March 1999

As highlighted in the diagram above, March 1999 saw prices in QLD higher than in the other regions of the NEM.

The chart included above (from NEM-Review) illustrates that:

• There were many separate instances of price volatility over the month (this is understandable, as the QLD region was not connected to NSW at this time); and
• These price spikes would have had the effect of growing QLD’s Cumulative Price to close to what we believe the Threshold would have been in the early years.

March 2000

The main chart (above) illustrates that prices in South Australia had climbed higher in March 2000, whilst prices in the QLD region had fallen (coincidentally to around the same level).

The chart included above (from NEM-Review) illustrates that two price spikes played a major role in the higher average price.

In particular, the price spike on Thursday 2nd March saw prices greater than $1000/MWh for a total of 9 x 1/2-hour trading periods.

Once again, in this instance we can see (in the inset on the chart) an indication of demand-side response (or embedded generation) active in the market, with a response of about 60MW shown.

March 2001

March 2001 saw:

• The average prices in SA and QLD decrease from their levels in the previous year; and
• Prices in NSW, SNO and VIC climbed, but only slightly.

March 2002

As was also the case in January 2002 and February 2002, prices dropped considerably in March 2002 (compared to the previous year).

These subdued prices were due to a number of factors, including:

• the market being relatively over-supplied with capacity (due to the recent commissioning of Callide C & Millmerran in QLD and Pelican Point in SA),
• Added to this, the greater interconnection capacity (provided by QNI) helped to reduce volatility.

March 2003

As noted in the chart above, prices were astonishingly low across all regions in March 2003.

March 2004

As shown in the main chart (above) prices in March 2004 were significantly higher than in the previous 2 years – in all regions but Victoria.

The chart included above (from NEM-Review) illustrates that two price spikes played a major role in this situation for March 2004.

1) The case of 8 March 2004
On 8th March 2004, a bushfire in the vicinity of ElectraNet’s transmission network was a precursor to a cascade of automated protection responses, leading to:

• A rapid output reduction at the Northern Power Station;
• The islanding of the South Australian region; and
• The loss of approximately 650MW of demand in SA through under-frequency load-shedding.

The following chart from NEM-Watch version 5 highlights what was happening in the market immediately after the event.

As is seen in this illustration, the price has hit VOLL in South Australia as a result of the islanding of the region, and the loss of load.

Note in this image that we had also prepared for Tasmania and Basslink, and also for the introduction of SNI (which did not eventuate).

This event has been further reported, and discussed, in the following documents (amongst others):

• Note that NEMMCO’s report into the incident (released on 5th July 2004) and its updated report (released July 2005) cannot be found online, following the restructuring of NEMMCO’s website;
• A report by SKM “Audit of ElectraNet SA Service Standards Performance Reporting” to the AER for the year 2004 (p15‑16);
• A letter by NRG Flinders of 11 January 2006 to the AEMC;
• A letter by ElectraNet of 27 March 2006 to the AEMC.

2) The case of 9th March 2004
The second snapshot (below) illustrates a time when prices were high in QLD, NSW and SNO for close to 6 hours over the day.

As noted on the image, this instance illustrates why we have added (within NEM-Watch version 8) the real-time display of the IRPM of the “Economic Islands” dynamically created by constrained interconnectors.

In this case, we see the NEM-Wide IRPM (pre-Tasmania) was a very reasonable 20% at the time.

However, the surplus generation capacity shown to be present in VIC and SA could have no impact on moderating prices in NSW and QLD because of transmission constraints. With version 8, we calculate an IRPM for the NSW+QLD Economic Island (active for the dispatch interval above) as follows:

“Local” Demand = 7393 + 11997 – 3084 = 16,306MW
this is the “Local” demand supplied by generators in NSW and QLD
Available Generation in NSW + QLD = approximately 18,800
(estimating off the graph)
Hence, IRPM for the NSW+QLD Economic Island
= (18,800 – 16,306) / 16,306
= 15%

This number is lower than that for the whole of the NEM (understandably) but is still reasonably high – hence there must be other reasons (in addition to simply supply/demand balance) why the price was sustained at a high level for so long.

March 2005

As noted in the chart above, prices were relatively subdued across all regions in March 2005. This was in spite of the fact that the NEM-wide demand (peak, especially) was considerably higher in March 2005 than in March 2004.

The above chart (from NEM-Review version 5.3) clearly illustrates why South Australian prices were higher, on average, than those in the other regions for the month.

The two price spikes were as follows:

1) The price spike of Thursday 3rd March 2005
At 08:30 on the 3 March 2005, the South Australian price jumped to $1730/MWh. We have not investigated this incident.

2) The price spike of Monday 14th March 2005
On 14th March 2005, an event (similar to what happened on 8th March 2004, as noted above) occurred in South Australia, resulting in approximately 700MW of load shedding. The price rose to $6500/MWh for one half-hour (07:30) trading interval.

From what we understand, the incident occurred when output from the Northern Power Station was automatically (and rapidly) wound back following a nearby transmission fault.

Following from this incident, NEMMCO constrained down the flow allowed over the Heywood interconnector (imports to South Australia). As a result of this, prices jumped in South Australia.

This situation can be seen through sequential snapshots, taken with NEM-Watch version

At 06:45, we see the market in its last “normal” state before the loss of load, the interconnector limit reduction, and the price spike.

Five minutes later (06:50) we see a number of significant events occur, as follows:

• We see the Heywood link reduced to zero flow (and zero transfer capability) temporarily; and
• We see demand in SA reduced from 1680MW (at 06:45) to 880MW (at 06:50);
• Finally, we see the price drop in SA, because of the loss of load.

Twenty minutes later (at 07:10) we see the price at VOLL.

Note in this screenshot, we see Heywood back to exporting at 11MW, whereas the flow limit is shown (in NEM-Watch) to be 460MW.

• Under normal circumstances, we would not expect to see price separation occur (like this) when interconnector flow appears not to be constrained;
• We understand that, in this case, NEMMCO defined the largest single contingency to be a loss of two units at Northern and, as such, reduced the capacity on the interconnector to prepare for this – meaning an actual flow capacity of very little (hence the VOLL pricing).

March 2006

As the main chart above shows, March 2006 showed very modest prices in all regions.

March 2007

As shown in the main chart (above), prices in March were higher than in the 2 previous years, for reasons including the impact of the drought.

This issue was discussed, in more detail, in the review for the month of April.

March 2008

South Australian’s won’t quickly forget the 15 straight days of extreme temperatures that they had to endure in March 2008.

This caused significant issues in the NEM, including the Cumulative Price reaching the Cumulative Price Threshold for the first time in our memory, as shown here:

Shortly after this time, we prepared this analysis of what was happening in the SA region over this period.

As has been helpfully pointed out by “Apol”, a contributor on WattClarity®, this analysis does simplify some of what happened over the period (in particular with respect to constraints in and around the Snowy Region).