Are millions of solar roofs making solar farms pointless?

solar roofs and a solar field

Are solar roofs making solar power stations uneconomic?

Utility scale solar, or solar farms, are fields of PV panels which generate electricity that is fed directly into the grid.  Currently we don’t have much of this in Australia.  Over 99% of our solar capacity is point of use which is mostly on rooftops and the juice it does produce is first used to power the household or business it is on top of and then after that the excess is generally fed into the grid.

While there are countries that have gigawatts of utility scale solar, sunny Australia had none until 2012 when a small 10 megawatt solar farm was opened in Western Australia and no more was built for two years until the completion of the 20 megawatt Royalla solar farm near Canberra last month.  This new development resulted in utility scale solar soaring from about a quarter of a percent of the country’s total PV capacity to a not at all massive eight tenths of a percent.  But there is still more to come.  A small slew of solar farms have either been approved or are under construction.  A 56 megawatt facility is under way in Moree and 102 megawatts in Nyngan.  Several additional small solar farms have been approved for the ACT and a 53 megawatt solar farm will be built in Broken Hill.  Altogether they could take the portion of Australian PV in utility scale projects to over 6%.  It certainly looks as though solar farms are finally taking off and soon the sun soaked Australian landscape will be dotted with fields of solar panels, not at all gently waving in the wind, supplying the nation with clean electrical power for air-conditioners, hair dryers, shock therapy, and cattle prod Tuesday down at Fight Club.  This is a reasonable conclusion to arrive at, but it is wrong.

The solar farms under way now only got the nod because the current Coalition government could not stop them.  For some reason they do not seem to like renewable energy.  I mean, they really don’t like it.  My impression is if it would not cost them votes they would be quite happy to agitate their caucus until it was raucous and then rush out and smash solar panels with hammers, jump up and down on the pieces, and then train dingos to eat the fragments until their kidneys’ turned fetid.  Now pieces of silicon don’t have that effect on dingo kidneys, but my opinion of the acumen of Australia’s current leadership is so low I doubt they have even a cursory understanding of dingo urology.

It seems very likely the Coalition will prevent any further expansion after the current projects are completed.  However, an unpleasant political climate is not the main difficulty that utility scale solar faces  in Australia.  After all, government policies change.  And they will do so either when we get a new government or perhaps close to the next election when the stinkiest part of a rotting fish is likely to get the chop.  But that won’t be enough to save utility scale solar in Australia. Economics also gets in the way because utility scale solar simply cannot compete with rooftop solar.

The cost of utility scale solar in Australia is quite frankly embarrassing.  Not one single project so far will be able to install PV at a lower cost per watt than local installers putting it on  rooftops.  That is the opposite of how it is supposed to work.  To pay more for solar off in a field somewhere when it is cheaper to put it on your roof is, quite frankly, nuts.  With experience the cost of building utility scale solar should come down, but it is not going to have the opportunity to get that experience, as its situation is only growing worse.

The main reason why rooftop solar knocks its competitor into a cocked hat is that solar farms compete with the wholesale price of electricity during the day which averages about five cents a kilowatt-hour, while rooftop solar competes with the retail price of electricity which averages about 28 cents a kilowatt-hour.  In other words the price that rooftop solar has to beat to turn a profit is more than five times higher than for solar farms.  Currently utility scale solar is not competitive in Australia without the government picking up a large portion of the costs while rooftop solar is extremely competitive and is the cheapest source of electricity available to Australian households and small businesses.  And that leaves utility scale solar lying in a hat full of cocks where it will no doubt be clawed and pecked and terrified by their horrible red wattles.

What really dooms utility scale solar in Australia and ensures we’ll never be likely to have more than a paltry amount is that rooftop solar pushes pushes down the wholesale price of electricity during the day.  People are installing PV panels on their roofs right now because it saves them money and the more that is installed the lower wholesale electricity prices go.  And since the retail price of electricity is so high in Australia and the cost of installing solar so low, people have an incentive to continue to install it even if wholesale prices during the day drop to zero and people get no feed-in tariff for the electricity they export to the grid. This means the natural end point is people keep installing rooftop solar until it ends up supplying the majority of Australia’s daytime electricity use.  So if you listen very carefully you may be able to hear solar photons banging into silicon in the panels on your roof.  Those utterly imaginary tapping noises are the sounds of trillions of little nails being hammered into utility solar’s coffin.

Rooftop solar is already clearly pushing down wholesale electricity prices across the nation.  In South Australia where we get about 6% of our total electricity use from rooftop solar it is not uncommon for it to supply over a third of total demand around noon on weekends.  It did not take us long to get to this point and our capacity is continuing to expand.  If you want to see an example of what wholesale electricity prices may be like across the country in the future take a look at this 30 Minute Demand and Price Graphs for South Australia that I nabbed from the Australian Electricity Market Operator at 12:43 on Saturday:


The lower graph has the price in “dollars per megawatt-hour”, or if you prefer, “tenths of a cent per kilowatt-hour”, the upper graph has demand for grid electricity in megawatts.

Time runs along the bottom of the graphs but since South Australia is in a different time zone from the rest of the National Electricity Market and it is a 30 minute graph you have to add half an hour or so.  As you can see if you have good eyes and roll your mouse over the lower graph, the wholesale price of electricity was slightly negative for  a considerable period during the day.  When I checked at 12:15 Adelaide time the price was negative three tenths of a cent per kilowatt-hour.

Now I want to be clear that rooftop solar was not the only renewable energy source pushing down electricity prices at this time.  South Australia’s wind power also played a significant part.  Rooftop solar would have only been supplying about 28% of total electricity use at around noon, but it is easy to see how further expansion of rooftop capacity would result in a similar effect on wholesale prices.  After all, it’s not as if we are going to run out of roof space.  Technically it would be possible for Australia to generate its entire electricity demand from rooftops alone.

And I would also like to be clear about the reason why the price was negative.  Because the state’s Northern Power Station is not a modern flexible coal plant it doesn’t like shutting down if it doesn’t have to.  After all, it is not as if they have to pay a carbon price on the coal they burn.  As a result it was still operating at about 60% of capacity and the surplus electricity it provided  pushed prices into negative territory.  Also, because of fossil fuel unreliability, a unit of the gas Torrens Power Station was probably also running at about a quarter of capacity to provide spinning reserve in case  a mechanical failure caused the coal plant to clunk out.

And to be even clearer, so clear that I become almost wholly transparent like a jellyfish, I would like to say that I think utility scale solar in the right place is wonderful.  After all, it is far superior to burning fossil fuels on account of how it doesn’t kill people and it has a bright future ahead of it in many countries that are not Australia.  Its future just isn’t as bright as some people think, for while Australia is an outlier, other countries are not immune to having point of use solar out compete it.  And even if there was some sort of miracle in Australia and retail electricity prices dropped to 15 cents a kilowatt-hour, utility scale solar would still not be out of danger.  For while the incentive to install rooftop solar would be greatly reduced, it would still be there, and its installation cost would continue to fall.  But currently in Australia it is clear we would see far better economic return and reduce greenhouse gas emissions at a much lower cost if we invested in rooftop solar rather than utility scale solar.

About Ronald Brakels

Joining SolarQuotes in 2015, Ronald has a knack for reading those tediously long documents put out by solar manufacturers and translating their contents into something consumers might find interesting. Master of heavily researched deep-dive blog posts, his relentless consumer advocacy has ruffled more than a few manufacturer's feathers over the years. Read Ronald's full bio.


  1. roncher bridges says

    Great article !!!!

  2. Ronald Brakels says

    Thank you!

  3. A medium-large coal plant produces 1200 MW. So one of those huge fields of solar panels producing 56 MW (as given above) is just 4.7% of a single coal fired plant!

    • Ronald Brakels says

      The 1,600 megawatt brown coal Hazelwood Power Station and associated mine in Victoria covers 3,554 hectares. In a sunny location in Australia 20% efficient solar panels covering that area would produce more kilowatt-hours than Hazelwood’s average output. Rooftop solar, which is what we mostly have in Australia and mostly will have, removes no land from its original use. Per kilowatt-hour generated Australian coal power removes much more land from agricultural and other uses than solar does.

      • Well, now this is interesting. You’re right – but by an amazing coincidence, not by much. If we assume we can perfectly plaster (no tilt, no spacing) the entire Hazelwood site with SunPower’s record-breaking 345 W panels, my calculations show we could produce 3.47 GWh/year/hectare on the site. Compared to… ready for this? 3.40 GWh/year/hectare that Hazelwood puts out now.

        Now, it’s a bit unfair to pick a coal-fired plant that’s notoriously inefficient. Loy Yang B manages a whopping 10.2 GWh/year/hectare! That’s including the adjacent 700 hectare mine – a bit unfair since that mine also feeds Loy Yang A.

        If we look for one more example – the solar thermal plant in the Mojave Desert in California ( ), I get 0.864 GWh/year/hectare.

        Calculations and references available (gregory.j.bell at gmail) or maybe Solar Quotes will invite me to write a guest article 🙂

        I’m not a solar hater, I just like doing looking at the facts. Obviously solar can be area inefficient and still a great idea for other reasons, like not making our planet into an Easy-Bake oven.

        • Finn Peacock says

          Hi Gregory,

          Thanks for the comment.

          You are more than welcome to pen a guest article. You can contact me at



        • Stuart Brown says

          Why doesn’t anybody get it Australia is 75% desert, plenty of space, so space isn’t a problem. Repeat 1,000 kms by 1,000 kms, is a trillion square meters, 1 kW per meter, times 4,000 hours a year is 4 qintillion Watt hours. If apart from 19 years of Regan/Thatchernomics deinvestment where the price of solar only halved. The price historically halves every 2.5 years, in a decade the price will be 1/16th, 1/2×1/2×1/2×1/2=1/16th. So if your saying that it needs to be 1/5 the price then in 6 years it will be 1/5 the price. In the meantime, rooftop solar, is likely to add storage in 4 years, so people can go off grid and skip the supply charge. In 6 years, people will start buying carbon fiber frame, aluminum body electric cars, like the $17,000 Chinese BMW. Charge at home, from cheap manganese batteries, induction charging when out.

          Industry will want trunk line power, which must be more efficient, than distributed residential. Because it will become cheaper, industry demands will increase. By year 10, it’s 3 times cheaper than carbon, now, that’s what industrial revolutions are about, plentiful cheap energy. At 2,500 times Australia’s energy needs, we could export a trillion tons of liquid hydrogen a year, using only 25% of our desert. Globally 25% of deserts could give us 25 times current world energy usage, how do you make solar panels, you use energy, what do solar panels produce energy. Can you see a self reinforcing pattern here, a punctuation in the equilibrium, a rapid bootstrapping point, long before year 6.

          Where people begin to invest their superannuation money, reverse mortgage, for retirement money.

          • Stuart,

            Ronald’s comment about area, and my subsequent dead-horse-beating of it was mostly an academic exercise out of curiosity. I like applying basic arithmetic and unit conversions to energy issues to see where the truth lies.

            Your optimism about the future is refreshing, but I believe you’re making one huge assumption, and possibly making one error.

            1) Much of the high tech you see around you now, and hope to see soon, is made possible by an advanced civilisation that runs on cheap, fungible and plentiful energy. A role played mostly by oil.

            2) All energy is not created equally. Most alternative energy technologies supply electricity – incredibly useful for things like heaters, lights, electronics and motors, but not so useful for transportation. Storing electricity relies on similar high tech oil-driven technologies as PV panels, as well as oil for all that plastic and those industrial processes that provide the fancy materials.

            Specifically, you’ll find those solar panels you’re relying on in your future scenario require an extensive supply chain that is, now, entirely oil and coal driven. Could it be solar powered? Certainly parts of it, but the diffuse nature of solar makes the use of it for high temperature industrial processes, and mining, costly in a way we’ve never had to deal with before.

            Converting electricity to liquid hydrogen is very problematic and, thanks to the Second Law of Thermodynamics, is lossy. Not to say these things can’t be overcome, but I’m not optimistic about it.

            I highly recommend Tom Murphy’s “Do the Math” blog. He’s a physicist who.. you guessed it… does the math behind a lot of our problems (some would say predicaments) and the hoped-for solutions. Read as much of his back catalogue as you can. “The Energy Trap” and “Sustainability Means Buntky to Me” are good starting points.

  4. EmpowerRepower says

    I still believe that Australia needs some large-scale solar with storage if we are going to replace coal-fired power stations. Being able to store the solar energy for use after the sun goes down is important, and can be achieved by having some concentrated solar thermal plants, as outlined in the excellent Beyond Zero Emissions energy plan for moving Australia towards 100% renewable energy:

    Alternatively, rooftop solar may start achieving the same thing once battery storage becomes more affordable. The level of future wind farm development in Australia will of course have an effect as well, but the federal government seems to have it in for wind farms, which are apparently “utterly offensive.” Ruining an entire clean energy industry with coal-centric changes to government policy seems totally fine with them.

    • Ronald Brakels says

      EmpowerRepower, I’m sure our current government hates solar thermal just as much as they hate wind power. Or at least they would if someone explained to them what just it was using small words. I think it is very likely that by the time we get a government that is serious about reducing greenhouse gas emissions, home and business energy storage will have really taken off in Australia.

  5. Richard Griffiths. says

    Reference you comment – “Also, because of fossil fuel unreliability, a unit of the gas Torrens Power Station was probably also running at about a quarter of capacity to provide spinning reserve in case a mechanical failure caused the coal plant to clunk out.”

    I think you’ll actually find that a major factor in the need for a background gas turbine generator capacity is the unreliability that renewables (sun & wind) feed into the grid, such as if a cloudbank rolls in to cover the sun or the wind drops out.

    Renewables output is intrinsically unstable and inimical to stable base load supply. Why oh why we don’t smooth out the fluctuating renewables curve, with the use of pump storage schemes here in Aus is quite beyond me. It would allow better usage of thermal power output too by harnessing now wasted cheap night-time off peak power generated by day to pump water uphill and release it to power hydro turbines at peak demand. You get transmission and pump inefficiency losses but they aren’t insurmountable.

    • Ronald Brakels says

      Richard, a grid needs enough spinning reserve to cover for the largest generator suddenly going offline so the grid won’t suffer blackouts. And with the current South Australian grid demand of 1,230 megawatts and the high price of natural gas I’m guessing that is the brown coal Northern Power Station with a capacity of 520 megawatts. None of our wind turbines come anywhere close to being that large. The biggest are three megawatts. Trivial next to the 520 megawatts of the Northern Power Station or the 1,280 megawatts of the gas Torrens Island Power Station when it has all its units operating at full capacity. Now it is possible for wind turbines to fail but that’s only going to cause the loss of 3 megawatts at most. The largest solar power system in South Australia that I’m aware of is 1 megawatt. Because wind and solar generating capacity come in such small units there has been no need to increase spinning reserve capacity in South Australia as we’ve gone from nothing to generating electricity from wind and solar equal to about 40% of our total consumption.


  1. Coal vs Solar - Which is more space efficient? - Solar Quotes Blog says:

    […] This is a guest post by Greg Bell – which was conceived in the comments section of another post on this blog. […]

Speak Your Mind

Please keep the SolarQuotes blog constructive and useful with these 5 rules:

1. Real names are preferred - you should be happy to put your name to your comments.
2. Put down your weapons.
3. Assume positive intention.
4. If you are in the solar industry - try to get to the truth, not the sale.
5. Please stay on topic.

Please solve: 30 + 7 

Get The SolarQuotes Weekly Newsletter