As Coalition hearts coal, solar prices crater

As the Coalition hearts coal today, the US National Renewable Energy Lab (NREL) has released staggering new cost profiles for US utility scale solar:

Solar PV capex cost cratered 29% year on year to a levelised cost of electricity of $0.43kWh. This is way faster than we have previously estimated was possible:

Utility solar + full battery is now $0.16kWh. There’s still some scope for utility PV costs to fall further but the big price drops ahead are in batteries. We remain of the view that with the mad dash in technological advancements underway that battery costs will continue to plunge. At -20% per annum the killer app of solar + battery, which is base load power, will be cheaper than coal a year before Liddell coal power station shuts in 2022:

And, as we see today, prices of solar + battery are falling much faster that -20% per annum.

Our best guess is that for the countries with the best solar resources (that is, Straya), coal power will be obsolete some time around 2021.

Unless some imbecile subsidises it.


  1. But what is the cost of each of these power generating sources to deliver on demand power at 6pm, 9pm, 12MN, 3am and 6am?
    It is overly simplistic to ignore the costs of storage or excess capacity needed to ensure supply 24/7/365
    It is also simplistic to ignore network and distribution costs needed to connect these renewables into the network at scale from their ideal locations.
    But it is also simplistic to ignore the externalities of coal, namely toxics, health and climate change.

      • On an industrial scale?
        We are also, supposedly, converting all transport to battery as well – the loads just from that alone are immense and way outside parameters anyone is thinking about.

      • It is clear you are not operating off grid. You have no idea. There are days anywhere in Australia when the sun shines for well under 1 hour. An example from earlier this year:
        It is impossible to go off grid using only solar and battery with less than 2 days energy supply.

        The best that is possible on a commercial basis using components produced from coal generated electricity is 50c/kWh. If the components used electricity produced from wind and solar sources in their manufacture their cost would be truly astronomical. They are not renewable. They have the illusion of being renewable because coal is being used to generate the electricity that is used in to make them.

    • I’m not sure where the figure 43 cents/kWh came from either.

      My calculations – $2.80/watt US = $3733/kW Aus (at 75cent exchange rate).
      Production – 1300kWh/year/kW installed (at capacity factor 15%)
      Installation cost – $2.87/kWh/year.

      A LCOE of 43 cents/kWh assumes complete payback is necessary in 6 years 8 months (and zero maintenance costs).

      If assumptions included a 20 year complete payback plus maintenance costs of 25% of the initial cost over the 20 year period, the the LCOE figure is 18 cents/kWh.

      All of this is assuming that a 5kW system fully installed costs $18,600 without any rebates or subsidies. You should be able to get it for well under $15,000, probably closer to $10,000, even without RECs or subsidies. Let’s call it $12,000, 20 years, $3000 maintenance over those 20 years – this would bring the LCOE down to 12 cents/kWh.

      • But you are grid dependent for overnight power unless you add in the cost of a battery and upsize/double your solar instalation to charge the battery. All the busniesses in the grid will want to remain viable and so connection costs and retailers will recoup the full costs of their operations irrespective of the volumes being drawn through it.
        Let me make it clear though that I support dramatic increase in renewables for Australia to be making emissions reductions consistent with our history as a historic large emitter, being well endowed with renewable resources and facing an huge risk from climate change.

      • @ Explorer – what you’ve said is exactly true. This will affect the cost/ benefit of rooftop PV compared with no rooftop PV and therefore the financial viability of installing rooftop PV, but it will not affect the LCOE energy produced by the rooftop PV itself. That’s a separate sum.

      • with our history as a historic large emitter

        We are only a large emitter if you use per capita metrics, which are elevated by the relatively small population. If you used per km^2, then we would not be a large emitter. If you used total emissions on a global scale, then we would be a small emitter.

        If you used a metric that measured the percentage of global emissions affected per vote (which would be a measure of the emissions effectiveness of our political will), then small again.

        Australia is very well placed to exploit solar energy, especially EVs. Why can’t the argument be about that rather than the sackcloth and ashes and ideological self flagellation?

      • The components needed to reliably meet an average daily demand of 15kWh from solar alone is 15kW of panels and 30kWh of batteries. That system will at least collect and store enough energy in southern Australia to ensure power for 999 days out of 1000.

        Even with low cost installation and low cost finance the 5500kWh produced over a year will cost 50c/kWh. The achieved capacity factor is 365/8760 or a lousy 4.1%. Once you understand the reasons for this you begin to appreciate why grid power is rising in price as more wind and solar is connected.

        You cannot rely on more than 2 hours of full sunshine equivalent in any two consecutive days. That is the basis of design.

      • @RickW-MB

        Nobody’s talking about off-grid. Nobody’s talking about finance – otherwise your LCOE would involve much less equity. Nobody’s talking about capacity factors of 4% because you’ve wasted 75% of your energy.

        And nobody else is using crazy-whacko assumptions written by the coal companies.

      • @Rick, the need for a FF generator is one of the things that makes Hybrid Cars attractive.
        A Prius has a 1.5L ICE coupled to a 20Kw electric generator. The output charges a Hv battery (350V from memory). The clear solution is to find ways to use this “free” 20Kw generator to charge your household battery pack. This is why I’m more interested in systems with 7 to 10Kw arrays and 10 to 15Kwh batteries…you just need to be sure the petrol tank is full if you expect a week of gloomy weather. .

  2. i dont buy the energy cornucopian view proffered by renewables advocates. not that i’m a fan of fossil fuel energy either, but i think its probable the future doesnt belong to either. rather, just less energy and less consumption in general.

    and if so, that’s okay by me. maybe we’ll even live in villages again.

  3. To discuss the issue sensibly we need something more than wishful thinking on battery technology and costing. There have been some very good posts by people who seem to have some sort of credible evidence and logic on the topic that cast, at a minimum, doubt on this assumption. Batteries are NOT computers. They have a large input of physical resources and this in itself puts limits on their widespread adoption across the world.
    Further, given the long term determination to make Aus the home of the poor white trash, it’s likely that, long term, we will be outbid for whatever resources are required.
    There are just too many fairy stories surrounding this stuff on all sides.

    • My evidence is above. In the numbers. Where’s yours?

      We have never argued that a variety of base load won’t be needed. And I note above that the solar baseload option is competititive for those with the “best resources”.

      You offer nothing but doubt and vague gesturing at interest rates.

      • I ran some numbers a while back on the chances of getting transport running on batteries & have been meaning to extend those to utility storage for some time now. I’ll spend some time and try to get them done for the weekend links. If they come out anything like the numbers for transport, then flawse will probably get the guernsey.

      • I have done the exercise. The NEM has a daily average energy demand of 550GWh. To supply that from battery storage with solar arrays located in central Australia requires 750GWh battery capacity and a 250GW solar capacity. The required solar capacity goes up somewhat if arrays are near the coast.

        With the numbers currently being thrown around the AUD1tr or so needed to build this is not far off the CSIRO projections of between AUD800bn and AUD1100bn for 100% renewable. What is not understood is that these costs are based on buying components built using coal generated electricity in China. Such a system is in no way sustainable if it was built using the electricity it produced. The ratio of reliable energy out to the energy it takes to make them is around 2.1. That is well below what is sustainable. It is an illusion of sustainability.

    • Fl you are correct. As currently offered solar +battery is good for about 4 days max. After that it is back to base load for the normal usage + to recharge the battery. the grid wont stand it especially if you add vehicle recharge.
      there are plenty of times during a year when solar is not available for period in excess of 4 days. say 20.
      so if you go solar with batteries you need your own generator for those 20 times per year HC take note
      WW solar guru.

      • Noted. In my experience in our area this is most probably in EC lows and ex tropical cyclones. Notably, Debbie and her cloud cover ran for, IIRC, 10-14 days straight of rain. (IIRC we had 500mm in a week and 750mm across the event.) Then we had, soon thereafter, the June 10 day event. So there’s the 20 days.

        Since then (and yes it’s a small sample size), we’ve had not one instance of a cloudy day since about June 14, until today.

        Our usage – deep freeze, solar HW, 3kW PV, computers, coffee machine on most of the day, fridge, etc, is about 7kWh.

        My plan is for a 14kW powerwall so in the event of an EC low or ex TC I’ll have an input for diesel gen (that hopefully can run on cleansed vegetable oil).

        But, we’ll be on the grid.

        For us it’s minimising grid reliance, and having power when we have na outage, to the extent power bills are $0. It’s not about going off grid.

  4. I’m all for renewables playing the major role.

    The only thing that concerns me is that we would be hanging the fate of our civilization on something less reliable. Come rain, doldrums, volcanic eruption, just about any other catastrophe you can name – coal, gas and uranium will always burn and run generators. They are reliable and we have oodles of them, even if we just give the gas away and let the companies shaft us at this point in time.

    There’s probably a millenia worth of coal in the ground in this country at current domestic usage rates – how long do batteries last in the event of a wide-scale catastrophe?

  5. Cost per DC watt, you’d have to add +12-15% for power loss for AC conversion.

    None the less still good news, I’m installing 24 X 240w panels in a fortnight, just the feed in tariff of 7cents puts myself on a 6-8% expected yield P/A.

  6. Thermal molten salt can store base load solar and has been in large scale use around the world for half a decade and is the fastest growing sector. China has just installed the single largest array on earth.

    Why ? Why is this never included in the mix when it is the most established and common form of Solar ?

    Secondly why is it that this LNP gets everything wrong with regards to everything. NBN, Energy, Census, etc, etc ?

    Are they simply Luddites ?

    • Di Natale again failed to mention pumped hydro and CSP this morning. Geez he’s useless.

      Andrew Broad was good listening on RN this morning. You can tell he’s got half a brain and now he’s spent time on this stuff, appears to get it

    • The LNP are in thrall to the Neo-liberal philosophy of Small Government.
      They expect that by giving away the leavers of government to private enterprise, that their responsabilities will be reduced.
      This is essentially lazy politics that leads to lazy government. After awhile it shows.

  7. This is really good news. Can someone explain why we even need a CET if solar + battery is soon going to be the cheapest baseload electricity? Won’t the market just sort it out?

    • You are correct. And the crash in solar panel prices had nothing to do with AUS. It was thanks to Germany, China, USA.

      The RET should be up to each state – not the feds. ACT wants a 90% RET – cool. California has a 50% RET. Good on them.

    • None of the cost data on wind and solar takes the intermittency into account. It is based on the assumption that the grid currently works like a big battery that can take in excess energy when it is available and delivery energy when the generation is low.

      The LCoE for wind is typically based on CF of 35% and solar and solar 18%. For low levels of penetration those capacity factors are possible. However as the connected capacity increases the operating capacity factors reduces. At 100% renewable wind ends up around 5% and solar around 4%.

      As an example SA has 1600MW of wind and 1000MW of solar connected to supply a demand of 1200MW. That installed capacity achieved 39% market share. To get to 100% it would need about 5 times more generating capacity plus massive storage. So cost go up dramatically with increasing penetration.

  8. What I’m REALLY interested in is their “price” for utility scale single axis tracking PV.

    $1.50/watt Australian (a smidge under)! The cheapest it’s been done in Australia as far as I know is just over $2/watt (AC).

    SAT PV is the big one – it has a higher capacity factor than fixed tilt (up to 30% on the sunny parts of the mainland, approx 25% in the sunny parts of Tassie, compared with a CF of just over 20% on the mainland if you’re lucky for fixed tilt). This allows it to produce more GWh per year for less MW of peak production, which means that less power needs to be stored for later or wasted.

    It also optimises PV production from as soon as the sun comes up (or is about 5-10 degrees above the horizon, depending on how the rows are spaced), until the sun is almost down, hence morning and evening production is much better.

    People talk about “sometimes the sun doesn’t shine for days”. That’s true for the coastal vales of Australia (Brisbane, Sydney, Melbourne), but if you ask people in Mildura or Broken Hill or Emerald they’ll probably tell you a different story. One overcast, stormy day, full sun the next day. The point being that less storage is required for utility PV than wind, and still less for SAT PV than fixed tilt PV.

    At $1.50/watt, 20 year payoff, 25% CF, and 25% of its capital costs as maintenance spread over 20 years, this equates to LCOE of $43/MWh, ie, 4.3 cents/kWh. I’d take those prices in today’s energy market. Plus, despite having a lower CF than wind, it’s probably has the most useful power generation pattern.

    • TT you have put some effort into that story, but go back to the start, tracking is not a viable option, otherwise everyone would do it.
      second. most of the strayan inland is non populated, unless you are in jv with jacob and r2m for the installation of hvdc the power generated there is stranded. being from launceston you know the value of the bass st link. Still no answer why that failed??
      why you guys dont fill the west coast of your joint with wind generators and pumped hydro is beyond me.??? politics???

      • Clearly SAT PV is hopeless for domestic systems, but it can, has, and is being done in Australia on a utility scale. Moree, Barcaldine, Gannawarra, Swan Hill.

        None of these have been built or have been proposed to be built for under $2/watt AC. My understanding is that fixed tilt is still slightly cheaper per MWh produced, and since the incentives are based on number of MWhs rather than the utility of the spread of when those MWhs are produced, then this is how most solar farms have been built to date.

        As for Tassie – wind farms – great idea. Pumped hydro – a bit trickier – high capital costs and potentially high environmental costs, so it really needs to be worth it. I’m not sure that the numbers are there yet – at least until we’ve got a lot more installed wind/ solar capacity. Until then, wind or PV generation will primarily be “stored” by displacing hydro generation which would otherwise have been produced at that time, hence the water in storage remains at the top of the dam instead of at the bottom of the dam.

        Also regarding pumped hydro – there are plenty of equally good sites on the mainland as there are in Tassie – we really don’t have a competitive advantage over the mainland in having that form of storage down here.

    • For me the big unknown is will these cheap panels really last 25 years?
      At each cost down new sealants are being used and new structural assemblies / assembly methods developed, these methods might prove adequate in benign climates but the Aussie outback is far from a benign climate. I guess only time will tell, but informed opinion seems centered around a 10 years service life for modern cheap panels.

      • cb you’ve missed it, my patent application for uses of the new generation of panels is near written
        look at what the new construction offers, not what it takes away.
        when you see it rolled out (hopefully by xmas,) you will immediately understand WW

  9. I’m concerned optimistic forecasts in the renewables area has got us into the mess we are in. What is meant by coal will be obsolete by 2021. Coal currently provides over 80% of our energy needs. It seems that renewables can create problems of their own when it comes to integration to the grid and we are still some way from large scale storage that can provide the baseload and dispatability we require to operate modern electricity grids in large cities.

    I’m not as confident as House and Holes, in fact, I’m leaning toward pessimism. I’m thinking here in Melbourne that the Andrews plan is a major risk. This comment from yesterday was an eye opener (funnily my response to Turncoat’s comment below was to suggest Macrobusiness take up the baton, I now realise that is unlikely!)

  10. Diogenes the CynicMEMBER

    We’d need less baseload if we put utility PV west of the major consumption centres i.e. they have sun for a bit longer than the consumption centres do in the evening. If we had a decent interconnector to SA you could have solar in the middle of Aus (west NSW, east SA) powering the Syd/Melb evening peak. If you paid a higher FiT for panels facing west in those cities you could also cut off a bit of the peak off as well. But energy policy in this country is crazy so it won’t happen.

Leave a reply

You must be logged in to post a comment. Log in now