Observation 13/22 from GenInsights21 – is the Semi-Scheduled category sustainable or scalable?

Here on this site, and also in real discussions in the offline world, we have been pondering this question for a number of years.

For instance, it was pondered in Theme 13 (‘What’s the future for Participation Categories in the NEM?’) within the 180-page analytical component of the GRC2018 released on 31st May 2019.  Since that time …

(a)  On 10th March 2020 we highlighted the significance of this question in this article.

(b)  On 28th June 2020 we highlighted the question about scalability and sustainability in relation to the AER Issues paper.

We released GenInsights on 15th December 2021 – containing 22 ‘Key Observations’ inside of Part 2 of the report.    In this report pondered about this question in the form of Key Observation 13 (of 22), and also in a variety of Appendices within the report.  Following from this report …

(a)  On Tuesday 5th April 2022 in this presentation to Smart Energy Council members, we discussed this observation (with respect to Appendix 17 within GenInsights21).  See the table underneath the recording for the specifics of when, during the 60-minute webinar, this was discussed.

(b)  It was also something discussed on 19th May 2022 in the ‘Bang for Buck’ session at CEC Large Solar Forum in Brisbane.

(c)  On 9th July 2022 in the article ‘AER published ‘Semi-Scheduled Generator Compliance Bulletin and Checklist’ on WattClarity it was flagged again.

(d)  On 14th September 2022 at ‘Smart Energy Queensland’, and in this subsequent article, the question was raised again.

This same question is now being picked up with the regular GenInsights Quarterly Updates.

For various reasons we’ve chosen to share here the whole of Observation 13/22 (‘Is the Semi-Scheduled Category sustainable or scalable?’) in its entirety (with links added, for ease of reference) …

————–

From 31^st March 2009, Wind Farms and Solar Farms in the NEM were able to register as Semi-Scheduled generators.

Prior to that time, a small number of Wind Farms had been operating as fully Scheduled generators which had posed significant challenges for them with our markets first few versions of auto-bidding.

It had also meant that a portion of their zero-carbon energy capability was routinely being spilled because of their need to fully comply with obligations to follow their dispatch. At that time (and even in 2021) spillage is regarded as something to be avoided.

The Semi-Scheduled category, following clarification in March 2021 due to a request by the AER, holds that a wind or solar farm unit registered as such:

· Requires the unit to comply with a MW dispatch level contained in its dispatch instructions from the AEMO (i.e. its Target) for all dispatch intervals; whereby

· When a ‘semi-dispatch interval’ is flagged the unit cannot be above its Target – but can only vary below its target by the variation of the wind or solar resource; and

· At all other times, the unit must only be above or below its Target by the natural variation of its wind or solar resource.

Through GenInsights21 several questions are posed and explored, seeking to explore the implications of the way in which the Semi-Scheduled category operates. In particular, we wonder whether measures introduced when wind production was a low percentage of the overall energy mix are still workable, and appropriate, as the share of wind and solar grows to become dominant.

13.1 A review of Aggregate Raw Off-Target

In Appendix 17 we continue an exploration of extreme cases of large aggregate Raw Off-Target across all wind and solar farms.

This exploration showed that semi-scheduled units are more off-target than other units, by the very nature of their fuel source, but also because of the forecasting mechanisms, be it through AEMO’s AWEFS or ASEFS system, or via the newer mechanism, Self Forecasting.

Editor’s Note –> This exploration continues, with this article published 9th March 2023 the latest in a series of analysis shared here on this site.

The next discussions in Key Observation 14 & 15 are also related to the long-term sustainability and scalability of the Semi-Scheduled category.

Editor’s Note –> Key Observation #15 (What’s the purpose of Self-Forecasting?) has also been shared more broadly here on this site a couple weeks ago.

13.2 The challenges of Semi-Scheduled units in a fully hybrid facility

Some current conversations around the registration of batteries embedded with either wind or solar farms seem to be based on an assumption that the wind farm or solar farm would be able to continue operating as a Semi-Scheduled facility and yet also not curtail when asked to for pricing, congestion, or system strength reasons, but rather spill that surplus energy into a back-up battery system.

This level of basic thinking is not compatible with real world NEM operations.

If there is any AC linkage between facilities on the same site (or in the same locality) then this would be incompatible with the desire to establish hybrids that operates independently of dispatch targets and other directions.

AEMC’s recent final determination about ‘Integrating Energy Storage Systems into the NEM’ (noted here) is relevant in this context.

13.3 Resource uncertainty still needs to be covered

Appendix 16 considers the role scheduled generating units play in meeting demand when expected semi-scheduled unit output potential changes with the renewable resource. While increased spatial diversity has come with new units entering the market and this has led to smoother generation potential overall, there are occasions when weather patterns impact vast areas at a time (explored for wind in Appendix 27).

Gaps between anticipated and actual semi-scheduled energy supply have increased as new capacity has entered (at the NEM-level, with respect to expectations for 24 hours ahead). Yet prior to 2020 a complimentary impact did not show up in the long-term trend of scheduled supply. However, it is possible the NEM has reached a tipping point here, whereby supply uncertainty in the semi-scheduled sector is now large enough to impact supply uncertainty in the scheduled sector: Since January 2020, an increasing trend in the average level of uncertainty of scheduled generator supply has emerged.