Large investments in centralised power stations will look very sickly as the fossil fuel plants are called on to work fewer and fewer hours each passing year and at lower prices.
The Spanish power company Abengoa recently opened another enormous concentrating solar power (CSP) plant in the south-west of the US.
Covering over 700 hectares of Arizona's Mojave Desert and costing over $1.6bn, the site captures energy using parabolic mirrors that focus the sun's rays onto a thin tube of liquid that rises in temperature to over 600 degrees.
The big advantage of CSP is that this liquid is that it can be used to store energy overnight. In reliably sunny places, the technology can therefore provide 'dispatchable' electricity.
The 280 MW project in the Mojave is intimidatingly expensive at about $6,000 per kilowatt of maximum power. The average output over a year will be about 25% of the peak production, meaning that capital cost is about $24,000 per kilowatt of average output.
That puts it perhaps three times the price of the proposed nuclear plant at Hinkley Point. Of course, a nuclear plant requires fuel and a lot of people to run it. Nevertheless, CSP is still much more expensive than many other sources of power.
CSP will be competitive with fossil fuels in a decade
My rough calculation is that it is delivering electricity at a cost of around 12p/19 cents per kilowatt hour in the American South West or in Spain. PV is much cheaper - but of course photovoltaics cannot provide power at night.
So is there any reason to be optimistic about CSP? I think there are two important points to be made. First, concentrating solar power will see substantial cost reductions as the number of plants grows, particularly in the US.
The US Department of Energy has a carefully reasoned support plan that targets a cost of electricity from CSP of only 6 cents (4p) per kilowatt hour in 2020. Even if this level is not achieved and the figure is 8 or 9 cents, concentrating solar power may roughly competitive with fossil fuels by the middle of the next decade.
In its plan for 2020 the DoE sees substantial reductions in all four elements of the cost of concentrating solar power - the field of mirrors, the fluid-filled tubes on which they focus light, the containers that hold the extremely hot liquids and the electricity turbines that are rotated by the steam generated from these liquids.
These reductions look feasible to me. CSP farms will benefit from scale and learning effects, much as PV has done over the last ten years. By the way, my estimate of 19 cents for the new Mojave plant is not inconsistent with the 13 cents suggested for 2013 by the DoE work. Construction of the Mojave plant began when costs were higher than they would be today.
Capital pools are turning fossil-averse
The second cause of optimism is very different and it is to do with the capital markets. These giant anonymous pools of money are going to be the engine which drives decarbonisation, not governments or international conferences. In its amoral and unreflective way, Big Finance is beginning to really dislike fossil fuels (and probably nuclear as well).
Large investments in centralised power stations will look very sickly as the fossil fuel plants are called on to work fewer and fewer hours each passing year and at lower prices.
Money is now nearly impossible to obtain in many parts of the world for the construction of new fossil fuel power stations. It is not climate change or 'unburnable carbon' worries that are driving this. It is the growing sense among banks and investment managers that the fall in the cost of renewables is unstoppable.
This will leave their large investments in centralised power stations looking very sickly as the fossil fuel plants are called on to work fewer and fewer hours each passing year and at lower prices.
We saw more evidence of this in E.ON's decision to split its financially weak coal and gas power stations into a separate company while putting its future emphasis on renewables and on smart grid infrastructure. Make no mistake, the old company will eventually go bankrupt, or be completely restructured.
Other big generators, such as NRG in the US, have made similar corporate splits that puts coal power stations into separate companies that can be allowed to die without affecting the financial health of the renewables companies.
Just one big technical challenge - cutting water use
Abengoa's new CSP plant in the Mojave is an important part of the Spanish company's so-called 'YieldCo', a self-standing company that can find its own finance cheaply and offer a relatively small but reliable and steadily growing income stream to investors.
This is what Big Finance wants now and CSP plants provide a good asset to hold. These sun-following power stations are extremely reliable, very long lasting and their storage capacity means that they can bid in the electricity markets to provide guaranteed power throughout the day and most of the night.
These are the good things about CSP. There's always a 'but' and in the case of concentrating solar power it is the high level of water use associated with the plants. Abengoa claims that the fresh water absorbed by its Mojave plant is only a fifth of what it would be if its 714 hectares was used for agriculture.
Not good enough, I'm afraid. I'd like to think that nobody in their right mind would even consider agriculture in the desert - but as the photograph (above right) shows, that's not actually the case - it's surrounded by irrigated farmland.
But what clearly is true is that in most places where CSP will work well, water is very scarce and likely to get scarcer in the future. Until different cooling systems are found CSP will not achieve the widespread success we need.
Chris Goodall is an expert on energy, environment and climate change and valued contributor to The Ecologist. He blogs at Carbon Commentary.
This article was originally published on Carbon Commentary.
Author's note for real energy geeks: the new Abengoa plant has an energy intensity of about 9 watts a square metre. MacKay estimated a slightly higher figure for desert CSP of about 15 watts.