AGL: Our baseload is 9GW overweight

Michael Fraser AGL Managing Director says baseload needs to go

As reported by Giles Parkinson at RenewEconomy and others recently, Michael Fraser made the comment that Australia needs to shed 9 GW of baseload generation, and more peaking generation needs to be brought in.

He says that slow demand growth and installation of rooftop Solar has put us in a position of having too much baseload which adversely impacts new market entrants.

The statements were based on research by his economist Paul Simshauser. Their solution to the problem is to introduce capacity payments that will incentivise new peaking entrants.

We’ve seen this work from Paul before

Paul Simhauser’s research isn’t new. In fact, 7 years ago in 2006 Paul submitted an almost identical piece of research to the AEMC. In that piece he was again arguing for capacity payments to be introduced, and reckoned that baseload generation was overweight. Though at that time it was 3.4 GW overweight. His research showed that in 1998 at market start we had a touch over 4GW too much baseload power.

Why are we short of peaking generation?

In Simshauser’s 2006 essay, he points at the volatile energy only market design as being to blame. For a generator to get built, it must be able to show the bank that it’s capacity has been sold for many years out. But with high price events occuring randomly it’s much more difficult to price a peak contract. Retailers are reluctant to hedge any significant amount of time in the future, because market conditions can change so quickly. This is why vertical integration has become the key to succeeding in the NEM today. However, despite the raging amounts of vertical integration, the level of peaking generation has still fallen further behind what Simshauser defines as optimal.

Why is there so much baseload generation?

We began the markets with a huge level of installed baseload generation, in the 15 years since the market start a further 6 GW of baseload generation has been installed. Which represents an increase of around 25%. That probably matches most industry expert’s predictions of load growth over the same period. Except the load didn’t grow, we’ve had very slow growth and an insanely high rate of rooftop solar installation both of which have exceeded expectations.

Paul says that we are 9 GW overweight. His “optimal” baseload target is the same level that was installed in 1998, that is around 20 GW. Meanwhile in 2006 he was predicting that we’d need around 30 GW of baseload power today (which is the level we actually have).

Perhaps all that is happened is the classic disruption model. The incumbent suppliers become extremely efficient at making and selling electrons from large centralised power generators. They didn’t see that rooftop solar would be a problem for them, until it was. Now the NEM is charging full steam ahead towards an uncertain future where our electrons are produced on our roof, instead of in the La Trobe valley.

How would a capacity market help?

Both in 2006 and now with two very different market conditions Paul has asked for a capacity market to be developed. How would it help?

Mentioned in this podcast

The top wind farms

Macarthur: $1M
Waubra: $560k
Lake Bonney II: $450k

Transcribed from the recording by machine, then edited for punctuation and paragraphing. Names, figures and misheard words have been corrected where the original show notes or the rest of the series settle them; anything they could not settle is left as spoken, rough patches and all.

Hi there, it’s Friday, 27th September, and you’re listening to WindFront, the weekly roundup of the wind industry from Australia and around the world. Today, I’ll round up the weekly top three. And talk about how AGL is claiming our base load is 9 gigawatts overweight.

Now it’s reported recently by Giles Parkinson at RenewEconomy, and others, Michael Fraser, the CEO of AGL, has made the comment that Australia needs to shed 9 gigawatts of baseload generation, and more peaking generation needs to be brought in. He says that the slow demand growth and installation of rooftop solars put us in a position of not of having too much base load, which adversely impacts new market entrants. The statements were based on research by his economist Paul Simshauser. Their solution to the problem is to introduce capacity payments that will incentivise new peaking entrants.

So I’m going to take a closer look at Paul Simshauser’s work. We’ve seen this work before. His research isn’t new. In fact, 7 years ago, in 2006, Paul submitted an almost identical piece of research to the AEMC. In that piece, he was again arguing for the capacity payments to be introduced. He reckoned that baseload generation was overweight, though at the time it was only 3.4 gigawatts overweight. His research showed that in 1998, at market start, we had a touch over 4 gigawatts, too much base load power.

So why are we short of peaking generation? In Simshauser’s 2006 essay, he points at the volatile energy only market design as being to blame. You see, for a generator to be built, it must be able to show the bank that its capacity has been sold for many years out, because banks dislike risk. But with high price events occurring randomly, which is when the peaking generator will run, it’s much more difficult to price the peaking contract, and retailers that are reluctant to hedge any significant amount of time into the future, because market conditions change, and large customers can switch to a new retailer pretty quickly. That’s why vertical integration, which is generation and electricity retail under the same company, has become key to succeeding in the NEM today. All the large players are doing it. However, despite the raging amounts of vertical integration, the level of peaking generation has still fallen further behind what Simshauser defines as optimal.

So why is there so much base load generation? We began the market start with a huge level of installed baseload generation. In the 15 years since, a further 6 gigawatts of base load has been installed, which represents an increase of around 25%. Now that probably matches most industry experts predictions of load growth over that same period, except a load didn’t grow, and we had very slow growth and insanely high rate of rooftop solar installation, both of which have exceeded wild expectations.

So Paul says that we are 9 gigawatt overweight. His optimal base load target now is the same that was installed in 1998. That is around 20 gigawatts. Meanwhile, in 2006, he was predicting that we’d need around 30 gigawatts of base load power today, which is actually the level we have. So we have a situation where back in 2006, he was predicting we’d need to grow to about 30 gigawatts of base load power, which we did. However, now that we’ve reached that level, his optimal base load target was way back in 1998.

Perhaps all that’s happened is the classic disruption model, and incumbent suppliers become extremely efficient at making and selling electrons from large centralised power generators, but they didn’t see rooftop solar would be a problem for them until it was. Now the NEM is charging full steam ahead towards an uncertain future for them, where our electrons are produced on our roof instead of in the La Trobe Valley.

Now, the big question: in both 2006 and now, with 2 very different market conditions, Paul has asked for a capacity market to be developed. How would it help?

Now on to the weekly top three, Macarthur, back over a $1 million for this week, Waubra and Lake Bonney II behind. So that’s it for me this week. And until I see you next time, keep buying those green electrons.