Concern #1 … DPV Contingency Events

Our reader clarifies that the DPV curtailment that happen in DPV Contingency Events is not due to the Minimum System Load (MSL) requirements; as AEMO’s market notices explain it’s to do with managing contingency sizes.

Normally the risk of a large block of PV tripping in SA is managed by keeping headroom on Heywood via constraints.

But in circumstances like the ones listed in the RHS column, then AEMO may take action to have rooftop PV curtailed to keep the “PV contingency size” within secure limits.

As an analogue, if there were a 600 MW single generating unit in SA then AEMO would certainly put constraints on its maximum output when SA was islanded, or when Heywood was a single-contingency risk.

Hence there are examples of when this occurred:

1)  In October 2022 when there is a Heywood outage limiting secure transfers:

(a)  Sunday 16th October 2022 was one such case in point.  As copied into that article, the AEMO’s message was:

‘AEMO has detected there is an elevated risk of the contingent disconnection of Distributed PV (DPV) which, together with the risk of loss of scheduled generation, exceeds secure thresholds in the South Australia region from 1030 hrs 19/10/2022 to 1400 hrs 19/10/2022.’

2)  In November 2022 when there was the frequency islanding following failure of towers near Tailem Bend in the ‘Heywood link’.

(a)  On Monday 14th November, with that article noting MN103113 speaking about:

‘AEMO has detected there is an elevated risk of the contingent disconnection of Distributed PV (DPV) with the loss of scheduled generation exceeding secure thresholds in the South Australia region from 1200 hrs to 1630 hrs 14/11/2022. The maximum forecast DPV contingency is 84 MW at 1330 hrs and the secure DPV contingency threshold is 80 MW.’

(b)  On Tuesday 15th November

(c)  On Wednesday 16th November

(d)  On Thursday 17th November

(e)  Not, on Friday 18th November?

(f)  On Saturday 19th November

3)  In February 2024, DPV total output is so high that constraints on Heywood aren’t sufficient to manage the size of a proportion of that PV tripping off

(a)  as Linton noted with ‘Too much rooftop PV in SA, almost, as forecast for 15th February 2024’, with MN114740 speaking to:

‘AEMO has detected there is an elevated risk of the contingent disconnection of Distributed PV (DPV) which, together with the risk of loss of scheduled generation, exceeds secure thresholds in the South Australia region from 1230 hrs 15/02/2024 to 1400 hrs 15/02/2024.’

(Possible) Concern #2 … Distribution System Issues

In some cases, disconnection of solar PV might have been an unintended by-product of challenges within the distribution system.

1)  In the recent article in question, we linked back to the article of 2nd October 2023 titled ‘Has rooftop PV in South Australia recently been curtailed with these low points for demand?’ …

2)  In that article we note that we …

‘… have not seen anything from AEMO (e.g. in its Market Notices or Media Releases or News Updates) to suggest that there’s been a coordinated action since that time.’

… so it seems unlikely to be a function of either DPV (Concern #1 above) or MSL (Concern #3 below).

3)  In comments on that article there was a message relayed from SAPN that they …:

‘… have conducted a series of tests on behalf of “Smarter Homes” with the primary objective of ensuring the capability to disconnect these inverters in emergency situations, thereby reducing potential risks.’

… but that some inverters did not re-connect.  So perhaps that was the general cause, in those instances?

Concern #3 … ‘Minimum System Load’ concerns

These alerts are concerned with ‘Minimum System Load’, as described in the AEMO Market Notices at the time.

Our reader explained that, while aggregate DPV output certainly drives this, the issue appears to relate to the total level of system load (grid demand) and the issues that creates for things like voltage control, keeping sufficient synchronous units online etc.

Curtailing DPV appears to be to be one of the few levers available for dealing with the low load issue operationally.

1)  In December 2023 was the first case in Victoria:

(a)  On Sunday 31st December 2023 was the first instance

2)  On 27th and 28th September 2024 was the second instance in Victoria:

(a)  There’s a collation of articles for Friday 27th September 2024;

(b)  There’s a collation of articles for Saturday 28th September 2024.

3)  This current weekend (12th and 13th October 2024) was the third instance in Victoria:

(a)  There’s a collation of articles for Saturday 12th October 2024 (i.e. today);

(b)  There’s a collation of articles for Sunday 13th October 2024.