Queensland: A little sunlight is needed on renewables policy

“The sun was shining almost directly overhead onto the sand, and the glare on the water was unbearable.”

Albert Camus L’Etranger, 1942

50% renewable target: Great slogan, but arguably as compromised as Brexit

You can still make an argument that Queensland is on track to achieve 50% renewables by 2030. The renewable share of generation in the March quarter at 11 per cent is double that of a year ago and there’s about 750MW of renewables still to come on stream.

The problem is about 700MW is needed every single year to 2030 and there is next to no QLD policy to get there. The newly formed CleanCo will provide some help but basically Federal Government policy is more likely to drive development in QLD.

The numbers show that QLD  Govt owned coal  generators are in good shape. Dividends are $620 million, almost double the level of two years ago. You can give a tick to the  Cleanco process although “clean” is a relative word as it also includes an ageing gas generator.

But….

The Powering North Queensland plan has been hidden in an unreleased Powerlink feasibility study. Hiding the results of a study into an election commitment is poor policy.

Stanwell and CS Energy’s profits will be destroyed if the 50% renewable target is achieved and these assets could have been sold from  the public sector years ago.

Let’s be honest, you don’t improve a budget by hanging onto dying assets and then giving their near-term cash flows away in the form of a one time $100 per household  bonus.

That is populist politics and not policy. Capex is less than depreciation at the  generators  and we don’t see a viable plan for Stanwell or CS in the medium term. Gladstone and Tarong power station will have ramping  issues by 2025. At Stanwell coal prices will likely jump in 2027.

Pool prices in QLD, still lower than the rest of the NEM, have jumped this year despite the increase in solar and wind generation, some of which is effectively being exported to NSW. Prices have jumped for every half hour in the March quarter average by time of day curve.(See Fig 8).

Queensland generation and demand- the data

A brief summary of  recent electricity generation in Queensland is shown in the following table:

Figure 1: Queensland electricity generation statistics. Source: NEM Review

The table reveals a sharp price jump during the quarter  compared to last year of about $27/MWh.

This, if maintained and passed into household prices, would add something like 10% to household bills and more for business. As the $50 per year two time credit has also expired, Queenslanders will get hit with higher prices going into the next election. And with no policy.

In total, power output increased about 220 MW on average for the March 19 quarter and if you are an optimist the renewable share virtually doubled to 11% compared to last year. Of course 11%  is a long way below the 2030 target.

Note that we added rooftop demand to “in front of the meter demand” to get the total.

Getting to the 50 per cent target

The 50 per cent renewable target seems to be specified in terms of production, if it’s specified at all as anything more than a slogan. Production is impacted by net exports. Queensland typically is a net exporter and if exports grew it would imply more renewables relative to 50 per cent of demand and vice versa.

To keep it simple for this note we assume generation is 108 per cent of demand and that demand grows, more or less in line with AEMO assumptions, at 0.8 per cent per year.

Also, the 50 per cent target is generally expressed in energy (TWh) rather than average power (GW).

About capacity factors in comparing power and energy

For variable renewable energy [VRE], solar single axis tracking capacity factors are probably around 30% AC. They are discretionary and depend on the DC/AC ratio.

The AC comes from the inverter and the DC from the panels. Panel costs are declining faster than other costs (inverters, grid connection etc) and so there is a tendency to put on more and more panels per inverter , this is the Inverter Load Factor [ILF].

Wind capacity factors go up in theory more than in practice. Well, perhaps that is tongue in cheek. Still we anticipate maybe 40 per cent.

So a 50:50 mix of new wind and solar would get you 35% average capacity factor but despite the fact that there is (much) more wind in QLD than I used to think and despite that QLD wind is supposed to be less correlated with the rest of the NEM – making it more valuable to the system (but not necessarily to the wind developer) – despite those things we still think there will be more solar than wind in QLD.

This leads to a simplified average VRE capacity factor assumption of 32%. It doesn’t make much difference in truth.

So we can express the overall task as follows:

Figure 2: 50% renewables in QLD. Source: NEM Review, ITKe

About 22 TWh of new VRE is required and almost equally a 24 TWh fall in thermal generation (coal and gas) is required.

We can also allow for generation under construction and the annual requirement is about 650 MW of utility-scale renewables. Possibly more if rooftop slows or electric vehicles get a move on, or pumped hydro really pick up growth in the latter half the 2020s.

Figure 3: QLD’s $20 bn, 10 GW task. Source: ITK

We already know that getting to 50 per cent renewables across the entire NEM only takes about another 16GW of VRE so proportionately the task is much harder in QLD.

Tasmania, South Australia and Victoria are now much further advanced. Queensland has relatively little hydro generation so it started in a difficult position.

The 750MW under construction is the Coopers Gap wind farm and Yarraneea, Haughton and Clermont solar farms. Basically, QLD needs to do almost that much  every year out to 2030.

There are also probably 500MW of solar farms that are only part way through commissioning.

Rooftop: We optimistically stat bullish despite no policy support

QLD’s tiny (3,500 houshold) solar solar/battery scheme is just a joke in the bigger plan and says more about the QLD budget than it does about policy.

Overall there  are about 590K houses with solar in QLD or about  37% of detached houses. Over the next 11 years perhaps 66% penetration is reasonable or say another 540 k houses. We allow construction for the commercial rooftop market of 85 MW per year, perhaps a bit on the ambitious side.

All up maybe  rooftop can  provide closer to 6 than 5 TWh of the additional 28 TWh required for 50% renewable generation in QLD by 2030

Figure 4: Rooftop solar estimates. Source: APVI, ABS, ITK

It’s probably worth adding that QLD Govt announced a target of 1 m households and 3 GW of solar PV in QLD by 2020.  The 3 GW target will be met if utility projects are included but there is no chance of meeting the 1 mn household target by that time. There’s a target badly missed. Without some further policy support we see material risk of undershooting Fig 4.

Pre-election policy – scorecard two years on

Before the last  QLD election the QLD Govt.  made various electricity  policy announcements.

These included:

1. 50% renewable by 2030;
3. Establishment of Cleanco to own the “cleaner” that is hydro and gas assets of the Stanwell and CS Energy, the two QLD Govt owned Generators:
5. A 400 MW reverse auction program was contemplated and expressions of interest were sought but that particular program has been abandoned in all but name. We expect it to start afresh under the Cleanco banner
7. Powering North Queensland $380 m commitment and strategy

Looking at these my scorecard reads

50% renewable target. You could say it’s on track based on the doubling share over the past year and the 750 MW under construction. However  at least another 500 MW of commitments are needed this year.

CleanCo has legislation and it also has an interim CEO, none other than former Infigen CEO Miles George. This completes a nice back to back  arrangement by which ex QLD Govt people move to Infigen eg  current Infigen CEO Ross Rolfe and Infigen people move the other way. I anticipate most policy near term will be executed through Cleanco (see discussion below).

Powering North Queensland, Dr Lynham seems to have pronounced it on life support, we’ll call you

Prior to the 2017 QLD election  Powering North Queensland was meant to provide transmission to otherwise stranded electricity projects West of Cairns and Townsville.

The QLD Govt announced $386 m would be invested in Powering North Queensland, but that of course was only going to be part of the cost.

Since then the industry has moved to a view that Powering North Queensland may well need to be tied to “Copperstring 2.0”.  Copperstring, is surprisingly not a Scott Morrison phrase but nevertheless refers to connecting Mt Isa to the grid.

An earlier Copperstring 1 was beaten out by APA and AGL who instead  built the Diamantina off grid combined cycle 240 MW power station or 300 MW when adding in old but still operational capacity.

However, Diamantina only worked due to a 5 year cheap gas supply from AGL. That deal is expiring  and gas at $9 or $10 GJ will mean fairly expensive electricity. For instance $9 GJ gas at heat rate of 7.5 GJ/MWh maybe a touch more is about $67 MWh of electricity in terms of fuel cost.

Another benefit of the original powering North Queensland strategy was reducing the Queensland community service obligation [CSO]. This is a subsidy paid by the rest of Queensland to regional Queenslanders and is budgeted at around $465 m per year. In prior years its been as much as $600 and $700 m.

Much of that subsidy is paid to North Queensland. Windlab noted in a 2017 presentation that Kennedy 2, a 1200 MW, $2.5 bn project could reduce that subsidy by say $40-$50 m per year. At a discount rate  of 7% the NPV of $50 m  annually is $660 m. That’s a lot of savings.

In any event in the two years since Powering North Queensland strategy was announced there has been no public information.

On March 5 “The Townsville Bulletin” wrote that the Government was considering a feasibility study by Powerlink but the Government in the form of Dr Lynham declined to provide any further information.

QLD’s State owned generators, run for cash

Controversially I think QLD should have sold its coal generators (and wires and poles) and reduced State Debt. It could have used some of the proceeds to boost Cleanco up much more quickly and by 2030 could have ended up with just as much generation as now but renewable instead of coal.

But socialist ALP ideology doesn’t allow for this. So Queenslanders are left with a 50% renewable policy that will effectively reduce the value of Stanwell and CS Energy very significantly by 2030 and as carbon policy inevitably tightens further will likely require them to be closed shortly after.

QLD Govt policy requires the State owned generators Stanwell and CS Energy to pay 100% of the NPAT to the Government in the form of dividends.

Dividends have increased sharply even though electricity pool  prices, effectively at QLD Govt. mandate  fell for  two years before rising again this year.

Figure 5: QLD Generation dividends. Source: Annual reports, auditor general

Queensland households received a one time, $100 rebate off their bills partly funded from these dividends.

Return on equity is a largely meaningless measure for these businesses as the assets have written up and written down , moved from 3 generators to 2 and now back to 3 again in a way as to make the balance sheet fairly meaningless.  As reported both businesses show  an ROE over 20%. So way better than AGL or ORG.

Ebitda grew over 10% in FY18  assisted by higher coal sales from Curragh mine revenue sharing  and increased output from Gladstone power station.

Figure 6: QLD generator ebitda. Source: Annual Reports

Coal

Most of the coal comes from wholly owned mines, the Callide and Kogan Creek mine for CS energy and the Meandu mine for the Tarong power station. Reserves at Meandu are likely sufficient to carry the power station through to 2031.

Stanwell power station gets its coal from the Curragh mine and Stanwell also makes over $200 m profit (although volatile) from Curragh exports. In FY26 the Stanwell/Curragh contract expires and at current coal prices one might assume Stanwell’s coal cost will rise significantly.

Cleanco – where there might be some action

The asset profile, pool generation and pool revenue for the 12 months to 30 March 2019 are as follows:

Figure 7: Cleanco business, 12 months to Mar 19. Source: NEM Review, ITK, Company

We expect Cleanco to restart the stalled 400 MW reverse auction process and even move up towards 1000 MW of renewable energy, but that may not all be achieved until 2025.

That’s say 15% of the total to 2030 target. We expect that by the end of this calendar year the business will be in operation and will have managed a reverse auctions although obviously the risk is that timetable will slip.

We expect Cleanco to have capital for direct or partnering investment of say $250 m, a working capital  facility and the QTC  finance guarantee improving its ability to support PPAs.

Looking at the existing assets.

Swanbank E was built in 2002 and was shuttered in about 2014 when CS sold its gas entitlement ino the LNG business. That was and is the the highest value use of the gas.

When it was nevertheless returned to service  in 2018 Stanwell made a write down on its value. It’s a single open/combined cycle turbine  in contrast to most combined cycle plants that have 2 or 3 open cycle units and a recovered heat unit to provide the “combined” bit. Our SRMC is based on $9 /GJ gas.

Obviously the great hope is to increase the value of Wivenhoe. 570 MW is certainly a meaningful amount of capacity about 8% of average demand in Queensland. Its well more than double Genex’s proposed Kidston project and only 75 km from Brisbane and it’s already built and ready to operate.

As Snowy has pointed out a pumped hydro operator makes about 40% of its revenue from selling capacity in the form of $300 MWh caps. Snowy can do this easily from its portfolio of  hydro and pumped hydro but it’s probably a little more difficult for a pure pumped hydro operator.

For instance if you sell a $300 cap, have to generate to protect the cap and your pumping cost is $200 MWh you won’t make any money. So this would be a problem in a period of continuous high prices when there is no opportunity to recharge.

In the past 15 months the highest 10 hour average price was $300 MWh and there were 0.4% of 10 hour periods where prices were >$200 MWh. Not too much of a risk. So Wivenhoe could devote some capacity to selling firming products and caps to some of the solar farms in the State.

Wivenhoe has a very juicy 10 hours of storage capacity.

Its efficiency factor before a recent 70 MW capacity increase was around 66%.  So to make any money at all the selling price needs to be 1.5X the buying price. Ie buy at say $66/MWh – 1.5 MWh – and sell 1 MWh at $100, profit = zero.

The following figure, interesting in itself, compares average half hourly prices in the March 2019 quarter with the same quarter a year ago. This is just one quarter, still it’s typically the peakiest quarter  that on a daily basis should offer the most average arbitrage.

Figure 8: QLD March quarter pool prices by time of day. Source: NEM Review

Actually for these daily averages we don’t think there’s enough margin for Wivenhoe buying from the grid.

Figure 9: Wivenhoe March qtr 2019, daily average arbitrage. Source: NEM Review, ITK

However, there will be days and times with much bigger margins. We need another article to go through this in detail.

In short Wivenhoe can do better but its opportunities on a daily basis just now seem to be limited.

For interest we show how the growth of wind and solar are impacting coal, gas, hydro and exports in the following four charts. Each compares the March quarter in 2019 with that in 2018.

Figure 10: QLD, changes in response to more solar and wind. Source: NEM Review