Large scale solar costs near parity with wind energy in Australia

Print Friendly

There are a couple of big and exciting things about to happen in Australian energy markets: one is the rollout of battery storage and smart software in households, businesses and networks. Another is the anticipated start of the boom in large-scale solar installations, as envisaged by chief scientist Alan Finkel this week.

Until recently, virtually the only choice available to developers looking to build under the federal government’s renewable energy target was wind energy, much to the displeasure of a bunch of folk, including the former prime minister, treasurer and many backbenchers and ministers in the current Coalition government.

That is now changing. Within a year or two, all things been equal and if the current investment drought ends, it is assumed that large-scale solar farms will be able to compete with wind energy on costs. Which is not to say that wind energy could, or should, be sidelined. It will just offer more choice and diversity.

The Australian Renewable Energy Agency has been conducting its well-publicised tender for $100 million of grant funding, which it says should help about 200MW of large-scale capacity get built.

It has turned into a giant fact-finding mission, with some 77 projects applying for money and 22 making it to the next round, where they will have to get serious and produce detailed plans and the ability to finance the project.

ARENA has now released some of the data on its website, and it reveals some interesting information about capacity factors, capital costs and operating costs of large-scale solar projects in Australia.

It suggests that solar energy could be built at around $110-$120/MWh in Australia, although with cheaper financing this could be lowered by around another $10/MWh.

This is significantly cheaper than the $135/MWh benchmark set by ARENA, although the agency suspected that this might be the result, given the prompting of some competitive bidding. It means that the additional subsidy (ARENA grants) might only amount to between 10 and 20 per cent of project costs.

“(This data) is showing that costs are coming down, and in fact they have come down dramatically, which is where we need to get to wind parity, which is one of our goals,” ARENA CEO Ivor Frischknecht told RenewEconomy.

Frischknecht said while wind is a relatively mature technology in Australia, large-scale solar is not, so there is potential to lower the costs of inputs that can be controlled in Australia – labour, brackets, frames, manufacturing, maintenance, and the cost of capital.

“If we can get the costs down as low as possible – then developers will have a choice. In southern Australia it might be wind, in other areas it could be sun. In some it might be both.”

This means diversity of supply – wind usually produces more during the night, solar PV is definitely a day-time supply.

Large scale solar is a relatively new market, with only a handful of completed projects built to date – the ARENA-funded Nyngan (102MW), Broken Hill 50MW) and Moree (57MW), along with the ACT-government contracted Royalla solar farm (20MW), and the 10MW Greenough River plant in WA.

There are three critical elements to the costs puzzle – the strength of the solar resource and the potential capacity factors of the plant (i.e. how much they can produce), the cost of material supply, labour, construction, frames, and maintenance, and the cost of finance.

On capacity factors (below), there are no real surprises. It shows that Western Australia and Queensland are the best places for large-scale solar in the country, with capacity factors of 28 per cent and 25 per cent respectively (compared to an average of below 23 per cent elsewhere).arena solar capacity factor

Solar plants with tracking technology increased the capacity factor in Queensland to 28 per cent. This high capacity factor may explain why Queensland projects accounted for 47 per cent of the applications, and nearly half of the 22 projects selected to go forward into the next round.

(It is important to note that the WA figure includes plants that were both fixed and tracking)

The next slide presents a bit of a surprise, with NSW showing significantly cheaper capital costs per watt than other states. The folk at ARENA are still scratching their heads about this one, but think it might have something to do with the location of the proposed projects, better supply chain or possibly lower labour costs.

arena solar capex

(DC – the blue dots, represents the amount of electricity generated by the plant. AC, the green dots, represents the amount that can be fed into the grid. As a rough guide, a capital cost of around $2.20/watt AC is expected to translate into a levelised cost of generation of around $125/MWh, based on ARENA’s 10 per cent weighted average cost of capital methodology. However, a 1% reduction in the assumed WACC translates to a $7 reduction in the “levelised cost of energy”, so if the borrowing cost can be lowered to around 7 per cent, then the cost of generation will fall to nearly $100/MWh).

The next slide is the comparison of operating costs. No great surprises here, but South Australia and Western Australia were significantly higher than their peers, possibly due to the location and therefore labour costs.

arena solar opex

The next round will be the critical one. That is when the 22 projects need to finalise their approvals, their siting, their design and, most critically, their access to finance.

Some projects may benefit from a power purchase agreement (essentially a contract to buy energy) from a retailer, while others will look at the merchant market (selling into spot wholesale prices).

As mentioned above, financing will be critical because a weighted cost of capital of say, around 7 per cent, could mean a difference of $20/MWh with those only able to get capital at 10 per cent.

That could also mean reducing the difference between solar and wind to around $10/MWh, although wind industry folk will point out the recent pricing at both the Coonooer Bridge wind farm and the Hornsdale projects, indicating costs are coming down with the right structures and wind resource, and competitive bidding.

In recent years, that differential might have been considered enough to give the go-ahead, because of the “premium” value of daytime energy, but that premium has been eroded by the proliferation of rooftop solar, now at a capacity of more than 4.5GW across the country.

The key could be in the ability to obtain power purchase agreements. These are typically indexed, suggesting a PPA of around $100/MWh could be good enough for a project with a cost of capital of 8 per cent.

“We don’t know how much costs are going to change – there is a lot of work that is going to happen between the expressions of interest and the final tender,” Frischknecht says. Submissions for the next round are due in June.

The data set can be found here.  

RenewEconomy Free Daily Newsletter

Share this:

  • Russell Harris

    My gut feel is that differences in $/watt are probably associated with productivity assumptions from the installation contractor rather than any particular “state” issues. A strategic and imaginative approach could deliver a significant advantage in this area.

  • solarguy

    Love to know how they come to the conclusion of 28 and 25% capacity factors.

    • Jo

      Agree, the capacity factor of roof based domestic PV plants is just around 16% in the Sydney area. How can they nearly double that?

      • Gary

        Tracking enables higher capacity factors.
        And optimal siting with no shading.

      • solarguy

        Well no mate, I think your misunderstanding, although I didn’t elaborate, I was thinking that 28% didn’t seem to be enough, especially for tracking arrays.
        I’m just north of Sydney near Newcastle, now as I understand capacity factor, it’s based on a plants 24/7 365 days production. PV without storage can’t get 100% of course, but with storage would on some days, so would be variable. With solar PV I don’t think it should be worked out like a coal plant, where it could be much higher, say 75%.
        On residential PV, take my 1.5kw Array, although I have more, today It produced 8.6kwhs, that’s 5.733 PSH out of max of 10.2kwhs- 6.8PSH,which is 15% less than max. So for today that’s 85% capacity factor for generation that works only during the day. Note, I’m not talking theory here.
        Worked out over a year at the moment I can’t tell you what CP would be with weather and seasonal variations, but I think it would be better than 28% fixed north Array.

        • Gary

          Sorry – Jo is right. Rooftop PV operates in the 15-20% range.
          Your rooftop array would produce 1.5*24 = 36kWh if it operated at 100% cf. 8.6 is 24% (a good day).

        • Jo

          Capacity factor is only related to the PV component (the energy generator) because relating the capacity factor to a battery backup makes no sense.
          Capacity factor is always related to 24 hours. So in your case of a 1.5kW system with a mathematical production of 1.5*24=36kWh over 24 hours and a real production of 8.6kWh the capacity factor is 8.6/36 = 23.9%. This is really a good value because we in the summer half of the year. The average over the year for your area is roughly about 18%.

          • solarguy

            Capacity factor does go up with using storage for renewables, simply because it can operate (generate) for a longer period of time. Of course that depends on the amount of storage which would determine what the capacity factor would turn out be.

      • David Osmond

        A few reasons why these large scale systems get high capacity factors compared to residential PV.

        The capacity factor of domestic PV is based on the DC size of the system (capacity of panels). If you look at the graph in the report, these large-scale systems are getting DC capacity factors of 20-22%. It is only the AC capacity factor that is getting up to 28%. These large scale PV are installing panels with capacity 20-30% larger than their inverters (ie. inverter ratio of 1.2-1.3), which means their AC capacity factor is much higher than their DC capacity factor.

        Second reason, as Gary mentions, is that these large scale systems have optimal alignment and tilt angles, no shading issues, and in some cases tracking, which can also boost output greatly.

        And finally, obviously the solar resource is better for these systems, particularly those in QLD, relative to the typical Sydney home.

  • Shane Oneil

    Being someone your readers have completely disagreed with on windfarms being situated to close to homes and lives, this is great news , if simon corbel wants to talk about situating a 20mw solar farm on my property he should come out and see me, already having 10kw system ground mounted l have observed the extra fodder growth that is a result of the shade so would work on another level . Also easily mounted fire suppression in bushfire season and would create a perfect refuge. .

  • Ian

    Large scale solar is said to be more cost effective than roof top solar. Here are my calculations:

    1x5x300 KWH/year per 1 KW installed
    22500KWH/KW over life of panels
    $1000/KW installed
    5% interest
    15year life

    Repayment 7.91/month

    7 year life 10500KWH total
    Repayment $14/month

    Rooftop solar is more cost effective than large scale solar

    Oh, I failed to mention, the only overheads for rooftop solar are the panels lying on the roof! No poles, no wires, no accounts department, $63/MWH is the cost to the household. Beat that coal, beat that wind, and large scale solar, hydro and gas what’s your best offer!

  • neroden

    My bet — NSW has shorter transportation from dockside or factory to installation location. I bet the capital costs everywhere else include higher transportation costs for the solar panels & hardware.

  • sunoba

    Terrific initiative by ARENA. I look forward to hearing about the successful applicants in due course. In the meantime, I’ve analysed the ARENA data in the context of my standard LCOE methodology, which is explicitly aimed at international comparisons. I get the average LCOE as AUD 144 per MWh. Details at (post for 9 March 2016).