The specific carbon emission data from electricitymap.org reject the claim from the headline:

> https://www.electricitymap.org/zone/AUS-SA

They’re currently at 369 g CO2 per kWh, that’s not clean by any means.

That's coming from the gas plants in SA that are required to provide network inertia

They're currently working on replacing that with battery

(this is explained in the article)

That's false, they're not at all working on replacing gas. In fact, they're quadrupling down on gas (because it's basically a necessity for baseload capacity). [1] [2]

[1] https://www.minister.industry.gov.au/ministers/pitt/media-re...

[2] https://www.murrayvalleystandard.com.au/story/6607101/650m-g...

Is there a date when the existing gas plants will be decommissioned and demolished?

Good question, all the emitted CO² stays for centuries in the atmosphere. What is happening if in the end they can't replace all those "dirty" backups ?

This whole plan will basically work only if this big "if" is solved and deployed.

Reading the article may help. The article states that it was a temporary event.

That's ~30hrs after the point in time highlighted by the article.

How does "synthetic inertia" work? That sounds like a big deal.

Inertia is how well a system responds to a change in frequency.

For example, if a large amount of loads switch off, the system frequency decreases. In response to this, synchronous generators will produce more power to bring the frequency back to normal. This is as a result of the big spinning mass in the generator that wants to maintain its rate of rotation.

Synthetic inertia involves configuring control loops in the battery inverters to cause them to have a similar response to a synchronous generator. The bonus is that you can tune this control as much as you like depending on grid conditions.

> Synthetic inertia involves configuring control loops in the battery inverters to cause them to have a similar response to a synchronous generator.

So you get a clock, monitor the grid, calculate the current that would be drawn from a short-circuited dynamo right now, and get the inverter to generate that current, all within a few microseconds, and without setting the battery on fire? I'm impressed.

Mostly.

https://www.nrel.gov/news/features/2020/renewables-rescue-st...

https://www.nrel.gov/docs/fy19osti/73207.pdf

Just to add on to this, here's a good talk that goes through the basics of VSGs (Virtual Synchronous Generators) and how they work.

https://youtu.be/2BHbUPDxwJc?t=668

Is explained in the article:

Conventional Generators have enormous energy stored in their rotating parts, and the grid ties all these generators together as if they were on the same shaft. This maintains the phase and frequency of the grids electricity by providing "synchronous momentum".

However this synchronous momentum is lost when the mechanical generators are switched off. And so giant batteries and inverters are used instead to provide this critical function.

To put it another way, the giant batteries are not used to power the grid, but instead used to lower the AC impedance of the grid, eg provide synthetic inertia.

This is pure speculation, but I wonder if they are using GPS time signals to synchronize all of the inverters perfectly in frequency and phase. Or maybe they just have them resist changes in frequency like an actual spinning flywheel would, and let phase drift.

In Europe the grid is controlled by the operators, inverters that are grid tied follow the operator commands to increase frequency or decrease frequency. The phase is kept synchronous with UTC.

Do inverters communicate with the grid operator, or do they just monitor the frequency they see on the mains? And how do they all agree on what UTC is with sufficient accuracy? At 50 Hz, a 1% deviation is only 200 uS.

Modern solar inverters communicate with the grid operators, less modern ones follow the grid frequency. The grid is allowed to deviate from UTC by a few seconds, in some edge cases it ramps up to minutes, but it eventually comes back in sync with UTC. Good enough for kitchen clocks, simple timers and the likes.

…and the rest of the time it's good old coal and gaz.

Each nights, when it's cloudy, when it's raining… during all those time periods (so way more than 50%) there's no sun. You can install GW of solar panels, they'll produce during those time close to 0MW of electricity.

Math are already done, "backing-up" things with batteries works to power up for a few hours a neighbor or hospitals. This doesn't work to provide electricity each nights and during cold-not-sunny-winter weeks or hot-AC-everywhere-summer-nights.

Even if you store things in water dam, that is today the easiest, quickest and most cost efficient way of storing electricity. In France where all the dam are already build we basically have only a few days of "electricity storage". Battery is around 2% of the total, so less than a few hours.

If you install "dirty" backup, such as gaz or coal, you're basically erasing all the benefit of installing win turbines or solar panels regarding CO² emission.

If you choose to install nuclear power plants as backup, regarding the building cost and the fact that such plan is having a big inertia (days) regarding the nuclear fission and that the cost stays roughly the same whatever change you're making. You better not use that as backup solution but better than a close to 0 CO² emission source (in France wind turbine is ~60g of CO² per MW, it's 6g of C0² for nuclear plants).