On Wednesday last week (6th September 2023) we had some of our interstate team members in our Brisbane offices, and were focused on some specific projects we’re working on.
… additionally, Dan Lee was busy posting about ‘Long or short on storage? A look into price signals for duration’ on that day (a day following his well-received talk at Smart Energy Queensland).
But with the help of the ez2view dashboards on our display wall in the office, we did notice the unusual case whereby many units in northern Queensland were simultaneously curtailed just before midday on the day … with the curiosity heightened because they were all the Semi-Scheduled (i.e. Wind and Solar) units.
Having experienced this before* but not having time to drill in**, we surmised at the time that these curtailments were due to some form of System Strength constraint … knowing that in the QLD region they are of a form (i.e. single DUID on LHS) such that are only ‘next day public’, so only visible to clients running ez2view against their own Private MMS.
* such as on Tuesday 13th October 2020, as explored afterwards in this series of articles.
** had we drilled more into other widgets in ez2view we’d have seen more information to help us understand this on the day.
(A) Some details published by Overwatch Energy
Readers here at WattClarity might also like to keep an eye on updates from Overwatch Energy on their blog here.
… occasionally we might see articles we copy into WattClarity as guest authored articles … like these articles by James Tetlow.
(B) Using ez2view to explore a bit further
Firstly, readers should note that the Overwatch chart above shows a smaller drop in output earlier that morning … which may have been due to economic curtailment (i.e. spot price below ‘negative LGC’) or due to some other reason.
I’m deliberately ignoring that one.
Instead, with only a bit of time this afternoon, I’ve time-travelled our ez2view software (version 220.127.116.11) back to Wednesday 6th September late morning and (with a combination of different widgets) want to quickly* explore what more I can see of what happened through that bigger drop just before midday.
* Reader Beware: this is a quick review, and there may well be errors as a result! It’s certainly incomplete…
B1) At 11:25 on Wednesday 6th September 2023
Let’s start at 11:25 with this snapshot (remember with all of our articles we try to set them so you can click on the image to open a larger full-screen view in a separate tab):
In particular I have highlighted three particular aspects of the market as follows:
1) Note in the northern part of the QLD region (i.e. the North, Ross and Far North zones) that many units are running heavily at the time.
2) In the ‘Bound Constraints’ widget, we see that there’s only the ‘Q>NIL_YLMR’ constraint equation bound in the QLD region … quite a few more in NSW and so on.
(a) Not shown here, but a quick review of the ‘Constraint Dashboard’ widget shows that this only contains MARYRSF1 and YARANSF1 on the LHS.
(b) Both DUIDs have a factor of 1
(c) But it is YARANSF1 that is being ‘constrained down’ by this constraint
3) And spot prices are currently –$23.82/MWh and forecast to drop to –$45.06/MWh in the P5 predispatch horizon.
B2) At 11:30 on Wednesday 6th September 2023
At 11:30 we see that there are a number of new constraint equations bound, with most of these in two particular (network outage related) constraint sets:
I’ve highlighted a couple particular DUIDs (and the related constraint equations) in the image above, but won’t discuss. Note also that the prices have risen a little, as a result of the reduction in output.
Opening a new window with 2 copies of the ‘Constraint Set Details’ widget focused on these two particular constraint sets, we see the following…
For these two constraint sets, note the following…
|The re-occurrence of
the ‘Q-BCNE_821’ constraint set
|The occurrence of
the ‘Q-CPWU_818’ constraint set
|There are 14 separate constraint equations bound in the ‘Q-BCNE_821’ constraint set. Of these …
1) three constraint equations have a very high Marginal Value.
2) another 5 have a Marginal Value of –$992.22/MWh.
Translating the shorthand AEMO description shown above, we noted that this particular constraint set relates to an outage on …
‘the (line 821) 275kV feeder between (H10) Bouldercombe and (H11) Nebo’
There’s a series of articles written about, and tagged with, the ‘Q-BCNE_821’ constraint set that provide useful background … but there is one in particular we suggest you read for more details, and that is ‘Case Study Part 5 – Why did those constraint bind – in QLD on Tuesday 13th October 2020?’
However the reader is reminded that the image above shows that this constraint set was last changed on 24th August 2023 on the basis of changing Powerlink limit advice … and we’ve not invested time to explore either:
1) What these changes were;
2) Nor have we invested time to see how many other changes were made since the incident on 13th October 2020.
It’s also informative to see that the current outage …
1) Began at 14:05 on Monday 4th September and that it
2) extends out till 11th October 2023 at the 17:00 dispatch interval.
|There are 2 separate constraint equations bound in the ‘Q-CPWU_818’ constraint set. Of these …
1) both have a Marginal Value of –$992.22/MWh.
The ‘Q-CPWU_818’ constraint set relates to an outage on …
‘Line 818 (which is 275kV), which runs between (H67) Calliope River and (H40) Wurdong … and that it relates to CQ-SQ voltage stability limit’
In contrast to the other outage, this one is shorter…
1) Began at 08:05 on Monday 4th September and that it
2) at the time-travel point shown extended out till Thursday 7th September 2023 at the 17:10 dispatch interval:
(a) Noting that (after the time-travel point) the outage plans might have changed…
(b) But a quick review now shows that it did indeed end as planned at the time.
Keep this in mind as we walk through another few dispatch intervals…
B3) At 11:35 on Wednesday 6th September 2023
Stepping forward to 11:35 we see that:
1) Output from a number of these Semi-Scheduled units in northern QLD continues to drop (though not as quickly as on 13th October 2020) as a result of bound constraints within the ‘Q-BCNE_821’ constraint set
2) But the bound constraints in the ‘Q-CPWU_818’ constraint set have disappeared (i.e. no longer bound) – though the two DUIDs that were bound by that one (i.e. KABANWF1 and SMCSF1) are now bound within the ‘Q-BCNE_821’ constraint set.
There’s a number of other things observable here, but I don’t have time to explain:
B4) At 11:40 on Wednesday 6th September 2023
Stepping forward another 5 minutes to 11:40, we see the following:
At this point, we see that many Semi-Scheduled units in northern QLD are now at 0MW, or very close to it.
B5) At 11:45 on Wednesday 6th September 2023
Five minutes later (at 11:45) we see this:
There’s minimal output from any Semi-Scheduled unit in northern QLD, with only SMCSF1 and HAUGHT11 and CSPVPS1 having any non-zero dispatch Target (remember that the QLD Schematic shows ‘FinalMW’ when time-travelled).
(C) For greater understanding …
We’ll leave it there for now … if you’d like to learn more, here are three options:
If you’d like to learn more
of the options
Wait for more details at www.WattClarity.com.au
Use Overwatch Energy
Worth noting that, in this short update from Overwatch Energy via LinkedIn, they noted that they could help their clients understand the implications of events like this (and, where circumstances allow, take steps to remediate).
That’s all for now…