing those that are the most expensive to operate out of the market.

The report states a number of things that are not included in that analysis and one of them is the need for backup capacity. The report did however mention several possible backup strategies and it seems that the authors believe the most in batteries as backup strategy, although they did not look into the economic effects of this strategy. To emphasize the viability of batteries, the claim is made that batteries are among the cheapest backup sources and battery costs are expected to decline substantially over the coming decade.

There are two graphs in the report that illustrate this point. The first one is a graph comparing the average short run marginal cost of different generation types (page 16):

Comparison Average short run marginal costs of generation technologies

Just like solar and wind, batteries apparently are among the generators with the lowest Average Short Run Marginal Cost (probably because, just as solar and wind, they don’t have fuel costs).

According to the report, batteries are also the generation technology with the lowest build costs (page 32):

Comparison Build costs of generation technologies

This comes with a caveat though: the build cost is lowest for batteries that could provide two hours backup. Batteries that could provide backup for a longer duration, move more to the left of the graph. The same with other backup strategies: the longer the duration they can provide backup power, the more they move to the left in this graph. Which should be obvious.

This raises the question whether such a two hours backup scenario is realistic? Why do the authors think that a two hour capacity might be enough? They looked at what would be the effects of the closure of the Liddell coal fired power plant (New South Wales) and found that solar and wind would be able to fill in a large part of the daytime demand, but this doesn’t work well in the tail end of the evening peak. This is one of the time slots used by coal (and gas) to eke out a living.

I seen that same principle in a previous post back in January where I played with the grid generation data of NSW: when I increased intermittent capacity while at the same time lowered dispatchable output, I indeed found that a gap formed during the evening peak.

So, aren’t batteries then the ideal cheap backup solution for New South Wales (of maybe even Australia)? If the exit of coal plants only leaves a gap of a couple hours in the evening and two hours battery backup is the cheapest to build and one of the cheapest to operate, then just replace those expensive coal plants by cheap batteries and call it a day…

Well, not so fast…

How certain is it that these gaps will only last two hours or less? Solar and wind are intermittent power sources and also demand can vary (for the better or the worse). Those gaps are not guarantied to last just a couple hours in such a system. Take a look at the first five days from the data that I used in that post:

chart 12a NSW battery model

Gaps are forming every evening, but their duration and intensity vastly differ. Two gaps are rather shallow and one of them is plugged easily (on a Saturday), the other one almost (on a Sunday). The rest of the gaps are larger and couldn’t be filled in by the battery capacity used in this scenario. The gap of the third day (on a Friday) was much larger and deeper. Building for grid security means dimensioning for the largest gap possible and that will be (much) more expensive to build than the proposed two hours capacity battery.

Not only that, there might also be other periods where solar and wind are insufficient. If such a lull follows a gap before the batteries would have the chance of reloading again during the day, that battery capacity needs to be even larger.

This will bring those batteries in the same territory where the coal plants are now. The mechanism that drives the coal plants out of the market is the competition of all those coal (and gas) fired power plants for the time slots when solar and wind power are not sufficient to meet demand.

The same principle also applies to battery backup. For example, let’s suppose that batteries are build that could fill in a six hours gap. This means that when a gap is only two hours, the rest of that capacity is sitting idle, earning no money for their owner. If those gaps are mostly two hours or less over time, than two thirds of those batteries will be sitting idle for the majority of the time, making them economically unviable. Solar and wind power, due to their intermittency, act more like parasites than like fellow power generators and will suck out the life of those technologies that need to fill in their gaps.

Currently this is not a problem for batteries yet. There are still plenty of coal and gas power plants to fill in the gaps left by solar and wind. Batteries are currently not even used as backup, but they tap into the lucrative market of frequency control (that was created by the introducing large amounts of intermittent power in the first place). This could be very different when fossil fuel power sources are pushed out of the market and batteries will have to bear the brunt of the backup themselves.

So yes, batteries for two hours backup might well be one of the cheapest backup strategy available, but it is not exactly realistic to assume that this could provide sufficient backup capacity. Meaning that solar and wind will not only eat the lunch of coal (and also gas), but also that of battery backup…


4 thoughts on ““Cheap” batteries to the rescue?

    1. trustyetverify Post author

      Yep, couldn’t help it 😉

      Glad you found it interesting and learned something.

      Next post(s) will probably on a different subject. I might however come back to this report later, I am for example wondering what would happen if I take the NSW or NEM generation data and model ever lower dispatchable capacities while limiting backup capacity to just two hours…


    1. trustyetverify Post author

      True, batteries need electricity to charge and that could be considered a “fuel cost”, thus a short run marginal cost. However, as I understand it, correct me if I am wrong, the current grid sized batteries are part of solar/wind installations in order to absorb excess production and doing frequency control and arbitrage. If that is true, then I don’t think that they would get invoiced for that power. Even if that would be the case, charging is generally done when there is a lot of excess power and that means buying at low (or even negative) prices and selling at high prices (the Hornsdale Power Reserve earned quite a buck by doing so). Which is bit different from coal or gas always having to pay their fuel at a premium.



Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s