That ‘Notification’ widget alert we’ve referenced before in ez2view triggered on Saturday morning 3rd January 2026 at 07:46 (NEM time), as captured here:
A 296MW drop in ‘Market Demand’ in South Australia in a 5 minute period is outside of the ‘normally expected’ bounds … which is why we set the alert at that level. A quick look at NEMwatch in this snapshot at 08:05 (NEM time) shows that there was some oscillations in ‘Market Demand’ around that time (of which the 296MW drop was the largest leg):
There are a few possible reasons why such an oscillation (including the drop) might have occurred – including:
Suspect #1) First suspect is some fluctuation in rooftop PV injections … which occur ‘behind the meter’ (from the perspective of the wholesale market) and so might have driven a corresponding
Suspect #2) It might have been some form of data glitch … we’ve seen a few of these in the past couple years? But the system frequency remained relatively stable, so does not suggest something like that.
Suspect #3) With the final report of the Nelson Review fresh in mind, perhaps there was some form of aggregated response of invisible embedded Customer Energy Resources?
Suspect #4) or some other reason …
We’ll leave this article here at this point.
The shape of AEMO’s estimated actual PV curve for SA looks pretty jagged even at half-hourly level. Fast changing clouds likelt part of the story.