Since 23^rd May we’ve seen big price spikes, to over $1,000/MWh at Otahuhu for example, consistently during the peak demand at the end of the day.  In this post we take a brief look at what’s causing these spikes.

By Greg Sise, 31 May 2018

Before I go any further, I have to state that I am not attaching any blame to any party.  Dramatic events are more often than not due to a random combination of factors, often beyond the control of the parties concerned.

The highest spike so far occurred last Monday, 28^th May, as shown below, in trading period 37, and whatever caused it impacted both Otahuhu in the north and Benmore in the south, with the price difference between the two a function of losses during northward transfer on the HVDC link.

If we now include reserve prices on the chart below, we immediately see an obvious correlation:  spot energy and reserve prices spiked up together.  Was this a coincidence?

Well in fact it was no coincidence.  It appears there was a general shortage of reserves, especially sustained reserves (SIR on the chart) which led to the dispatch of reserves at a high price. 

I haven’t looked at the data in detail, but often when there is a shortage of reserves the dispatch model has to turn generation down so that it can provide what is known as “partly loaded spinning reserve” (PLSR) which simply means that a generator that is partly loaded can quickly increase output (“ramp up”) if there is a sudden and unforeseen event which causes a large generator to disconnect from the grid.  In these cases, more expensive generation has to be dispatched to replace the output reduction at the generator that is turned down to provide PLSR.

But this scenario suggests a deeper shortage of cheaper generation, which brings us to those coincidences I mentioned earlier.

After a period of unsettled weather, it’s now settled down, causing a drop-off in the output of wind farms.  On the 23^rd the wind farms got to over 300 MW overnight but by trading period 37 they were down to 200 MW.  Then yesterday (30^th) they actually got down to zero output – now that is unusual in these windy isles in which we live!  (Note that we don’t have access to data for all wind farms, only those that are grid-connected).

On top of that, the 200 MW Stratford Peaking station in Taranaki was notified for an outage of one unit (100 MW) from 2^nd to 21^st May, but then on the 18^th this outage was extended to 28^th September.  In addition, on the 23^rd the second unit at Stratford was put on outage through to the end of May, but then on the 25^th this was extended to 30^th June.  So we now unexpectedly have 200 MW at Stratford out for an extended period:  we’re not sure but there must be a significant and unexpected problem discovered at the station.

Looking at the data for yesterday, the cost of reserves appeared to get so high that the dispatch process turned the output of the e3p and TCC combined cycle stations down to reduce the reserve risk.

Now although these spikes are occurring due to a combination of unexpected circumstances, I have to reiterate an important point:  back in 2015 two large gas-fired generators in Auckland were shut down and we lost over 500 MW in capacity when that happened.  That means that the margins are slimmer and that these chance events are more likely to occur unless we experience a reduction in demand for electricity.  Is that likely?  No, very unlikely.  Another 50 MW of constant demand will be added later this year when the Tiwai smelter restarts its fourth pot-line (that’s 1% of total demand on average) and our population and economy are growing rapidly.

To learn more about how dispatch and spot pricing for energy and reserves work in the New Zealand electricity market, check out our Nodal Pricing course.