October 27th, 2012

From a report (pdf, page 5) on the German solar PV market by a government trade promotion agency:

Sorry for the lousy screengrab of a brilliant original. The rising black line is the retail price of electricity; the falling green line is the installed cost per watt for systems up to 100kw.

What it shows is a discontinuity in history, an event of critical importance, [rant] a beam of hope through the black clouds of selfishness, lies and despair threatening the survival of our civilisation in a very Oilocaust [/rant]: the arrival of residential solar grid parity in one large, rich, and northerly country.

The German feed-in tariff (FIT) for solar electricity is already below the average retail rate of 25€c per kwh: 21.4€c deducting the cross-subsidy of 3.6€c for renewable energy, or 4.7$c. This high price by American standards does include a component for greater reliability and amenity; you rarely see an overhead cable in a German town.

The FIT is still well above the wholesale rate, but so what. The datum is crucial confirmation of the insiders’ prediction I trumpeted here that solar grid parity will now rapidly spread through most of the inhabited world, driven by technology and economies of scale alone, more or less regardless of policy (as long as it isn’t actively obstructive).

It will happen soon in the USA. How soon?
Let’s try to translate the German chart to the USA. We need to make two adjustments.

The retail price of electricity is lower and more variable than in Germany. The range for the lower 49 states is between 7c and 18c per kwh (2011); the national average is 10.2 $c. Curiously, the NE-SW gradient of electricity prices is more or less a mirror image of the insolation map, so grid parity may arrive more or less simultaneously in a good many states. Let’s take Omaha, the metropolis nearest to the geographical centre of the lower 49 states, as a conservative representative, with a retail electricity price around 10$c per kwh.

Germany is much more northerly than the USA. Berlin gets no more sun than Alaska. A world insolation map:

Munich has an insolation of 1088 W/hr/yr. Only a few areas round the Great Lakes and the Pacific NW have as little. Albuquerque has 1814 or 66% more. Taking Omaha again as our representative, it has 1453, or 33% more.

Let’s try to adjust the German graph for Omaha conditions, dropping the FIT information. First, we relabel the units from euros to dollars.This makes no difference to the relationship between the left and right y-axes. Next, we should raise the left axis numbers by one-third because of the greater insolation. Eyeballing, an Omaha LCOE of 10c/kwh – look at 13c on the chart – requires an installed system cost of $1.5 (€1.2) per watt; in Albuquerque, $1.8 (€1.4). The current German residential installation cost (new index!) is €1.82, or $2.35.

The problem is of course that real US installation costs are much higher than German ones – at least twice – through market fragmentation, complex and unpredictable subsidies, and Byzantine permitting. Surprisingly, prices in Britain and India are quite close to German levels. According to solar entrepreneur Jigar Shah, writing in Clean Technica :

Today, the average large commercial solar installation in the UK is installed and sold to investors for less than $2/Wdc – same as Germany. In India, where the basic building block is a 5-MW utility scale project, the systems are installed in less than two months for about $1.70/Wdc.

This British commercial source gives £2.5/watt (€3.0, $4.0) for a 4kw house rooftop system in the UK, higher than Shah’s number for utility-scale projects and the German index, but well below US prices.

You really can’t explain low British and Indian costs by appealing by analogy to Germany’s market size, stable regulatory environment, and superior technical education.

This is old satire not data, but it’s fair to say that American workers are probably more skilled than British and less than German.

Britain does share two things with Germany: no permitting for residential rooftop panels, and a (shambolic) FIT rather than quotas. Lowering BOS costs in the USA to British levels should not prove too difficult, and there are strong market and political forces pushing in that direction. On this basis, I predict residential grid parity for Omaha in 2015 (without subsidies but with feed-in priority), just a year later than the untweaked model; and correspondingly earlier anywhere sunnier.

Jagar Shah suggests that the reason for high US costs is over-generous subsidies. They certainly look badly designed. Take Germany as the benchmark. It has no industrial policy to pick winners in the solar business (I reluctantly side with Romney on this one, though he’s in bad faith); it funds world-class applied research – like the DoE – and has a declining national FIT, embedded in federal law. This is reasonably well designed to keep demand high and stable enough to keep installations on the learning curve. That’s it. In contrast, the USA has a maze of federal, state and municipal grants, quotas, loan guarantees and tax credits. It takes expert middlemen to navigate all this, and they extract substantial rent for the service.

Consider the impact on the cost of capital, a crucial input to the LCOE. The chart uses a German figure of 6%, well below the 10% typical in the USA. But it’s quite realistic for a very low-risk investment, even high. That’s the rate for an unsecured personal loan in Germany; mortgages are around 3.1%. Secured home improvement loans should be in between. If a household has savings, it can get much less for an investment of equivalent security: a 10-year Pfandbrief (a secured debenture) yields only 2% wholesale. The panels are less liquid; the only way to cash them in early is equity withdrawal. Taking all this into account, 4-5% would be a reasonable range for for a household’s cost of capital. I don’t see why this doesn’t also apply to US and British households.

The FIT only applies to the electricity actually supplied to the grid by the panel-owner (EEG Art. 16.1). The opportunity cost of the self-consumed electricity is the full retail rate of 25€c and rising. The investment calculation is thus based partly on the FIT rate, partly on the full retail rate, making solar even more attractive. Abandon both FIT and feed-in priority, and at some price it still pays households to invest.

In contrast, the US reliance on tax credits makes the financing of solar panels and wind farms depend on matching projects to rich individuals with tax liabilities to offset. Their alternative investments will be higher yielding than those of typical households, and the difference is driven up further by the bloated margins of financial intermediaries.

A solar FIT has three enormous benefits which German experience shows outweigh any theoretical advantages of tax expenditures and traded quotas, and its drawbacks as touted by the IEA without evidence.

  • FITs create an even playing field for everybody. It’s not flat but tilted: all players know the FIT will drop in future (by an uncertain amount), largely removing the financial incentive to wait and free-ride on the price drops due to the investments of early adopters.
  • As a subsidy to electricity produced, not to installations, an FIT maintains at full strength the normal market incentives to offer and seek out the best deals and the most cost-effective technology.
  • FITs tap the low real and opportunity cost of capital of households far removed from Frankfurt, the City or Wall Street. By creating a wholly new low-risk and low-overhead form of investment, they make the national capital market work better.

Germany provides a fascinating window into the future of energy politics. Grid parity isn’t Utopia, but it changes the game. I’ll come back to this.

If the German grid parity chart is only the second most important chart in the world, What’s the first one?
This :

The familiar Mauna Loa data are from here. I added the exponential trend line fitted by OpenOffice, and the three reference levels. The Copenhagen 2°C/450ppm limit is a political not a scientific one. There’s no reason to think it would be safe in any normal sense. James Hansen, basing his opinion on actual science, describes the target as “a prescription for long-term disaster.” What we’re currently heading for is a 4°C future, which, says Kevin Anderson, another leading climate scientist:

is incompatible with an organized global community, is likely to be beyond ‘adaptation’, is devastating to the majority of ecosystems, and has a high probability of not being stable.

To avoid this, we require not incremental technological progress, but a highly disruptive innovation. Solar PV is the only one on the table. Well, there’s always this old standby. That’s the Pentagon talking, not me.

Post a Comment

Logged in as James Wimberley. Log out »