AEMO warns of increased outages from ageing coal generators

Stormy weather and leaky boats:

“Should you find yourself in a chronically leaking boat, energy devoted to changing vessels is likely to be more productive than energy devoted to patching leaks.”

Warren Buffet

One of the great things about markets is that they impact all the participants. Price is going up, it’s basically good for all producers, in the short term at least, and bad for consumers and vice versa.

Doesn’t matter if your a coal generator, a hydro generator or own the lowest cost solar farm in Australia. If the electricity price goes up and you aren’t fully contracted it’s great. At the same time these guys aren’t your team mates, they are the enemy, always trying to get your lunch.

At the same time when price goes down the industry will tighten its belt. Capex will be reduced, opex will be worked on. A falling price is almost definitionally because there is competition for market share. Its in bad markets that you find out whether you are a stayer or just a fly by night. Optimistic assumptions get found out for what they really are in a down market.

So why the soapbox declamation?

Well, it’s no secret that the most enduring and pressing problem in Australia’s electricity system is replacing our increasingly clapped out fleet of coal stations in NSW and Victoria, and to a lesser extent, in QLD.

We think a combination of high coal prices in NSW, increasing competition from new renewables, increasing costs in Victoria and the pressure of having to operate hard in the past 18 months have increased both economic and reliability risks.

In reviewing reliability management it was interesting to read AEMO’s additional information provided to AEMC as part of its request for a RERT rule change.

AEMO provided some stats on the forced outage rate of coal stations over the past few years. Incredibly the forced outage rate for brown coal in FY18 was just about 8%.

This may just be a one year blip but the trend seems to be clearly up. And quite sharply. And it’s well-known that in Summer when demand is high reliability declines. Last year the most difficult time was on a weekend. There is an only 2/7 chance of that happening again on a weekend.

Our conclusion from looking at the above figure is that consumers should continue to hold their breath this Summer.

My own strong belief is that we get on with building new supply, mostly wind and solar, but some battery and some pumped hydro.

We also have the 830 MW of gas that reentered service last year and about 500 MW of AGL’s reciprocating gas under construction in South Australia and NSW.

Policy should be oriented to a steady supply of new wind and solar and some firming technology. Snowy Hydro 2 is not necessarily the best choice and certainly should have been required to go through a competitive tender but in the end will help with reliability when built.

Cash flows will be tighter

At the same time as reliability is declining pressure is coming on market share and price. If we look over the 12 months to June 2018 and ignore hydro then we can construct a “merit order” chart of average demand and supply as follows:

The chart shows the average quantity and an estimate of variable cost of the various fuel classes. Price will normally be higher than variable cost [SRMC] because the marginal generator, gas in this case, has to recover its fixed costs.

The merit order is no secret and neither is the broad impact of what happens when more variable renewable energy [VRE] with next to zero marginal cost enters service.

Basically, everything else is pushed to the right. If demand remains constant some generation will not operate because it can’t cover variable cost. Or it can’t cover variable cost by enough to recover all of its fixed costs.

For instance imagine that wind and solar expand from 2 GW on average to 8 GW on average and demand falls a bit.

The figure shows that the extra wind and solar bid in at close to zero displaces gas and also displaces, in this example much of the available NSW coal.

If the NSW coal is bid in at SRMC then the price falls from $65 MWh in Fig 2 to $28 MWh in Fig 3.

In theory NSW coal sets the price and should be prepared to set it just below the next unit of generation cost. In this example though there is not enough market share for all the NSW generators to have their desired market share and so they will compete on price.

A more finely grained chart would show that Bayswater will have lower costs and will end up setting the price with higher cost coastal units (Eraring and Vales Point B) losing share.

Overall consumers are big winners but even though high cost generators are the biggest losers the entire generation industry sees a decline in profit.

The above figures are entirely static. Of course the daily and seasonal numbers are much different.

When the sun is shining and the wind is strong the price is lower, at peak times in the evening much stronger. For instance, in another familiar figure, this is the average generation profile for NSW in FY18.

Based on ITK projections of committed but also as yet unannounced new wind and solar investment around the NEM, then one scenario for FY25 might be something like the below:

Imports from Vic/SA and QLD come into NSW because they can outcompete NSW coal and have first displaced much of the gas generation in both regions. We under project NSW rooftop in this model, but have held demand flat.

Price will be higher in the evening when in this model, hydro will be setting the price at times as will gas.

In this scenario there is no need for all the coal generators in NSW even after Liddell retirement.

We recently added in the numbers from Sunset Power, owner of Vales Point B, FY18 P&L filed with ASIC. They are the only NSW coal generation profits which we can actually get relatively clean numbers for.

Note that the sales price achieved was still well below the average pool price, although the gap closed in FY18 compared to FY17. Assuming raw materials the coal price was equal to $33 MWh despite Vales Point getting some of its coal from the insolvent Chain Valley coal mine. Chain Valley is insolvent partly because it sells coal too cheaply to Vales Point.

Vales Point profitability is highly sensitive to electricity price. Even assuming little further change in coal price there is a massive difference between a $50 MWh average price and a $70 MWh average price.

If the volume projections we show in our scenario above come to pass we expect price competition to push average coal generation prices towards the lower half of the range.

David Leitch is a senior utilities analyst and now runs his own energy consultancy ITK. He is a regular contributor to RenewEconomy.