A quick first pass through ENERGY bids (for Victorian units) on Tuesday 13th Feb 2024

We noted yesterday that we’ve started to delve into the AEMO 4-second data … and as such published this ‘14 hours of trended (mainland) frequency on Tuesday 13th February 2024’ …

1)  In the process flagging 7 different periods over that 14-hour period that will be of particular interest.

2)  Some of which we’ll delve further into as part of what I called ‘Timescale #2’ in this earlier list of questions.

In parallel, we’re also looking into market participant behaviours – including bidding.  In this first pass we take a top-down look at bids by Fuel Type / Technology using the ‘Bids & Offers’ widget in ez2view, time travelled back to 00:00 on Wednesday 14th February and looking back 24 hours … such that:

1)  It shows ‘final bids’ (i.e. for each dispatch interval it shows the bids that NEMDE took into account in its dispatch process for that dispatch interval)

… i.e. so we only see the final outcome of rebidding behaviour, not (in this article) looking at ‘that other dimension of time’ in how the rebids were progressively developed.

2)  It is filtered to VIC region only; and

3)  Just looks at the ENERGY market.

This uses an internal build of ez2view version 9.9.1.21, which is being prepared for the IESS market changes that go live in June 2024.

Disclaimers

First and foremost, readers should keep in mind the following disclaimers:

1)  This analysis performed and published on Saturday 17th February 2024 as a quick first pass

(a)  It’s deliberately intended to not show the whole story, but instead to flag top-down things we’d like to delve into in more detail in the days and weeks ahead, time permitting (i.e. we actually have other things to do as well!)

(b)  It’s done quickly, so there’s increase of mistakes in opting for speed.

i.  If any readers spot mistakes, it would be appreciated if you could point them out (in comments below, or directly);

ii.  If we become aware of mistakes we may update this article, but it’s not guaranteed.

2)  We’ve noted below that ‘the NEM is a complex place’, and readers should keep in mind the interconnected nature of the NEM, in several dimensions:

(a)  Firstly, it’s probably obvious to most people that just zooming in on bids in Victoria will overlook any impacts that came in the ENERGY market of other regions;

(b)  Perhaps less well understood will be the ’interconnections’ across all 11 commodities (i.e. ENERGY and 10 x FCAS commodities).

i.  Just focusing on ENERGY will almost certainly overlook outcomes related to co-optimisation.

ii.  As an example of this which I can point to, the increase in mainland Raise Regulation Requirements for part of the day Tuesday 13th February 2024 is bound to have had some impact on other commodities (including ENERGY).

(c)  Finally, worth also noting that this analysis just focuses on the ‘physical’ ENERGY market operated by AEMO, and that there are likely to have been impacts (both ways) with related financial markets as well.

So reader beware!

Tips in reading these ez2view charts…

A few background tips on how to read these charts – which are from the ‘Bids & Offers’ widget in ez2view:

1)  We’ve added the spot price for VIC on these charts to provide you a reference point for the bidding activity.  Keep in mind that (as noted in this ‘sequence of events for what happened in Victoria through the day on Tuesday 13th February 2024 (via Timescale #1)’)  that the spot price:

(a)   spiked for the 13:20 dispatch interval, and

(b)  remained high until it dropped negative in the 15:20 dispatch interval.

2)  However also keep in mind that (also noted in this ‘sequence…’)  load shedding under LOR3:

(a)  commenced for the 14:20 dispatch interval; and

(b)  the Actual LOR3 was cancelled at 15:15 (i.e. last dispatch interval with the price at the Market Price Cap).

3)  Users can create their own bid price buckets, and assign colours of their choice – but it makes sense (at least in this top-down review) to use the ‘Default Price Buckets’ which use three basic colour ranges:

(a)  BLUE = $0/MWh and below (with –$1,000/MWh being light blue … there’s a lot of that!).

(b)  GREEN = Above $0/MWh to $300/MWh

(c)  PINK and RED = Above $300/MWh.

(d)  GREY = eagle-eyed readers will spot a bit of grey, which are ‘Fixed Load’ (i.e. ‘must run’ price-taker bids).

4)  Remember to click on an image to open a larger view in another tab.  Reading through on two windows might help, if you have two screens?

So let’s get started…

ENERGY Bids in Victoria on Tue 13th Feb 2024 – all units

With the above in mind, here’s the trended bid stack for all units in Victoria:

Note that bids for consumption (and aggregate targets for consumption-side units) are shown as negative quantities … but a bit dwarfed by the supply-side contributions.

There’s a massive drop in capacity offered into the market shown that appears to be almost coincident with the spiking of the price … but remember that (also noted in this ‘sequence…’)  the price spiked at 13:20 but the biggest drop in capacity offered did not  occur until the 13:25 dispatch interval.

Hence the slight mismatch in timing we see above.

ENERGY Bids in Victoria on Tue 13th Feb 2024 – coal units

Here’s the trended bid stack for all (coal-fired) units in Victoria:

No bids for consumption with these units.

With this chart it’s useful noting:

1)  The total volume offered at the start of the day was lower than total installed capacity because:

(a)  Yallourn had two units out (YWPS1 and YWPS2), and

(b)  also because YWPS4 was offering capacity lower than its MaxCap.

… something to explore in a further article (Observation/Question #1).

2)  The loss of 2,200MW of capacity offered at Loy Yang A from the 13:25 dispatch interval was a noticeable drop.

… something to explore in a further article (Observation/Question #2).

3)  Nothing immediately jumps out with respect to Loy Yang B

… so let’s move on.

ENERGY Bids in Victoria on Tue 13th Feb 2024 – gas units

Moving along, here’s the trended bid stack for all (gas-fired) units in Victoria:

No bids for consumption with these units.

With this chart we can see that:

1)  Some capacity was unavailable through the whole of the day

… something to explore in a further article (Observation/Question #3).

2)  Of the capacity that was offered to the market:

(a)  At the start of the day all capacity was offered at or above $10,000/MWh;

(b)  But by 14:50 all capacity that was available was offered down at –$1,000/MWh

(c)  and there was an increase in volume offered (which was not a surprise, as some slower-start units that don’t expect to run in a given day take the approach of bidding unavailable)

… all something to explore in a subsequent article (Observation/Question #4)

(c)  What’s notable is that prior to the tower collapse and subsequent spike:

i.   some volume was shifted down to –$1,000/MWh as early as 12:20 (i.e. 60 minutes prior to the spike), hence receiving targets shortly afterwards.

ii.  By the 13:15 dispatch interval (i.e. last before the spike), there was 625MW offered down at –$1,000/MWh … much of it receiving targets.

… hence that’s definitely something else to explore in the a subsequent article (Observation/Question #5)

ENERGY Bids in Victoria on Tue 13th Feb 2024 – hydro units

Finished with fossil fuels,  here’s the trended bid stack for all (hydro) units in Victoria:

No bids for consumption with these units (i.e. the Snowy pumped hydro units are in NSW).

With this chart we see:

1)  Some volume unavailable for the whole of the day

… something to explore in a further article (Observation/Question #6).

2)  A more complex pattern of bids than was shown in the (red to blue) shift for gas above … but the end result being almost all volume offered down at –$1,000/MWh by 14:30

… something to explore in a further article (Observation/Question #7).

ENERGY Bids in Victoria on Tue 13th Feb 2024 – solar units

Into the VRE, here’s the trended bid stack for all (solar) units in Victoria:

No bids for consumption with these units.

With this chart we see a number of interesting things:

1)  Through the middle of the day, peak energy-constrained availability was close to aggregate Max Cap

… which suggests that most plant were available, though still something to be explored later (Observation/Question #13).

2)  There’s a noticeable volume of energy offered in (102MW in GREY) ‘Fixed Load’ up until 13:40 for reasons unknown:

… something to explore in a further article (Observation/Question #14).

3)  There’s a very up-and-down aggregate production shape (i.e. the black line) shown through the day … which:

(a)  could be due in part to the underlying weather system (i.e. fast moving rain bands) … or possibly bushfire smoke; but

(b)  also might be due to network constraints, which we’ll need to use other widgets to explore

(c)  possibly other factors?

… to be explored further in a subsequent article (Observation/Question #15).

4)  Readers will see (remember click on image for a larger view) some staggeringly large ‘Aggregate Raw Off-Target’ deviations in the dotted grey lines that began about 12:30 and grew worse in subsequent hours

… presumably also as a result of fast moving cloud cover (and possible smoke?), but to be explored further in a subsequent article (Observation/Question #16).

5)  I also see a somewhat counter-intuitive outcome in bidding from ~13:40 (i.e. with the spot price is at the Market Price Cap) in that volumes were shifted from –$1,000/MWh upwards to ‘negative LGC’, and even some above $10,000/MWh for reasons not immediately obvious?

… perhaps as a result of some crude auto-bidding logic – definitely something to explore later in a subsequent article (Observation/Question #17).

ENERGY Bids in Victoria on Tue 13th Feb 2024 – wind units

Next in terms of VRE, here’s the trended bid stack for all (gas-fired) units in Victoria:

No bids for consumption with these units.
With this chart we see a number of interesting things:

1)  Throughout the day, peak energy-constrained availability was well below aggregate Max Cap:

… probably just a seasonal low for wind … but I’d like to explore, later, if there were units unavailable in the day (Observation/Question #18).

2)  Until ~09:50 the aggregate production shape (black line) is fairly steady – but from that point the up-and-down aggregate production shape grows worse for the rest of the day… which:

(a)  could be due in part to the underlying weather system (i.e. fast moving wind bands); and

(b)  possibly including high-speed wind cut-out; but

(c)  also might be due to network constraints, which we’ll need to use other widgets to explore

(d)  possibly other factors?

…  all to be explored in a subsequent article (Observation/Question #19).

3)  Readers will see some noticeably large ‘Aggregate Raw Off-Target’ deviations in the dotted grey lines that began about 10:00 and are worse from ~11:00

… presumably also as a result of wind patterns (but interestingly do not appear to be as large, in relative terms, as for solar above) … all to be explored further in a subsequent article (Observation/Question #20).

4)  Factoring in here is the sudden outage of STOCKYD in the 11:40 dispatch interval (which was noted as Event #1 in this ‘sequence of events …’)

… something to be delved into separately in a subsequent article (Observation/Question #21).

5)  There’s also the coincident trip of DUNDWF1, DUNDWF2 and DUNDWF3 in the 13:10 dispatch interval (which was noted as Event #2 in this ‘sequence of events …’ and also by Allan is his first pass on the 4-second data).

… but I have not specifically looked, yet, to see if there were others that dropped offline at this time.

… something to be delved into separately in a subsequent article (Observation/Question #22).

6)  As was the case in aggregate for solar farms above, I also see a somewhat counter-intuitive outcome in bidding from ~14:40 (i.e. with the spot price is at the Market Price Cap and the AEMO has instructed load shedding) in that volumes were shifted from –$1,000/MWh upwards to ‘negative LGC’, for reasons not immediately obvious?

… perhaps as a result of some crude auto-bidding logic – definitely something to explore later in a subsequent article (Observation/Question #23).

ENERGY Bids in Victoria on Tue 13th Feb 2024 – battery units

In firming new technology, here’s the trended bid stack for all (battery) units in Victoria:

In this case the chart looks significantly different because the Installed Capacity for discharge (i.e. positive numbers) appear the same size as the Installed Capacity for charging (i.e. negative numbers).

With this chart we see a number of interesting things:

1)  Throughout the day, peak availability for discharge was well below aggregate Max Cap and the volume offered to charge was substantially more than offered to discharge throughout much of the day.

… I’d like to explore, later, what the reason – e.g. commissioning delays or SIPS or something else (Observation/Question #24).

2)  Throughout the day, especially when charging, there were substantial aggregate ‘Dispatch Error’ ….

… being BESS there are more reasons why this might be the case (compared to Solar and Wind) so am not calling it ‘Aggregate Raw Off-Target’ at this stage – would like to explore further in a subsequent article (Observation/Question #25).

3)  Strangely, when the crisis hits for the period when prices are sky high, the volume of energy offered to discharge was substantially lower than earlier in the day … off the top of my head, I wonder what the reason(s) for this might be:

(a)  Were they actually empty;

(b)  Was this somehow to do with FCAS provision?

(c)  Or is this another example of what we wrote about with respect to the Energy Crisis in 2022 where energy limited plant (with batteries having the most severe limits) withdrawing capacity to preserve what little they energy they have?

(d)  Or is it something else entirely?

… all to be explored further in a subsequent article (Observation/Question #26).

ENERGY Bids in Victoria on Tue 13th Feb 2024 – negawatt units

For completeness (an remembering that we already know registrations in WDRM has been lacklustre), here’s the trended bid stack for all negawatt units in Victoria:

Readers are cautioned here to remember the mental gymnastics required in Tripwire #3 in that a ‘negawatt’ is a supply-side unit (i.e. positive number) not a consumption-side unit (i.e. negative number).

With this chart we see a several things:

1)  Throughout the day, peak availability for discharge was well below aggregate Max Cap (at 8MW offered of 30MW registered that’s only 27% participation, which seems even more underwhelming).

2)  What’s more, there was a delay of a couple dispatch intervals before anything was offered:

3)  For unknown reasons from 14:45 the dispatch of the few negawatts even offered to the market dropped to 0MW, despite it being offered at ‘green’ prices well below the Market Price Cap.  It’s not immediately obvious why.

… all of this could be explored later, if we have time (Observation/Question #27) … but honestly it seems the least important thing on this list.

Wrapping up …

Having been through (above) this quick and top-down assessment of bidding through the day on Tuesday 13th February 2024 we’ll now be better placed to zoom into”

1)  Particular DUIDs of interest; and also

2)  Particular subsets of time through the day.

… but that’s another article!