The International Energy Agency published a report today (August 31st 2015) that focuses on the rapid decline in the cost of renewable energy. More precisely, it says that electricity costs from wind and solar have plunged, a word rarely used by international civil servants. On good sites around the world, renewables are now cheaper than fossil fuels.

Bizarrely, the IEA says that new nuclear is also inexpensive, a conclusion strikingly at variance with the rampant inflation in construction costs around the world. It may be that the absurd optimism over nuclear is influenced by the joint author of this report, the Nuclear Energy Agency. (The cost estimate of $50 a megawatt hour is one third of what Hinkley C will be paid).

This note looks at how today’s figures compare with the last edition of this report. It'll be no surprise that expected solar PV costs are now little more than a quarter of the figure of just five years ago. We are living through a truly remarkable decline in the costs of PV, driven by the huge increases in the volumes of solar panels being installed.

The 2010 report

In 2010 the IEA said that solar costs ‘could drop 70% from the current $4,000-6,000 per kilowatt down to $1,200-1,800 by 2030’. It targeted reductions of ‘at least 40%’ by 2015 and 50% by 2020. These apparently aggressive assumptions presupposed ‘rapid deployment driven by strong policy action’.

Five years later, the IEA says that solar PV costs in the most competitive country (Germany) are now $1,200 per kilowatt for large-scale installations. In other words, costs have already fallen to the level that the Agency said ‘could’ be achieved in 2030 under very favourable conditions. What the IEA said would take 20 years actually took 5. Solar farms installed in low cost areas are now half the price that the IEA’s 2010 estimates suggested might be possible.

The lower capital costs have fed through to reduced electricity production charges. In a very good location, the 2010 IEA report said it would cost $215 to generate a megawatt hour. (This figure is calculated by working out how much electricity is going to be produced over the life of the panels and spreading the full cost of this installation over this total). This calculation used a cost of capital of 5% a year, which adds to the implicit price of electricity produced.

By 2015, the combination of a lower interest rate and reduced capital costs had cut that the cost of electricity to a low of $54 per megawatt hour in the US, parts of which have some of the best sun in the world. That’s a reduction of very nearly three quarters in five years, or 32% a year compounded. Although German installation costs are lower than in the US, better solar radiation more than makes up for this, leaving the cost per megawatt hour lower in places like Texas and Arizona.

Does the $54 figure correspond to the offers that solar farm owners make to electricity buyers? Yes, in parts of the US recent agreements between solar and utilities have been lower than $60 a megawatt hour, even after adjusting for the subsidies received by the PV industry.

What other technologies have ever achieved this rate of improvement? The early semiconductor industry achieved compounded rates of improvement of at least 35%. The cost of DNA sequencing has fallen by 90% since 2010, a rate equivalent to over 60% improvement a year. But apart from these two outliers virtually no technology has got better faster than solar PV. Importantly, although some experts suggest that semiconductors might now be approaching the limits of improvement, the scope for better PV is nowhere near exploited. The reduction in the costs of generating electricity from solar panels sitting in fields will continue for many more years.

Is PV competitive with fossil fuel technologies yet?

Where does this leave PV in relation to competing ways of generating electric power? The IEA doesn’t make comparisons easy because it uses a high interest rate of 10% in its own charts. Renewable technologies such as PV usually have high installation costs and low running costs whereas fossil fuel plants are cheaper to build but more costly to run. If interest rates are as high as 10%, this penalises those types of generating plant which need more upfront money to build.

At a 10% rate, PV in the best countries produces electricity at around $100 a megawatt hour, even when penalised by high interest. This compares with about $70 for the cheapest gas and just over $80 for new coal plants. This comparison makes solar PV still not quite competitive with fossil fuels.

Look at the numbers using a lower (and more realistic) interest rate and the picture changes markedly. In the chart below, the cost of PV in the US is lower than gas as long as the interest rate used is below about 4%. Is this a reasonable rate to use? Yes; new PV developments are now routinely financed at lower rates than this around the world.

The picture is even clearer in China, where gas for electricity production is much more expensive than in the US. There, PV beats gas at all interest rates. The significance of this probably hasn’t been fully realised.

It’s also striking that the in the five year period in which solar PV costs have fallen dramatically, most of the competing technologies for generating power - gas, coal and nuclear – have seen increases. The minimum cost for electricity from a new coal power station was put at below $40 a megawatt hour in 2010 and is now over $80. The same figures for gas are $45 rising to $70.

Nuclear costs are also assumed to have risen, although the people at the IEA still think it is possible to build a nuclear power station to deliver electricity at around $50 a megawatt hour with a 10% interest rate. Have they spent the last five years on the Philae comet or somewhere equally remote from Planet Earth? For a realistic comparison, the strike price actually agreed for Hinkley C is around $150 a megawatt hour, or three times as much as the IEA hypothesises. Other nuclear power stations currently in construction are similarly priced at multiples of what the IEA says is possible. But, for completeness, one does need to say that the IEA does conclude that nuclear is cheaper than PV at all levels of interest rate. However their data seems remarkably, almost absurdly, divorced from reality.

What about wind? The IEA says that onshore wind has reduced in cost by about 30% since 2010. In the best US locations the figures for wind are now as low as $33 a megawatt hour, down from $48 in 2010 if we use a 3% interest rate. At the moment, wind can be cheaper than PV. But its cost is falling much more slowly than PV. If current trends continue, PV will cut below wind within three years and the difference will then continue to widen.

Or perhaps not. The foolish policy changes of the UK government may be mirrored around the world. It is the sheer volume of PV being installed that is crashing the price of solar. We need this hell-for-leather growth to continue for a few more years, supported where necessary by tax and regulatory support. Although PV is almost certainly cheaper than any other technology in the Middle East, much of the Indian subcontinent, parts of Africa and Latin America, large rich countries need to play their part in keeping global demand for panels surging. If a few more countries act precipitately like the UK, which during the first quarter of this year was probably accounting for 20% of global panel sales but now almost zero, then the rate of PV price decline will inevitably tail off. This is in nobody’s interest (except the fossil fuel companies).