Deserts vs Roofs, Round 2

Following Scientific American’s foray into the arena of grand, green projects, which I looked at a couple of days ago, I see that, in New Scientist, Fred Pearce has been looking at the Desertec plan to bring solar electricity from the Sahara to Europe.

Fred’s article compares Desertec with the feed-in tariff fueled PV roofing of Germany.  My immediate observation is that it’s hardly a case of either one or the other.  The Desertec plan, which is backed by “20 major German corporations” aims to raise €400 billion to meet (only) “15% of Europe’s electricity needs by 2050”.  And unless I was asleep when I read the relevant chapter in David MacKay’s book, PV panels on roofs alone can’t provide all our electricity. The quote from Germany’s rooftop PV champion is therefore rather puzzling:

“Some of the opposition to Desertec comes from an unexpected quarter. Hermann Scheer, the German member of parliament who masterminded the programme that has put solar panels on 100,000 of the country’s roofs, declared Desertec to be an unnecessary and expensive distraction that would divert investment from projects in Germany itself. Europe could generate as much solar energy as it needs domestically with more rooftop PV panels, Scheer says.”

Maybe he’s joking.

Pearce rather confusingly compares PV in Germany with solar thermal in the Sahara. That’s one two many dimensions for me. The argument is confused by the need, for example, to use water for the solar thermal. If we do a quick thought-experiment comparing PV in Germany and the Sahara the economics becomes obvious. PV in the Sahara would generate at least twice as much electricity per area of PV panel on German roofs, because of the lack of cloud in the Sahara, more overhead sun (reducing losses in the atmosphere) and lack of any tracking ability whatsoever and much greater possibility of being shaded at the ends of the day of fixed panels on roofs (I could say 3 times as much electricity – though I would then have to check my facts more thoroughly, since 3x might be per area of land, which is a slightly different matter – but don’t need to). So if we give a €400bn budget to each location we would have to spend only €200bn on panels in the Sahara.

But we have to allow for transmission losses – let’s say 20% (the article gives 10% for the high-voltage direct current (HVDC) links from Africa to Europe but let’s be pessimistic). So for the equivalent of €400bn of rooftop PV we need to spend €240bn in the Sahara.

HVDC is a proven technology and the article says we need 20 links from Africa to Europe. Let’s be pessimistic again and price these at €5bn a pop [Postscript: I’ve now found that David MacKay gives £1bn for 2000km of HVDC + £1bn for the cost of the land it uses (p.216), so my guesstimate implies he’s a hopeless optimist!]. That’s $340bn we have to spend on our Sahara project.

And maybe another wodge – say €10 billion (I’m basing this on a rather generous €1000 per rooftop worth, estimating rough equivalence with Desertec – and market saturation of suitably angled roofs – at 10 million chunky 1kW average output rooftop systems, 100 times as many as at present) – to enable tracking of the sun to ensure we achieve at least double the output of the German rooftop PVs. I’ve managed to get us up to €350bn.

OK, let’s add on 10% extra capacity to allow for the accidental disruption of supply, political risk, bribes, earthquakes cutting the cables, outages caused by sandstorms and so on. That gets us to €385bn.

I’m really trying to push up the cost here. I’m not even making an allowance for ease of installation and maintenance, the ability to standardise and optimise panel dimensions, the lower labour costs and so on, of arrays in the desert compared to on German roofs.

Even being really pessimistic, we’re €15bn up on the deal. I’ll take that. Plus we have the option of expanding the project further, exploiting economies of scale, after 2050 – whereas in comparison we’re running out of roofs in Germany.

The reason why the Desertec guys have gone for CSP, of course, is it’s even cheaper than PV [something like 8 times cheaper on an average power delivered basis according to David MacKay, p.216]. As Fred himself says: “The clincher is cost. Building a power-station-scale solar thermal installation costs only a fraction of PV generators with the same output.”  And it comes with integral energy storage (read the article) whereas the German rooftop panels will generate most energy when it’s not needed (OK they can sell it to parts of Europe where they have a daily air-conditioning peak, but Desertec can do that too).

Look, feed-in tariffs to roof-owning Germans was a way of funding the early development of PV technology. It’s just not a sensible financial mechanism for rolling out PV. And German rooftops are not a sensible place to put the things.

CSP in the desert wins hands-down over PV on German roofs. If there really is a choice between the two, I don’t know why we’re even discussing cost.

But Fred’s article raises some other strange objections to Desertec:

• “…it could make Europe’s energy supply a hostage to politically unstable countries” say critics.  Nein, nein, nein!  It’s trade.  And trade cements peaceful international relationships.  It’s when trade breaks down – like in the 1930s – that wars start.  Compare the relationship of the West over the last 30 years with, I don’t know, Afghanistan (little trade) and Saudi Arabia; or North Korea and South Korea.  I find it rather strange that a global problem – climate change – is spawning a retreat to local solutions.  Maybe it’s some kind of innate human response – akin to cowering in Mother’s skirts – when bad things happen.  We need collaborative, mutually beneficial solutions if we’re going to crack this one.  Any doubters should just pretend we’re facing some other kind of common threat – a 1km asteroid hurtling towards us at 40,000kph, say.  We’d need to work together on that, wouldn’t we?  Same difference with global warming. In any case, building Desertec diversifies Europe’s energy supplies compared to where they are now, reducing reliance on any one regime. The theoretical problem of supplier-power will be lessened for at least 30 years, until we switch off the gas, by which time we might have come up with other energy sources. Even in the worst case, it’s difficult to imagine reliance on renewables being worse than reliance on dwindling fossil-fuel supplies. So why don’t we just get on with it?

• “Europe should not be exploiting Africa in this way”.  What?  Make your mind up.  How is it exploitation?  You help us make some electricity, we give you some euros and you can buy stuff.  What exactly is the problem with that? This “argument” is hand-wringing liberalism gone mad.

• “But capital cost is not the end of the story. While a solar thermal power plant requires a round-the-clock crew, PV installations pretty much run themselves.”  I seriously doubt this.  Think about it.  You’ll have a few maintenance teams in the desert each looking after a vast amount of output.  Compare that with the army of bureaucrats alone needed to manage millions of rooftop PVs and the owners’ accounts, sort out disputes, investigate fraud (what if they just sell normal electricity back to the grid and put a piece of cheap plastic on the roof?) und so weiter, together with the thousands of small businesses fixing the things when – like any domestic appliance – something goes wrong.  Not to mention the army needed to get up on people’s roofs every now and then to clean the birdshit off the things.  Someone’s in fantasy-land, and it’s not me.

• “What’s more, PV power plants can grow piecemeal: they can start generating power for the grid from the day the first panel is installed, while solar thermal mirrors are useless until the entire power station is completed.”  So?

I’m sold on Desertec. How can I invest?