Mixed sources of semi-scheduled availability on March 23, 2026

Monday’s market suspension in the NSW region impacted the dispatch availability expectations of semi-scheduled units.

We have a collection of articles about the the event on WattClarity. This article delves into how the source of the dispatch availability changed across the day, focusing on the suspension period.

Dispatch availability can come from different sources

For a background, we suggest: What inputs and processes determine a semi-scheduled unit’s availability.

In this article we focus on whether:

• Availability was suppressed through the energy bid’s MaxAvail,
• Availability was forecast from AEMO from a forecasting system (the simple forecast or AWEFS or ASEFS),
• Availability was forecast from a self-forecast,
• Availability was adopted from SCADA (persistence from interval-beginning output reading),
• Availability was adopted from the last target (persistence from the last dispatch level).

A shape for each dispatch interval indicates the source of availability

We use the following chart to see how semi-scheduled unit availabilities changed their origin. With interruptions to SCADA across NSW, many of the inputs that an availability forecast depends on could have been impacted.

Just enough commonalities as there are differences among the series. Apologies if this drives you to bring  forward your visit to the optometrist! Each row is a semi-scheduled unit.

AEMO > Last target > SCADA > AEMO

Example from main chart

There is a group of units that were originally seeing availability from AEMO (AWEFS or ASEFS).

At interval 11:15 (generally) we presume the necessary SCADA datapoints were not available, the source of availability switched to Last target.

Some time after 12:00 (12:20 in the example) we presume the some SCADA had resumed for these units to permit the source of availability to be based on the initial MW SCADA. One solar unit’s bid MaxAvail set the availability for 4 intervals at this time.

At 13:05 use of AEMO forecasting resumed.

Last target used from 10:20

Example from main chart

Some units had ended up with availability set by the last target as early as the 10:20 interval. This was followed by a brief period of ‘AEMO’  – 10:40 to 11:20 approximately – before use of SCADA-based  (initial MW) availability until 17:05.

 Site data unaffected?

Example from main chart

Some units sailed through the period with either full use of the self-forecast or partial use, switching to ‘AEMO’ at times. We presume the ‘partial use’ came from self-suppression of the self forecast for those units. Closer inspection at at least one of the (in the example, solar) units exhibiting source-switching suggests impacts of being under an intermittently-bound constraint:

A quick introduction to the gyrations on the ‘N>>NIL_060_051’ Constraint Equation during NSW Market Suspension

Really unaffected?

Example from main chart

Some sites under a self-forecast or AEMO forecast saw the source of availability unchanged throughout the period. These are seemingly unaffected from the perspective of availability forecasting.

Some in this category may still have seen some SCADA signals drop out (at least, as indicated by what reached AEMO and ended up in the MMS). Potentially, a self-forecasting system may have been able to access required measurements through a channel separate to the one AEMO uses.

Others in this category may have truly been unaffected for the purpose of availability forecasting (to say nothing of other or broader impacts of the event).