Tasmanian demand drop on 16 August 2025 – part 2

We follow on from A 292MW drop in “Market Demand” in Tasmania, on Saturday 16th August 2025 enlisting this “part 2” to understand the sequence of events using 4-second data.

To recap, the part 1 article noted:

• 292 MW demand drop in Tasmanian demand to the 14:15 interval
  • This is a forecast of demand to be met by the market (hence ‘market demand’). It is an interval ending figure set at the start of the interval, therefore the disturbance must have occurred between 14:05 and 14:10 for it to get picked up into the dispatch of the 14:15 interval.
• Basslink tripped
  • The market notice informed us that at “1409 hrs 16/08/2025 there was an unplanned outage of Basslink.”
• No market notices informing of load shedding were issued.

Data suggests frequency disturbance then Basslink trip?

The chart below captures

• The flow on Basslink.
  1. It was flowing southward at about 310 MW.
  2. At 14:09:32 (first dotted line) readings show it to be operating at that ~310 MW level.
  3. It wasn’t until 14:09:48 that the trip registered in the MW series “Basslink MW”.
• NEM South frequency
  50. Spiked to almost 50.1 Hz 14:09:32 (first dotted line).
  51. It was otherwise stable above 50 Hz.
• TAS North frequency dipped early in the measurements (14:09:36).
  • A later dip also occurred at 14:11, appearing unrelated to Basslink flow, not explored further in this article.

But the measurement timestamps may not be aligned

We understand that SCADA (measurements such as these) are timestamped at the time of receipt of the data into AEMO’s systems. If there is any latency in the sending from network element to AEMO the timestamp will be off. To muddy the water further, that latency and lag-effect will likely be network element-specific. Such that data from Basslink might be received with a different latency to data from the Hazelwood BESS, for example.

In this case Basslink timestamps appear to be delayed. Basslink was acting as a supply source for TAS. If that were to trip, frequency in TAS would reduce with the associated loss of supply. This is what we see in the TAS frequency series, indicating Basslink tripped first and that triggered the frequency disturbance in TAS (310 MW represented 30% of market demand around that time). We didn’t observe a trip of any other large supply sources in TAS (chart below on MW deviations confirms this).

Therefore, we suspect the Basslink MW data is slightly delayed relative to the frequency readings.

Supply-side deviations from target increased at the time of the event, but were small relative to the loss of the Basslink flow

The next chart shows how units (those with a deviation greater than +/- 2 MW at any point) did generally increase output around the time of the event. We’ve kept the same dotted lines to help assess timing relative to the previous chart.

Yet the TAS Aggregate series (from the supply-side units) shows the net increase in deviation was from roughly 35 to 50 MW.

This 15 MW deviation (could be uncertain if latency impacted timestamps) would not be enough to compensate for the loss of 310 MW from Basslink.

Hence, we anticipate a reduction in load somewhere in TAS made up the difference. And of course, that drop in load was reflected in ‘market demand’ in the next interval. The sequence of events must have been close together as the frequency disturbance lasted for only about 8 seconds in the measurements.

SPS and FCSPS

Its likely that the Basslink System Protection Scheme (SPS) was triggered. We understand this would have entailed the Frequency Control System Protection Scheme (FCSPS) acting to control system frequency by quickly shedding load after the Basslink trip.

A TasNetworks + HydroTas document on managing a high-renewables system was the first found in my search to glean further detail on the scheme:

“Contracted load blocks and generating units that participate in the scheme, are automatically ‘armed and disarmed’ as necessary to meet the calculated requirements. If Basslink flow is interrupted, the armed loads or generators are tripped in protection clearance time (within hundreds of milliseconds). The scheme allows system frequency to be maintained within the operational frequency tolerance band limits as defined by TFOS, even though Basslink could be operating at up to 630 MW export or 478 MW import.”

TFOS: Tasmanian Frequency Operating Standard.