Australia has a problem. Contracts for wholesale electricity prices across the National Electricity Market have soared in the last 12 months to some of the highest in the industrialised world. Meanwhile our electricity supplies are the second most carbon pollution intensive in the developed world, second only to Estonia.
In our report – Overcoming ideology to support new power plant investment and reduce power prices – Could Judith Sloan and Chris Kenny hold the answer? – we outline how we got ourselves into this mess and how we might get ourselves out of it.
This probably begins by Malcolm Turnbull following the advice in his own February speech to the National Press Club when he stated,
The next incarnation of our national energy policy should be technology agnostic. It’s security and cost that matters most, not how you deliver it.
Policy should be all of the above technologies, working together to deliver the trifecta of secure and affordable power while meeting our emission reduction commitments.
Our Resources Minister Matthew Canavan thinks he’s got the solution. The lights are going out in South Australia, he opines, due to too much renewable energy. So what’s required is for the government to directly underwrite a brand new coal power station several thousand kilometres away in a marginal seat in Northern Queensland.
Of course, he’s somewhat onto something in that Queensland wholesale power prices were not far behind South Australia’s last year ($59.99/MWh vs SA’s $61.67/MWh), and this year they’re tracking at almost $100 per megawatt-hour and hot on the heels of SA’s price.
But we’re not so sure this is for a lack of baseload coal in Queensland. All up, Queensland has a tad over 8,000 megawatts of good ol’ baseload coal capacity when its average demand for electricity is 6000 megawatts.
Minister Canavan also reckons new coal power stations are really clean these days. He has suggested that rather than encourage renewable energy, building lots of new coal power stations will deliver our emission reduction targets.
Now you might not agree with Minister Canavan, but rather than argue about it, why don’t we let the market sort this out?
This is where Judith Sloan, columnist for The Australian newspaper and general all round hater of renewable energy, has come up with what could be a very bright idea.
In a January 24 column she bemoans what she claims is the high cost of the Renewable Energy Target (RET) and suggests,
There is a solution the government should consider, and that is to revise the RET to be more broadly based and technology-neutral. It won’t be a case of ditching the RET, but improving the RET. Let us call it the Clean Energy Target that will use underlying clean-energy certificates that are weighted on the basis of the emissions intensity of the source of the electricity. The CET could formally run to 2030 and beyond, thereby giving all players some certainty.
The aim of policy should be to achieve a given reduction in emissions at least cost, not to favour one form of energy at the expense of all others. The CET meets this criterion in a way the RET does not.
Now, in case anyone in the Liberal-National party thought Judith had lost the plot suggesting we extend the RET, former Liberal Party ministerial adviser and right-wing political commentator, Chris Kenny observed,
But the best idea for immediate action has come from my colleague Judith Sloan.
She says that rather than fund only renewable energy the RET certificates could also fund non-renewable projects that are more efficient. They could support clean coal or gas generation, so long as they are lowering emissions, comparatively.
We think this is a brilliant idea, especially given the government has ruled out emissions intensity trading scheme, and the existing Renewable Energy Target will effectively cease to support new investment and abatement in a few years’ time.
In fact, we think you could go even one step further in helping the Liberal Party claim this policy as their own – merge it with aspects of the Direct Action Emission Reduction Fund.
The biggest problem with schemes like the RET and also an Emissions Intensity Trading scheme or the prior carbon price is that low carbon investments live under a shadow of fear that Tony Abbott might arise from the backbench.
He, or someone like him, could amend the scheme’s legislation rendering the abatement certificate revenue from the project worthless, leading to bankers knocking at the door.
The Direct Action Emission Reduction Fund circumvents this by providing a contract with the government to the lowest cost bidders in auctions. If the government breaks the contract proponents can seek compensation via court action. Something unavailable to them in the event a RET or carbon price is abolished.
Why environmentalists and the renewable energy sector should welcome Judith Sloan’s idea
Now before all you environmentalists choke on your Weeties at the prospect of letting fossil fuels into the renewable energy scheme– let us let you in on a secret.
Provided such a scheme is genuinely designed to achieve our 2030 emission reduction target it will be almost entirely be dominated by renewable energy projects. The thing is that Sloan, Kenny and probably also Minister Canavan don’t seem to have a clue about the latest economics of renewable energy projects relative to fossil fuels.
We note that RenewEconomy’s editor Giles Parkinson has highlighted concerns that an emissions intensity scheme (which is what Sloan is unwittingly proposing a derivation of) would drive a large expansion of gas power and very little additional renewable energy, based on modelling by Frontier Economics.
However, Frontier Economics’ modelling is based on an out of date assessment of the economics of wind and solar projects, and also has some incredibly optimistic assumptions about the future price and availability of gas.
Our report goes blow by blow through all of this, but we think it is particularly worthwhile highlighting some of the misconceptions with ground-mount solar project economics. The dramatic cost reductions achieved by solar farms in this country are a complete game changer for the costs of emission mitigation policy which analysts must urgently recognise.
For those interested in the nitty gritty detail you can read on in the addendum below (or better still read our report).
However, to cut the story short for the rest of you, Frontier estimated that ground mount solar projects would cost $117/MWh in 2016 declining to $85 by 2040. By contrast our understanding is that solar farm power purchase agreements are being signed right now in Australia for between $75 to $85 without ARENA grant funding.
(Ed: Ex Hazelwood boss Tony Concannon confirmed this price – $75/MWh – and has started building a 220MW plant at that price in South Australia through his new venture at Reach Solar).
If Frontier Economics got up to speed with what wind and solar project developers can achieve, and adopted more realistic gas price assumptions, they’d have found that the predominant driver of lowest-cost abatement under an emissions intensity scheme would be through expansion of wind and solar power.
They’d have also found that expanding renewable energy targets would have delivered very substantial reductions in consumer power bills relative to if we keep climate policies unchanged.
Addendum: Where Frontier Economics went wrong on Solar Costs
Judging from Frontier’s report, it appears they were misled by data reported below by the Australian Renewable Energy Agency on the costs of solar projects they funded (see table below). Frontier note in their report that they used these costings as the basis for their base case estimates that power from ground mount solar projects would cost $117/MWh in 2016 declining to $80 by 2040.
|Developer||Project||MW – AC||MW – DC|
|Total project cost||$/kW (AC)||$/kW (DC)|
|Origin Energy||Darling Downs Solar Farm||110.0||132.0||$216.7 m||$1,970||$1,642|
|Edify Energy with Solar Choice||Whitsunday Solar Farm||58.1||69.7||$122.4 m||$2,107||$1,756|
|Neoen Australia||Parkes Solar Farm||50.6||60.7||$107.9 m||$2,132||$1,777|
|Genex Power||Kidston Solar Farm||50.0||60.0||$126.2 m||$2,524||$2,103|
|Manildra Solar Farm||Manildra Solar Farm||42.5||51.0||$109.3 m||$2,572||$2,143|
|RATCH Australia Corporation||Collinsville Solar Power Station||42.0||50.4||$95.9 m||$2,283||$1,903|
|Neoen Australia||Griffith Solar Farm||25.0||30.0||$54.6 m||$2,184||$1,820|
|Canadian Solar (Australia)||Oakey Solar Farm||25.0||30.0||$47.5 m||$1,900||$1,583|
|Neoen Australia||Dubbo Solar Farm||24.2||29.0||$55.6 m||$2,298||$1,915|
|APT Pipeline (APA Group)||Emu Downs Solar Farm||20.0||24.0||$47.2 m||$2,360||$1,967|
|Goldwind Australia||White Rock Solar Farm||20.0||24.0||$44.5 m||$2,225||$1,854|
|Canadian Solar (Australia)||Longreach Solar Farm||15.0||18.0||$28.7 m||$1,913||$1,594|
At Green Energy Markets we looked at this data and were left puzzled. These costs are in many cases substantially higher than what those in the rooftop solar sector can achieve doing projects that are a tiny fraction of the size of the projects above.
We know of large suppliers using good quality tier 1 panels and inverters that will deliver a fully installed 10 kilowatt solar system for $1400 per kilowatt of direct current (DC) capacity (excluding STC rebates and GST).
They’re also notably more expensive than what is being achieved internationally for large ground mount projects. Now to be fair the projects above were quoted in terms of their inverter’s alternating current capacity, and ARENA information indicates the amount of panel DC capacity is about 20% higher.
Yet after adjusting for this we still end up with costs that on average are 30% higher than what is typically achieved for a 10 kilowatt system. Admittedly the rooftop systems avoid costs involved in ground-mount support framing.
They also face far less onerous grid connection requirements. Yet you’d expect projects that are thousands of times larger in capacity would more than make up for this via efficiencies in labour costs and reduced equipment prices from buying in bulk. What’s also interesting is the wide variation in costs quoted. While Canadian Solar’s projects at just under $1600 per kilowatt DC were very close in cost to what we see in the rooftop segment, others were as much as 30% greater than this.
Location might help explain the higher costs faced by the Kidston project given its remote far north location, but Manildra is just 300 kilometres from Sydney.
So we started asking around the industry talking to project developers, equipment suppliers and construction companies to understand whether in fact the industry could manage to achieve costs far better than reflected in the ARENA bid data.
It turns out ARENA may have been more successful than it realised in overseeing a reduction in the cost of ground mount solar. Some developers using fixed tilt framing are achieving costs as low as $1,300 per kilowatt (DC) including grid connection, and the general consensus is that $1,500 per kilowatt (DC) employing single axis tracking is readily achievable.
Once you take into account the oversizing of DC panel capacity relative to the size of the AC grid connection, the consensus is that costs of around $1800 per kilowatt on an AC basis can be achieved.
So why is the ARENA data so wrong? We suspect it could be a product of the fact that proponents prepared their project costings for the ARENA process several months before this data was published. At the time these costings were prepared the market for construction of ground mount solar projects was almost non-existent.
In the intervening period the ARENA process spurred several construction companies and equipment suppliers to enter or prioritise the Australian utility-scale solar market (assisted by the downturn in mining construction activity). This led to much greater levels of competition supported by significant international price declines in solar PV modules, inverters and trackers over the selection process period.
Another area where Frontier isn’t on top of latest developments is the capacity factor of ground mount solar projects (the amount of output you get out of a given amount of solar capacity).
Frontier aren’t entirely clear about what capacity factor they assume for solar in their base case estimates, but the report indicates that 22% was what they considered the best that could be achieved on an AC rated basis. It turns out that this is a likely to be a very significant underestimate of what Australian ground mount projects are likely to achieve.
The table below details the AC rated capacity factors achieved for Australian operational megawatt-scale solar farms for which we have more than 12 months’ generation data. Three of the five have capacity factors comfortably exceeding what Frontier estimates to be the best that could be potentially achieved. The other two, which are located in Victoria and the ACT – some of the poorer solar irradiance locations considered for solar farms – fell only just short of what Frontier considers as its best case for solar.
|Project||State||AC Capacity (MW)||Measurement period||Capacity factor|
|Nyngan Solar Plant||NSW||102||Jan-Dec 2016|
|Broken Hill Solar Plant||NSW||53||Jan-Dec 2016|
|Greenough River Solar Farm||WA||10||Jan 2013 – Dec 2015|
|Royalla Solar Farm||ACT||20||Jan-Dec 2015|
|Mildura Solar Park 1 – Koorlong||VIC||3.2||Jan – Dec 2015|
Furthermore, most solar farm projects being pursued in Australia are now looking to employ single axis tracking frames which rotate the panels with the sun. These can be expected to achieve significantly higher output yields. Statements by project proponents on the expected output of projects employing tracking such as Clare, Kidston, and Moree indicate capacity factors in the realm of 28% to 31% (pre transmission losses).
What is all rather odd though about Frontier’s solar capacity factor assumptions is that while they adopted ARENA’s published cost estimates for solar projects, they seem to have failed to notice ARENA also published data estimating these projects’ capacity factors. This indicated average capacity factors close to 25% (now likely to be too conservative given widespread adoption of trackers), rather than the 22% Frontier took as its best case scenario.
Again it is worth emphasising the supply of potential solar farm sites capable of achieving these types of economics in Australia are very large. You’re looking at a geographic area stretching down as low as Mildura and bordered by the western section of the Great Dividing Range. With judicious upgrades to transmission infrastructure the major constraint is actually the level of day time demand for power.
Taken together these indicate that developers can construct solar farms for around 18% less than Frontier assumes while achieving between 13% to as much as 41% greater electrical output. Our understanding is that developers are capable of financing solar projects today at prices of around $75 to $85/MWh (where a long-term power purchase agreement is in place) and the expectation is that this would decline noticeably over the next ten years. This marks a stark contrast with Frontier’s base case assessment that costs would not reach $100/MWh for around a decade and not reach $85/MWh until 2040.
We would point out that the information outlined above is widespread knowledge amongst those actively involved in the supply and construction of wind and solar farms and those involved in the gas market.
They are not a tightly held secret that only Green Energy Markets has been privy to. The chart below, taken from a presentation given by Origin Energy to investors back in August 2016 largely concurs with our own feedback about the likely contract price required to support wind and solar projects going forward.
The challenge for analysts of this market and policy makers is that the economics of these projects have been changing quite rapidly. Data that is just a year old is almost next to useless particularly in relation to utility-scale solar in Australia. Indeed it’s worth noting that Origin Energy in a subsequent February presentation this year is now indicating PPA prices for 2017 lie between $65-$80/MWh covering both solar and wind.
It’s possible things may settle down in another twelve months given Australia’s utility scale solar sector is only just emerging. But it is incumbent upon policy makers, analysts and commentators that they regularly communicate with market participants to inform their views, rather than rely solely on published information that often lags the market by 12 to 24 months.
Ric Brazzale is the Managing Director and Tristan Edis is a Director with Green Energy Markets. Green Energy Markets assists clients make informed investment, trading and policy decisions in carbon abatement markets.