Utility solar may be low cost, but it is even lower value

 

Now that the transmission issue is “solved”, getting value back into solar is the No 1 policy priority. Actually, it’s way too early to say that transmission issues are indeed solved, but the path is clear. If the regulatory process (RIT-T) test doesn’t work, and it doesn’t, then State Governments, with the assistance of the Energy Security Board, will find a way around it.

Even the Federal Government has, on paper, offered to assist with underwriting of transmission in NSW.

So let’s move onto the next problem, which is the lack of value in solar. ITK’s attention was redrawn to this issue by Paul McArdle, CEO of Global Roam who constantly, and correctly, rabbits on about value not cost.

Then, in preparation for interviewing Matt Kean, NSW Energy and Environment Minister for the latest episode of the Energy Insiders podcast, and what a good interview it was, ITK reviewed our forecasts for what’s committed, or under construction but not operating right now.

A summary of those numbers is:

Despite no state support, NSW has about 40% of the new supply under construction and that new supply in NSW is equivalent to about 10% of NSW demand.

But that’s not all: On top of this, behind the meter solar will likely grow by about 20% over the next couple of years.

In NSW, ,though, there is a mismatch between the solar and wind development, and frankly solar is an easier political sell in NSW compared to wind.

But from the developers’ perspective, and from NSW consumer, big and small, the overweighting of solar should be a concern.

Bear in mind that before considering grid connection costs, which are a case by case analysis,

Solar has lower capacity factors than wind, let’s say 28-30% v 38%-40%;

Wind is highly variable but its output is reasonably evenly spread over the 24 hour daily cycle. By contrast solar operates through the middle of the day;

On an LCOE basis wind is basically still cheaper than solar. I get that there are good solar farms and bad wind farms, but right now wind still, in ITK’s perspective, has the advantage;

There are more economies of scale in solar than in wind. Wind sites tend to have an optimal number of turbines before they start interfering with each other. By contrast the bigger the solar site the more the construction teams can get their costs down and probably the lower the unit vegetation management costs.

Utility solar’s biggest problem is rooftop solar

Utility solar’s increasingly and overwhelmingly large problem is that it has to compete with rooftop solar.

Essentially there is already close to oversupply of total power at lunchtime and that oversupply is set to increase sharply.

That pushes middle of day prices down to below variable cost for utility solar and in Summers to come if nothing changes prices will be consistently close to zero during the day.

Rooftop doesn’t see that price signal directly so they won’t care. Eventually, though, feed in tariffs will reflect value to  system ($0) and this will incentivise household or community storage.

ITK is just finalizing itsJune Quarter  electricity price forecasts update but we can see what’s going to happen in NSW by looking at QLD where even in Winter the combined rooftop and utility solar share of production is 9%.

You can kind of see the issue here and noting that the price problem would be worse if QLD didn’t export surplus power to NSW particularly in the middle of the day.

However, we can bring the issue into more focus by looking at the coal, gas, utility solar and wind load weighted average price graphs.

To be clear, the load weighted price, takes the average generation of a fuel, say coal in a given half hour as a proportion of total coal generation and multiplies it by the price in that half hour and then sums that for all 48 half hours in the average day.

Figure 3, Source: NEM Review

Even in this half, even allowing for the fact that lunchtime surplus power is exported to NSW, even allowing for demand being relatively high in the half its still a fact that utility solar averaged $38/MWh.

Solar is cheap but the required price is still more than $38/MWh even if your MLF was 1, and it isn’t.

Even in Queensland the problem is going to get heaps bigger, not only because more solar is coming (see Fig 1 and add in some rooftop) but also because the increase in wind across the entire NEM will lead to more wind production when solar is operating as well as at other times.

Everyone else will be “giving it” to NSW

NSW, as we have said on many occasions, is the first home of every other State’s surplus.

Provided there is enough transmission when the big increase in Victorian wind comes through from Stockyard Hill and Moorabol the first instinct of the brown coal generators will be to shove the excess power up to NSW where brown coal’s variable costs are lower than those of NSW black coal generators.

Admittedly some could go to Tasmania, but only if Hydro Tasmania wants that to happen.

QLD’s increase in generation has only one place it can go, and that’s NSW.

When Project EnergyConnect – the proposed link between South Australia and NSW – is eventually built, excess solar and wind in South Australia will flow to NSW as well as the existing flow to Victoria.

Another possibility is that more generation closes.

Why would a corporate sign a PPA with a solar developer?

There may still be reasons for storage owners to sign PPAs with solar developers, or to develop their own solar farms, but any corporate with half a brain is going to buy out of the spot market, since the spot price will be lower than the LCOE of a new solar farm.

Corporates really only want to contract three years forward anyway.

Of course solar farm developers may give away PPAs as loss leaders to get their project up, hoping to make it back in the spot market with the “equity” portion, but that is likely to end in tears.

It’s not much fun for existing coal either

NSW coal generators will lose volume in the middle of the day, but they also lose out in the midnight to dawn market when Victorian brown coal and Qld black coal and wind all undercut the higher NSW variable cost.

We can already see this quite strongly with NSW coal on average having to ramp up 2000 MW every day from trough to peak, and in fact they already have to ramp down temporarily in the middle of the day before ramping up again for a very short evening peak.

NSW coal generators have great ramp characteristics so they can cope with this, but its going to get more extreme before Liddell closes.

Not only are there the technical issues of ramping but as stated the ramp downs result in lost volume and revenue.

Also of interest is how narrow the evening price peak is already.

It may, or may not be economic for a battery to arbitrage out that peak, but if the battery already existed for frequency control it could potentially do quite a lot of damage to coal revenue by eating into that short peak price.

In short, I think the vulnerability of the NSW coal generators is already evident and we have only just started the 2020s. I left January out to avoid bushfire distortions.

Figure 4, Source: NEM Review

What’s the role of policy here?

So far in NSW the emerging policy is to push transmission along, put in a reliability standard and talk about the low cost and environmental benefits of wind and solar.

Given the level of interest in the NSW REZ (renewable energy zone) , as well as the new solar under construction one could be forgiven for thinking the policy is working.

But cost is not the same as value and policy needsd to bear that in mind. There are some significant risks;

The most obvious is a repeat of Hazelwood. This is where everyone can be grateful to AGL for standing up to bullying by the Federal Government on Liddell.

By giving very long term notice of Liddell closure the market has plenty of time to build new supply in advance. But things move quickly in electricity markets and Vales Point could easily find itself under pressure sooner than it thinks.

The Vales Point structure doesn’t have any debt but neither does it have any assets other than the power station. In any event it’s quite eviden that Vales Point needs to be replaced starting fairly soon.

It took years for enough new supply to be built to replace Hazelwood and now Liddell and at least there was the RET target to incentivize that production.

Now with only “Technology not taxes” and Angus “the day trader “ Taylor busy punting taxpayer funds on the oil price the clear risk is that new supply pipeline drys up. You can’t replace Vales Point in three years.

The main reason is that you spend the first two years arguing about how to do it. Because your starting point is zero policy incentive.

You may have to spend 5 more years building transmission. The process of replacing Vales Point needs to start now. Same frankly for Gladstone and Yallourn, but that’s for another day.

Secondly, excess solar is probably not the system wide lowest cost way of doing things. Nearly every study of the NEM has a bigger role for wind than solar.

It’s not only that wind is lower cost, it’s not only that NEM wide correlations of wind are lower than those for solar, it’s not only that wind operates over the whole day, but basically the higher capacity factor of wind means that a wind biased system has a lower firming cost than a solar biased system, probably about 30% lower (40% capacity factor compared to 28% capacity factor).

So you need a policy incentive that favours wind. And politically that’s tough in NSW.

A plan for NSW would be a good start

If I was Matt Kean, and seeing that renewable energy and the environment are election winners and politically popular and that basically NSW voters by and large like hearing the coalition talking about decarbonizing NSW I would:

Develop the equivalent of the ISP for NSW, but with a target portfolio subject to a carbon budget. I’d model my plan on that of California. I might even invest in a second copy of Plexos.

Publicise the plan so that developers and large consumers can see what’s coming and point out what’s wrong. Based on a two year cycle version 1 of the plan is likely to be pretty rubbery but by version 3 every one should be in the groove particularly as regards what input costs actually are;

Decide on the best way to implement the plan. In my view this could be done with some reverse auctions, but maybe there is a better way. Reverse auctions give very clear signals, allow for targeting by fuel and also produce strong competition. And they are fair with objective criteria that can survive probity testing. I think there is little doubt about that. As things stand reverse auctions could favour wind over solar. Equally if there was a need for firmingcapacity, an auction lets all the competing technologies compete, including those from out of State (eg Tasmania) If entrepreneurs are given a clear carbon budget then that can be allowed for in new investment. Investors looking to gas can get their costs far more accurate if carbon risk can be measured;

Spend some of my R&D dollars in the software and inverter development space. Most of the exciting technology is around control of the system. There are already a number of small firms like “SwitchDin” and “Planet Ark” often in regional centres like Newcastle that have a lot to offer. There is going to be a big market for grid forming inverters and grid controlling software. It’s the sort of thing Australian electronics engineers and our excellent computer science teams would love to have a go at. One way of doing it might be to emulate in the remote areas of the State some of the distributed microgrid approaches being adopted in West Australia and Queensland. But this is probably too specific. Incentivise community batteries in new apartment blocks. Imagine if each ofthose inner city towers had say 2 MWh of centralised storage in them. These moves would be very popular politically in the City. NSW doesn’t have as many committed anti renewables groups as say QLD, but more to the point there is a large population of middle class Liberal voters that would love to see a combination of technology and economics being brought to bear by a progressive Liberal Govt. It is technology and economics that do it, but that is not the role of policy.

The role of policy is to set the target and provide the mechanisms. Will it happen?

Probably, not but it’s worth a dream or two.