Dispatch of the market’s more complex at present … with SA on its own island (and TAS still in pieces)

Some of our automated alerting has triggered this afternoon (Tuesday 15th November) to let us know that the NEMDE dispatch process at the AEMO is struggling a bit to deal with the increased complexity of the dispatch process with:

1)  South Australia on its own island from Saturday 12th November 2022; and also

2)  The Tasmanian network still more complicated by what happened on 14th October and is still being repaired.

Here’s a quick snapshot of one of our monitors showing file creation latency for ‘DISPATCHIS*’ and ‘P5MIN*’ files has been escalating since the islanding – and particularly today (in part because Intervention’s happening, perhaps?):

Just to give you some sense of the added complexity, here’s two (constraint-focused) widgets from our ez2view software at the 12:30 (NEM time) dispatch interval.

Quick notes with respect to the bits highlighted:

1)  At this point in time there are close to 1,200 constraint equations invoked, or forecast to be invoked in predispatch…

… compare this to more ‘boring’ times in the NEM with a more integrated network and no outages and you might see between 600 and 800 constraint equations listed in the ‘Constraint Equations’ widget on the right of the display;

2)  In terms of bound constraints, which are at the top of both widgets, we see:

(a)  A number of constraints pertaining to Torrens Island … which would presumably be the result of directions to keep the SA grid stable; and

(b)  Off topic, but will be of particular interest to some readers – note that the ‘N::N_UTRV_2’ constraint equation is currently bound (i.e. the one discussed in some detail in this Case Study here).

Challenging times!