The Semi-Scheduled Category

In practice this covers the majority of dispatch intervals.

When the SDC is OFF, the DUID can operate as it likes without needing to listen to* the AEMO’s Dispatch Targets.

* Whilst this is true, in terms of the way the rules are written (i.e. the Semi-Scheduled plant Conformance Status is automatically set to ‘Normal’ when the SDC flag is OFF), there is still a cost applied indirectly to Semi-Scheduled plant operating away from their Dispatch Targets during this period – which comes about through recovery of Regulation FCAS costs using the ‘Causer Pays’ method.

Historically this might have been a very low percentage of dispatch intervals – however in practice this is increasing for some Semi-Scheduled plant as a result of both:
Case 2a)  Being ‘constrained down’ – due to transmission constraints being bound, affecting the unit
Case 2b)  Being ‘dispatched down’ – due to the increasing incidence of negative prices in dispatch (i.e. below the bid prices of some Semi-Scheduled assets).When the SDC flag is ON, then the unit needs to take account of the Dispatch Target when it operates.Readers need to understand that ‘taking account of the Dispatch Target’ in these cases is still different from what’s required for fully Scheduled plant:

1)  For a Scheduled plant, the Target means what the name suggests (i.e. the unit needs to meet the ‘target’ or it is deemed Non-Conforming):
(a)  This test applies if the unit is higher than the Target by a small tolerance (the operations procedures talk about ‘Triggers’); and
(b)  This test also applies if the unit is lower than the Target by a small tolerance (the operations procedures talk about ‘Triggers’).

2)  For the Semi-Scheduled plant, the Target is perhaps mis-named and would be better named as a ‘Cap’:
(a)  The Unit can be deemed Non-Conforming if the unit is higher than the Target by a small tolerance (the operations procedures talk about ‘Triggers’); but
(b)  The unit cannot be Non-Conforming if the output is below the Target (even if the difference is a large MW amount).

At the start of the NEM generators supplying the NEM were registered as Scheduled or Non-Scheduled. The Semi-Scheduled registration category did not exist.

The first* wind farm (that’s visible in production data) commences operations in the NEM.

* re ‘first’ we mean the first wind farm that is visible in production data (Starfish Hill Wind Farm).

The earliest wind farms were (smaller and) Non-Scheduled, but others came later that were (larger and) fully Scheduled.  These pre-dated the commencement of the Semi-Scheduled category (which began in 2009, as noted below).

The Semi-Scheduled category came into existence.  Remember that at this time:

1)   there were only Wind Farms existing at the time.  Prior to the introduction of the Semi-Scheduled category, these Wind Farms were registered as Scheduled or Non-Scheduled (depending on size, and date of registration).

2)  Solar Farms did not arrive until some years later … such as in March 2015 with ‘First production from Nyngan Solar Farm seen over the weekend’.

With the commencement of the Semi-Scheduled category, there were changes in terms of how these wind farms operated.

Bidding

Because some of the larger Wind Farms had been operating as fully Scheduled units to that point in time, their re-registration as Semi-Scheduled meant that there was a substantial reduction in the volume of bids submitted, as can be seen in the chart included in this article here.

Unconstrained Intermittent Generation Forecast (UIGF)

The reduction in the volume of bids submitted was made possible by the introduction of the Unconstrained Intermittent Generation Forecast (UIGF) for each Wind Farm DUID.  This UIGF value then set the Availability figure in the bid used by NEMDE, and the MaxAvail figure in the bid was ignored (up until 7th August 2023, as noted below).

At this time, utilising AEMO’s Version 1 of AEMO’s AWEFS forecast (described in more detail here) to provide Availability forecasts for time horizons including:

Time Horizon #1)  In the ST PASA  time horizon (out 8 days into the future);

Time Horizon #2)  In the P30 predispatch time horizon (out until 04:00 tomorrow or the day after);

Time Horizon #3)  In the P5 predispatch time horizon (out eleven dispatch intervals into the future); and

Time Horizon #4)  In the dispatch interval time horizon (i.e. a forecast, at the start of the dispatch interval, of what was possible for the end of the interval).

As described in more detail on the Glossary Page for ASEFS,  Version 1 of AEMO’s (Large Scale) ASEFS forecast (the AEMO calls it ‘ASEFS phase 1’) commenced on 30th May 2014.

As described in more detail on the Glossary Page for ASEFS,  Version 1 of AEMO’s (Rooftop) ASEFS forecast (the AEMO calls it ‘ASEFS phase 2’) commenced on 30th March 2016.

Trials of Self-Forecasting were commenced with the coordination of AEMO and ARENA – with ARENA calling for expressions of interest on 28th March 2018 from various parties.  That Media Release notes:

‘In partnership with the Australian Energy Market Operator (AEMO), ARENA is seeking to demonstrate wind and solar farms can provide more accurate forecasts of their output into AEMO’s central dispatch system.’

Further details, including some Knowledge Reports, are provided here on the ARENA website.

In particular it’s worth highlighting ARENA’s report from January 2020 ‘Short-Term Forecasting Trial on the NEM:  Progress Report (April to October 2019)’ which they said (p5/15) is …

‘to summarise insights and progress from initial reports submitted by the 11 participants of the Short-Term Forecasting (STF) trial that is taking place between March 2019 to mid 2021.’

We released the Generator Report Card 2018 (GRC2018).

This widely read report analysed aspects of, and posed questions about, the Semi-Scheduled category.

As described here with more context (from September 2019), suitably approved Self-Forecasts were allowed to substitute* for the AEMO-developed forecasts in AWEFS and ASEFS.

*  It’s important for readers to note that these Self-Forecasts only provided an alternate forecast for Time Horizon #4 (i.e. within the current dispatch interval), whereas AWEFS or ASEFS continued providing inputs for the other 3 x Time Horizons.

From Monday 12th April 2021 the AER-requested rule change relating to Semi-Scheduled units came into effect – clarifying dispatch obligations for Semi-Scheduled units under both conditions:

(a)  When the Semi-Dispatch Cap flag is true; and

(b)  When the Semi-Dispatch Cap flag is false … including during periods in which the price outcomes might not be to the liking of the generator.

We’ve collated a series of  articles about what we’ve sometimes referred to as ‘the AER Rule Change’ in this category on WattClarity.

On 9th December 2021 three Non-Scheduled Wind Farms in South Australia were transitioned to operating with Semi-Scheduled obligations.

These three Wind Farms were:

This was the first lot (of 3) Wind Farms transitioned in this way … there were others transitioned later, on 1st February 2022, as noted below.

This was discussed in the article ‘How are South Australia’s non-scheduled wind farms participating in the market now?’ by Marcelle Gannon, published on 21st February 2022.

On 15th December 2021 we released GenInsights21, which explored and discussed several aspects of the Semi-Scheduled category including as follows:

1)   With respect to the Semi-Scheduled nature of their operations …

… we asked the question ‘is the Semi-Scheduled category sustainable or scalable’ (as Key Observation #13 (of 22) within GenInsights21).

2)  With respect to auto-bidding …

… we asked questions (to add in later).

3)  With respect to Self-Forecasting:

(a)  we asked ‘What is the purpose of self-forecasting?’ (as Key Observation #15 (of 22) within GenInsights21).

(b)  In Appendix 6  within the report, we explored in more detail.

On 1st February 2022 three Non-Scheduled Wind Farms in South Australia were transitioned to operating with Semi-Scheduled obligations.

These three Wind Farms were:

This was the second lot (of 3) Wind Farms transitioned in this way … there were others transitioned earlier, on 9th December 2021, as noted above.

This was discussed in the article ‘How are South Australia’s non-scheduled wind farms participating in the market now?’ by Marcelle Gannon, published on 21st February 2022.

As noted in ‘Enhancement of ASEFS and AWEFS forecasts provided by AEMO – from Wednesday 23th November 2022’, AEMO began to use:

1)  Version 2 of AEMO’s AWEFS forecast (described more here) and

2)  Version 2 of AEMO’s ASEFS forecast (described more here) .

Coincident with this:

1)  the AEMO paused any new suppressions for Semi-Scheduled plant that failed the ongoing self-forecasting ongoing assessment; and

2)  notice was given that this ‘Grace Period*’ would end on Tuesday 28^th February 2023:

(a)  ~3 months after the upgrade.

(b)  which has almost completely elapsed now.

* note that the term ‘Grace Period’ is our term, not the AEMO’s.

We released GenInsights Quarterly Update for Q4 2022, which looks specifically at the period covering 1^st October 2022 to 31^st December 2022 (so including the period since 23^rd November 2022).

AEMO 3-month ‘Grace Period’ following the upgrade to ASEFS and AWEFS ended.

This change was requested:

1)   to allow the operators of Semi-Scheduled generators to utilise the rebid process to inform the market of any technical limitations of their plant …

2)  as distinct from utilising pre-existing processes via ASEFS/AWEFS or their self-forecast) to alert the market of any energy limitations in the underlying solar/wind resource.

Articles pertaining to this change are collated here … including articles advocating for the change, and articles analysing use of this functionality.