(The author is a Reuters market analyst. The views expressed are his own.)
By Gerard Wynn
LONDON Feb 22 (Reuters) - Roof-top solar power is increasingly cost-competitive with retail power prices, with far-reaching implications for solar manufacturers, utilities and rival generation technologies.
Data gathered from U.S. installations by the Department of Energy suggests it is cheaper to generate electricity from roof-top solar panels than to purchase power from electric utilities, if applied to European retail power prices.
The economics of unsubsidised solar depends on the balance of self-generated solar power which is used at home, displacing more expensive purchased electricity, compared with the surplus which has to be exported back to the grid at much lower wholesale prices.
Retail power prices are higher than wholesale because of a mark-up by utilities, plus state levies and charges to cover the cost of grid transmission and renewable energy.
A report titled "The unsubsidised solar revolution" by UBS analysts last month estimated that households in southern Germany installing unsubsidised solar power could already make a net saving over the 20-year lifetime of the panels.
The analysts estimated a positive rate of return on investment of 2 percent already, rising to more than 6 percent by 2020.
The economics of solar will continue to improve as the installed cost continues to fall, retail power prices rise and residential battery storage becomes increasingly competitive, allowing households to displace more purchased electricity.
FALLING INSTALLED COST
Prices of solar panels, or modules, have more than halved in the past three years, because of a global glut after manufacturing ramped up in China.
The remaining installation costs, chiefly labour, are often referred to as "balance of system" and vary according to the maturity of the supply chain.
The U.S. Department of Energy's National Renewable Energy Laboratory (NREL) has developed an open project database detailing the combined full installation cost, excluding incentives, of projects based in the United States.
The NREL database can be found here:
Utilities, installers and the public volunteer the data, which NREL monitors to ensure quality.
"Data validation occurs on each record in the database on a regular basis. The database is continually analysed for corrupt records, bad or invalid data, and outliers such as an abnormal cost to watt ratio. Records found to contain questionable data are flagged and are dealt with on a case by case basis."
As expected, full installed costs have fallen less precipitously than modules, given the labour component.
Median calculations are more meaningful than averages given the non-symmetrical data which includes a minority of utility-scale projects.
The NREL data show median, full installed costs fell 17 percent between 2010 and 2012, and are now around $4 (3 euros) per watt.
That is higher than some analyst estimates.
For example, UBS last month assumed current full installed costs at 1.9 euros ($2.5) per watt, perhaps reflecting a more developed supply chain and lower costs in parts of Europe and especially Germany, compared with the United States.
According to the NREL data, costs fell to a median $3.6 per watt in 2013 to date (sample size of just 7 records), from $4.9 in the last three months (99 records); $5.5 in calendar year 2012 (9,747 records); $6.3 in 2011 (31,388 records); and $6.6 in 2010 (35,906 records).
Regarding size, projects are the equivalent of large residential roof-top installations, with a median size of 4 kilowatts in the 2013 year to date; 7 kW in the last three months; 5.2 kW in 2012; 5.4 kW in 2011; and 5.5 kW in 2010.
Assuming the full cost of a new roof-top installation is $4 per watt, it is straightforward to calculate a levelised cost of electricity (LCOE) using various assumptions.
An LCOE measure estimates cost per unit of power output, in dollars per kilowatt hour, and can be compared with rival energy technologies and actual retail and wholesale power prices.
In the LCOE calculation, lifetime cost includes the initial investment plus financing and operating costs, after accounting for a certain discount rate and tax savings on depreciation and interest.
Lifetime power output is calculated according to the initial output of the installation, depending on local annual sunlight, discounted according to the cost of capital and an estimated annual degradation of the solar modules.
A simpler model can ignore the benefit of tax savings and assume the present value of future loan payments equals the actual initial investment.
Such an approach generates an LCOE of $0.24 per kilowatt hour (kWh), or 0.18 euros.
Assumptions include: $4 per watt installed cost; about 2,500 hours sunlight annually for a 20 percent capacity factor; cost of capital of 8 percent; annual degradation of solar module output of 0.5 percent; a lifetime of 20 years; and an operating cost of 2 cents per kWh.
In a useful reality check, NREL has its own simplified calculator which calculates the LCOE for a comparable installation at $0.25 per kWh.
The link to the NREL viewer can be found here:
Chart 1: www.energy.eu/
Chart 2: goo.gl/ve3Ob
In the European Union, an LCOE of 0.18 euros per kWh is below the retail power price in 11 European countries. (Chart 1)
Countries exceeding 2,500 hours sunlight annually include southern Germany (retail power price 0.26 euros); Cyprus (0.23 euros); Italy (0.22 euros); Spain (0.19 euros); and Portugal (0.19 euros). (Chart 2)
Many of these countries face severe capital constraints and austerity in the near term, which will limit household outlays.
The economics of unsubsidised solar are still only at break-even, but the evidence suggests that utilities would do well to prepare for a new wave of unsubsidised projects.
They have suffered enormously in Germany from the impact of subsidised installations, which have driven down wholesale power prices and fossil fuel power plant load factors. ($1 = 0.7563 euros) (Editing by Anthony Barker)