Part 2 in a review of bidding, on Monday 16th December 2024 (a focus on VIC Wind Farms)

We’ve already posted ‘A quick initial review of bids, for Monday 16th December 2024’ early in the morning on Tuesday 17th December 2024.

In that article, flagged that amongst other aspects that we’d like to drill further as time permits, were questions raised with respect to Wind Farms in Victoria – not least because of two interest phenomena that (at first glance) appear to have been occurring through the day:

1)  It seems highly likely that there was considerable network congestion affecting wind farms in western Victoria

(a)  With a big clue being the large volume of additional energy bid down at –$1,000/MWh (i.e. light blue, above) that did not receive a Target

(b)  Which is not a surprise, given what we saw and noted yesterday.

2)  Additionally, it appears that (because of high temperature effects, generally poor UIGF or other reasons) there were one or more Wind Farms that had considerable difficulty following their Target.

In this article we’ll peel off another layer or two of the onion, looking specifically at Wind Farms (and only Wind Farms in Victoria – though readers should note that there appear also interesting questions about Wind Farms elsewhere as well).

Full Disclosure

Let’s start by noting that in early 2024 we (at Global-Roam Pty Ltd*) contributed a seed investment in another early-stage start-up focused on providing self-forecasts to a growing number of Wind Farms (including some listed below).  This investment has followed from our growing interest in helping to make the energy transition actually work, by providing assistance in various ways, as we can.

*  Global-Roam Pty Ltd is both:
(a)  the providers of this WattClarity service
(b)  and developers of the ez2view software (amongst other software products)
(c)  amongst other initiatives.

The fact that we have now made this investment gives us one more reason to take a keen interest in the operations of wind farms.  Hence we share some insights with WattClarity readers here today.

Walking through all 24 Victorian Wind Farm stations*

Continuing to utilise the ‘Bids & Offers’ widget in ez2view, which we’ve time travelled back to 00:00 on Tuesday morning 17th December 2024 and is looking back 24 hours, we’ll quickly walk through all 24 Wind Farm stations* to quickly ascertain which of these are showing signs of either of the phenomena as above.

* noting that there are a few more DUIDs than stations due to stations like Dundonnell and Golden Plains, which have multiple DUIDs in a station.

Note that we don’t have time to fully answer (at least at this point) any of the questions we raise in what follows … just aiming to reveal more detail with one more layer of the onion peeled.

Stations are in alphabetical order of STATIONID.

VIC WF station #1 … Ararat Wind Farm

Starting at the top of the alphabet, we have the Ararat Wind Farm:

Now, longer term readers here on WattClarity® might remember that Ararat Wind Farm was the focus of the quiz/article ‘Which was that ‘mystery DUID’ that apparently ‘can’t stand the heat’*?’ from June 2019 …

1)  so perhaps the high temperature limitations are one factor contributing to the pattern of output through the day:

(a)  Maxing out at 208MW (at 10:15 NEM time), below MaxCap; and

(b)  declining therafter;

2)  but there may well be a more common explanation (that being just a general limitation on the wind resource).

Also of interest are:

1)  the drop in output between 15:00 and 16:00 (NEM time) … which (my first guess) may well be due to network congestion.

2)  the change in bid bands several times during the day … where again my first guess is due to the operation of constraints.

VIC WF station #2 … Bald Hills Wind Farm

Next on the list we have the Bald Hills Wind Farm:

In this case we see the unit’s Availability up-and-down in a way that seems ‘normal’ for an underlying variable wind resource and (with volume bid at –$110.24 at the RRN):

1)  most of this volume consistently dispatched (i.e. implying no congestion effects on this unit);

2)  except for two dispatch intervals (13:40 at –$183.01/MWh and 17:10 at –$219.70/MWh) where it appears the bid was ‘out of the money’ and so the unit lost volume for economic reasons.

VIC WF station #3 … Berrybank 1 Wind Farm

Next on the list we have the Berrybank 1 Wind Farm:

With this unit, there are three interesting things we can highlight here:

1)   there’s a general pattern of decline in output through the day (which is correlated with the increase in ambient temperature, but that may be independent or interdependent);

2)  There’s also a span of time (from morning to until early afternoon) where the unit fairly consistently did not follow its Target.

3)  Also we see a period of time where the unit (whilst Available and bid down to -$1,000/MWh at the RRN) appears fully dispatched down to 0MW … probably due to the action of constraints.

Time permitting I’d be interested to explore this further (especially with some commonality across the portfolio).

VIC WF station #4 … Berrybank 2 Wind Farm

Next on the list we have the sibling station, Berrybank 2 Wind Farm:

There’s definitely some similarities here between Berrybank 1 and Berrybank 2.

VIC WF station #5 … Bulgana Wind Farm

Next on the list we have the Bulgana Wind Farm:

This unit also displays a sustained period (for a longer period of time, but not as large a relative ‘Dispatch Error’) where its Actual Output is consistently under its Target.

VIC WF station #6 … Cherry Tree Hills Wind Farm

Next on the list we have the Cherry Tree Wind Farm:

In this case, the unit ramps up its output until early afternoon and then holds steady (though not at full capacity) before dropping quite low (including to 0MW) between 21:00 and 21:30 (NEM time).

VIC WF station #7 … Crowlands Wind Farm

Next on the list we have the Crowlands Wind Farm:

This unit sees a number of instances of sharp reductions in output, and then reversals which (in most cases) appear to follow Targets fairly closely.  Because the volume is bid down to -$1,000/MWh, one suspects this yo-yo behaviour is the result of constraint action.

VIC WF station #8 … Dundonnell Wind Farm

Next on the list we have the Dundonnell Wind Farm:

Remember that this station contains three separate units … but I don’t see why that should make any difference to the interpretation of this station-level bids chart.

In the case of this station we see a few brief periods of 4~6 dispatch intervals where it appeared there was some difficulty in meeting the collective Target across the 3 units.

VIC WF station #9 … Elaine Wind Farm

Next on the list we have the Elaine Wind Farm:

For reasons that have not been explored at this point, this station was not operational throughout the day on Monday 16th December 2024.

VIC WF station #10 … Golden Plains East Wind Farm

Next on the list we have the Golden Plains East Wind Farm:

Remember this station contains multiple units … and also that it is very much in commissioning. As we can see above, it was unavailable the whole day.

VIC WF station #11 … Hawkesdale Wind Farm

Next on the list we have the Hawkesdale Wind Farm:

This station displayed some commonalities with what happened at Berrybank 1 and Berrybank 2 … which might perhaps be due to common portfolio?

VIC WF station #12 … Kiata Wind Farm

Next on the list we have the Kiata Wind Farm:

We can see that there’s some commonality here with Hawkesdale (though note that Kiata WF is in a different portfolio), specifically with respect to:

1)   A contiguous run of dispatch intervals with some difficulty in meeting Target; but also

2)  A downward trend of output through the main daylight periods.

VIC WF station #13 … Macarthur Wind Farm

Next on the list we have the Macarthur Wind Farm:

This one is a different case that’s interesting for different reasons … for reasons not explored (though the rebid suggests some physical reason) the unit/station was fully unavailable from around midday.

This would be one that I’d be curious to delve into in a subsequent article … time permitting.

VIC WF station #14 … Moorabool Wind Farm

Next on the list we have the Moorabool Wind Farm:

There’s a relatively brief period in the morning where there are sequential Dispatch Error (output under target), but more noticeable is a long period where the unit has bid down at -$1,000/MWh but is still dispatched down considerably (presumably due to congestion, but we’ve not looked further to confirm at this point).

VIC WF station #15 … Mortlake South Wind Farm

Next on the list we have the Mortlake South Wind Farm:

In this view what jumps out to be first is a period of almost 4 hours to 16:00 when the unit is completely unavailable … for reasons we’d like to explore later.

VIC WF station #16 … Mt Gellibrand Wind Farm

Next on the list we have the Mt Gellibrand Wind Farm:

This unit appears to have some limitation on output at ~100MW compared with a MaxCap of 132MW … for reasons we’d like to explore later.

VIC WF station #17 … Mt Mercer Wind Farm

Next on the list we have the Mt Mercer Wind Farm:

This unit sees several periods where it appears to be fully ‘constrained down’ … despite (in the later periods) bidding down to -$1,000/MWh.

VIC WF station #18 … Murra Warra 1 Wind Farm

Next on the list we have the Murra Warra 1 Wind Farm:

The unit sees a long period through most of the afternoon/evening where the output was limited to not much at all.  But even in the other periods, the output seems to be limited to a maximum of ~145MW compared to a MaxCap of 225MW.  Reasons for both have not been explored at this point.

VIC WF station #19 … Murra Warra 2 Wind Farm

Next on the list we have the Murra Warra 2 Wind Farm:

We can see that this sibling unit operated differently from Murra Warra 1 … perhaps because of the different stakeholders involved.  This is one of the wind farms seeing a steady decline in output through the morning and into the afternoon, for reasons not yet explored.

VIC WF station #20 … Oaklands Wind Farm

Next on the list we have the Oaklands Wind Farm:

This is a curious case where the output was limited from ~07:10 … for reasons not explored at this point (but potentially due to constraints).

VIC WF station #21 … Ryan Corner Wind Farm

Next on the list we have the Ryan Corner Wind Farm:

This is another wind farm that shows some similarity to others in the portfolio (Berrybank and Hawkesdale) in that there is:

1)  A steady decline in output for a period of time in the morning

2)  Plus many instances where the unit is substantially unable to meet its target (with the largest Dispatch Error seeming to have occurred between 13:00 and 14:00).

VIC WF station #22 … Salt Creek Wind Farm

Next on the list we have the Salt Creek Wind Farm:

For this Wind Farm we see:

1)  a gradual/steady decline in output through morning and into the early afternoon

2)  plus instances of (relatively small) Dispatch Error

3)  then more perturbations on output later … perhaps due to the underlying wind resource?

VIC WF station #23 … Stockyard Hill Wind Farm

Next on the list we have the Stockyard Hill Wind Farm:

For reasons not explored at this point, this station seems to be the one with the most consistently full output profile … at least until after 16:00.

VIC WF station #24 … Yendon Wind Farm

Next on the list we have the Yendon Wind Farm:

Unlike the Elaine Wind Farm (above, which was offline for the day), this unit dispatched at close to full output until around 17:00.  There are a number of periods highlighted where the unit Actual Output was consistently under its Target.

Summing up…

Summing up, there appears to be a few different phenomenon represented across the 24 different cases above.

Phenomenon #1 … variability of resource … but also some commonality of patterns

All of our readers will understand that there’s variability involved in the underlying wind resource, but readers will see from the 24 images above that there are a number of wind farms that exhibited common patterns.  One pattern common to a number of the 24 Wind Farm stations was a trend down from morning through until afternoon:

1)  which might be just because of a decline in the underlying wind resource,

2)  or might have been for other factors (at least in part).

Phenomenon #2 … difficulty following Targets (for some Wind Farms)

It appears that there were some Wind Farms that had considerable difficulty following their Target.

1)  This would have calculated out mathematically as a Dispatch Error

2)  But note that the the ‘Bids & Offers’ widget in ez2view, does not show whether the Semi-Dispatch Cap flag was set during …

(a)  so we don’t know whether these units would have been classified by AEMO under their Conformance Status process

i.  we’d need to utilise other widgets (like the ‘Unit Dashboard’ widget in ez2view) to see this

ii.  which we might do later in some cases, as a follow-on article (time permitting)

iii.  albeit that a Semi-Scheduled unit might be not ‘Normal’ only when the SDC Cap is TRUE and the Output is above the Target … whereas in most cases we see the reverse is true.

(b)  but it might (depending on what the Frequency Indicator was doing during those periods) have contributed to a disadvantageous Causer Pays Factor … which is used, after a lag, in allocating Regulation FCAS costs (at least until FPP begins in mid 2025).

(c) and perhaps, in some cases, the AER might take an interest (given their clear caution given at the AEMO Intermittent Generation forum at the start of December).

Of the 24 stations above it looks like these stations, in particular, fall into this category:

1)  Most of those identified above showed contiguous periods of many dispatch intervals where their output was below Target:

(a)  Berrybank 1 and Berrybank 2 and Ryan Corner and Hawkesdale all share a common portfolio (so perhaps this is a factor) and all share some similarity of pattern

(b)  Plus others as well

2)  But there was one stations that showed the reverse (i.e. output above Target).

There are several reasons why this might have been the case:

Possible Factor 1)  It might have been because of high temperature effects …

Possible Factor 2)  It might have been because of a generally imprecise UIGF (Unconstrained Intermittent Generation Forecast)

… whether from AWEFS or an individual Self-Forecast (the specific source is not explored in this article).

Possible Factor 3) possibly the Availability estimates, and hence Targets were reasonable … but the unit Dispatch Systems had some issues in achieving what should have been possible?

Possible Factor 4) and, of course, there may well be other factors as well.

It’s likely that (across the Wind Farm fleet noted here) above there are a range of factors at work.   We’ll need to follow up with a subsequent article (part 3?) in order to use the ‘Unit Dashboard’ widget (or others) in order to see more for this subset of units.

Phenomenon #3 … strong Network Congestion (for some Wind Farms)

It seems highly likely that there was considerable network congestion affecting wind farms in western Victoria

1)  With a big clue being the large volume of additional energy bid down at –$1,000/MWh (i.e. light blue, above) that did not receive a Target

2)  Which is not a surprise, given what we saw and noted yesterday.

Of the 24 x stations shown above, the following show signs that they were affected by constraints (noting that the ‘Bids & Offers’ widget in ez2view focuses on the bids that NEMDE considered for dispatch, but not other factors, such as specific constraints).  We’ll need to follow up with a subsequent article (part 3?) in order to use the ‘Unit Dashboard’ widget (or others) in order to see more for this subset of units.

Phenomenon #4 … some Wind Farms off entirely

Two Wind Farm stations were off for the entire day (ELAINE and GPWFEAST) whilst other Wind Farms (INSERT HERE) experienced time periods through the afternoon where their Availability (and hence Target) was set to 0MW.

We’ll need to follow up with a subsequent article (part 3?) in order to use the ‘Unit Dashboard’ widget (or others) in order to see more for this subset of units.

Time permitting we might explore more later … but that’s all we have time for in this thread today.