Since you’ve been gone

Wednesday marks 8 weeks since the closure of Hazelwood power station. Following the announcement of its closure in November last year, there was much speculation about what the consequences might be: for prices, for reliability, and for market outcomes. With almost two months having passed since Hazelwood breathed its last, this week we examine what has happened since it’s been gone.

Disclaimer: At the outset, we emphasise that these outcomes will neither necessarily be enduring, nor should they be attributed solely to the closure of Hazelwood. It will doubtless take some time before the full effect of such a massive change in the power system becomes clear.

Chart 1: Price versus demand before and after the closure

Chart 1 shows dispatch prices versus demand in Victoria for the period from 1 January 2015 to 20 May 2017. By way of explanation, each point represents the price and demand in Victoria for a single dispatch interval. We have coloured the points:

  • prior to the closure on 28 March 2017 in blue; and
  • following the closure in red.

To aid the visualisation, we have restricted the vertical axis to $0 to 400 per MWh – we note that there have been no high price events in Victoria since Hazelwood’s closure.

There is a clear change in prices following the closure of Hazelwood – the cloud of red points sits markedly higher than the blue cloud. Even at very low levels of demand, we have regularly seen prices of $100 per MWh.

Chart 1 – There is a clear change in prices in Victoria following the closure of Hazelwood

Victoria dispatch prices versus total demand – 1 January 2015 to 20 May 2017

Chart 2: Average price by level of demand before and after the closure

So how much have prices increased on average? Chart 2 shows the change in average price by level of demand. By way of explanation:

  • the top panel is simply chart 1, which we have provided to aid the visualisation; and
  • the bottom top panel shows the average Victoria price before and after the closure for each level of demand (divided into bins of 500 MW).

For example, we have added labels for the average price for demand from 6000 to 6500 MW. You can see that the average price prior to the closure was $64 per MWh, and following the closure it has been $171 per MWh. There is a clear increase in average prices at all levels of demand observed since the closure.

Chart 2 – The supply curve has shifted markedly since the closure of Hazelwood

Top Panel: Victoria dispatch prices versus total demand – 1 January 2015 to 20 May 2017.

Bottom panel: Victoria average dispatch prices by level of demand.

Chart 3: Histogram of flows on the Vic-NSW interconnector

In addition to prices, we have also examined the change in flows on interconnectors. Chart 2 shows histograms of the flow on Vic-NSW. By way of explanation:

  • The vertical panels show different years: 2015, 2016, and 2017.
  • The horizontal panels show different months, ie, April or May. We have restricted our analysis to these two months, so that we can compare outcomes since the closure with similar conditions in previous years.
  • Within each panel, the horizontal axis shows the flow on the interconnector from Victoria to NSW, and the vertical axis shows the number of dispatch intervals for which that flow occurred.
  • We have coloured the chart according to whether the interconnector was constrained at the time: red for binding, and blue for unconstrained.

The change in outcomes is stark. In 2015 and 2016, Victoria was consistently exporting to New South Wales, with the interconnector regularly operating at its dynamic limit. But since Hazelwood has closed:

  • when the interconnector is flowing from Victoria to NSW, it seldom reaches the dynamic limit; and
  • the interconnector is now often flowing from NSW to Victoria, and is regularly reaching its dynamic limit.

Chart 3 – The operation of Vic-NSW has altered dramatically since the closure of Hazelwood

Histogram of flows on Vic-NSW for April and May – 2015, 2016, 2017; colour scale shows whether the interconnector was constrained.

My two cents

  • Hang caution and our disclaimer – the changes are the result of the Hazelwood closure. That we are already seeing such enormous changes in the power system begs the question: what’s in store for winter, and next summer?
  • The river of cheap energy that New South Wales has come to rely upon has stopped flowing. Indeed, Victoria is now asking for water to be sent ‘back upstream’. But does NSW really have the energy to spare? And even if they do, how much of it can they push across the interconnector, which is often binding at very low limits (ie, -200 to -400 MW).

As always, all comments are most welcome.