Monday, 4 June 2012
Reconsidering the economics of solar power
It's been suggested in comments that I have a good look at a technical paper from Bloomberg called "Reconsidering the economics of solar power".
So I did.
For starters, just note that most of the co-authors of this paper are employed by companies or organisations that have a vested interest in pushing solar power. I'm not saying that's bad - I'm just saying, is all. I'm happy to read research from Big Oil and Big Coal, so why not read what Big Solar has to say as well?
The paper itself is quite interesting, but the best bits are in the footnotes (which no one ever reads - they're great places to bury information that you need to disclose, but you don't want anyone to actually take note of).
For starters, the price of solar panels has fallen rapidly over the last few years. This is partly due to the growing scale of manufacturing, partly due to plumetting cost of silicon feedstock (from $450/kg in 2008 to $27/kg in 2012), partly due to improvements in manufacturing techniques and technologies and partly due to a flatlining of demand due to countries like Spain going broke. Supply greatly outstrips demand. Supply is over 50GW per year and demand is only 26-35GW. That great crashing sound you hear is solar manufacturers going broke. This is acknowledged in the very last paragraph of the paper - a great place to bury bad news, as very few people ever grind through a report to the very last paragraph.
"Current PV module prices are considered by some to be below manufacturing cost, and consequently, as unsustainable, in large part because several leading non-Chinese firms in the industry have recently announced losses, cutbacks or massive write-downs or filed for bankruptcy"
PV modules did cost up to $4.00/W in 2008. They're now below $1.00/W - at the factory gate. Prices at the factory gate in China have fallen as low as $0.85/W (before tax etc). However, that's because:
Note that sub-$1/W is largely seen as significantly below actual manufacturing costs -- First Solar is an important exception and thus the benchmark -- and is therefore unsustainable for many if not most manufacturers.
And you can see the impact of that price war in this share price chart for First Solar:
If it's unsustainable, that means one of two things. Either manufacturers need to cut their costs further, or enough of them need to go broke to allow prices to bounce back.
However, a module is not the only cost of a solar system. You've also got all the bits and pieces that are required to wire it together and plug it into your power system, and the labour and transport costs and taxes involved in getting it from a Chinese factory to your roof. According to Bloomberg, these costs (which include installation) average $1.85/W across all countries. Bigger, industrial scale systems that aren't roof mounted can be down around $1.43/W. However, there is no breakdown on a country by country basis, so we don't see what the costs are like in a high labour-cost country like Australia versus say Indonesia. Or Malawi. Or Texas, where you can make use of illegal aliens (the Mexican kind, not the probing kind).
As one solar installer noted here:
Our installed PV costs for simple, common asphalt rooftops are running under $4 / watt right now. But Bob - you are right in that the majority of the cost of a solar PV system is now becoming labor / project management costs. There are currently many time saving fasteners and solar equipment advances being used. It just takes a long time to crawl up on a randomly different roof, lay it out, and put in the roof support structure (flashed, water proofing...). These steps take human intelligence, variable problem solving skills, craftsmanship and TIME.
Labour costs aren't coming down soon - unless a recession really starts to bite.
The paper then delves into a lot of mathematical blather about the LCOE (Levelised Cost of Electricity). LCOE is defined as:
The cost of electricity generated by different sources is a calculation of the cost of generating electricity at the point of connection to a load or electricity grid. It includes the initial capital, return on investment, as well as the costs of continuous operation, fuel, and maintenance.
The paper notes that Solar LCOE dropped from $0.32/kWh in 2009 to $0.17 in 2012, but the footnotes disclose that "some LCOE figures from the US quoted in this paper may be post-Federal tax rebates and may also include local capex rebates in some cases." So take those numbers with a pinch of salt.
The paper also promotes a power auction in the US where 9 of the 11 winning bids were solar, with the highest price being 9c/kWh - but the footnotes again state that "federal tax credits likely make these prices look lower than they would otherwise be". So again, be generous with the pinch of salt.
In another paper, "BNEF identify the most important determining factors of the levelized cost as being capital costs, capacity factor, cost of equity, and cost of debt", and "the loan repayment method is one high-impact assumption."
So in other words, you can fiddle (ie, lower) the LCOE by providing capital grants that lower the apparent capital cost, providing low or zero interest loans on very favourable repayment terms that dramatically reduce the cost of debt and making the loan repayment method appear more favourable.
Also, "capital-intensive renewables, such as PV, are more sensitive to electricity prices, risk adjusted interest rates, maintenance costs and insolation levels"
Again, that's financial-speak for saying you can make a PV project look more palatable by removing the sensitivity to electricity prices by setting fixed, very high prices for PV power (like the Labor government did in NSW), and again, supplying soft loans (like Obama did with Solyndra - that worked out well, didn't it?)
The very last footnote in the paper notes that "Early 2012 Japan decided that solar will receive JPY 42/kWh for 20 years". The yen is currently worth about 1.3 cents Australian, which means the Japanese are paying 54.6 cents per kWh for solar power. Gee, that sounds attractive to me. Not.
A lot of spade work appears to be going on with the aim of redefining what "grid parity" means:
"Depending on the scale of the PV project, grid parity normally refers to the LCOE of PV by comparison with alternative means of wholesale electricity provision – often an inappropriate metric as discussed previously. While for large-scale PV, these alternatives may indeed be assessed as alternative wholesale generation projects utilising different technologies, for small-scale domestic or commercial PV systems, the appropriate alternative should be the purchase of electricity at a relevant residential or commercial tariff. The latter case is where grid parity actually took its name – such PV applications are not competing against wholesale generation but, instead, the delivered price of electricity through the grid."
As for reaching grid parity:
"Contrary to the view that the arrival of grid parity is still decades away, numerous studies have concluded that solar PV grid parity has already been achieved in a number of countries/regions"
"Calculations ... suggested that grid parity of wholesale electricity in Germany will occur around 2013-2014."
What isn't mentioned there is that electricity prices are now so high in Germany, the country is de-industrialising, with companies moving to countries with cheaper power. Remember that power from horrible coal and nuclear plants enters the grid at between 4c and 6c per kWh. That's why the solar industry is desperate to change the definition of "grid parity" to the price the consumer pays - which has in turn been inflated by the cost of subsidies for solar, wind, geothermal and biomass projects and our old favourite, the "carbin tax".
A paper cited in the References takes us to the US Energy Information Agency:
The average retail price of electricity in the United States in 2010 was 9.88 cents per kilowatt-hour (kWh). The average prices by type of utility customer were:
Residential: 11.6¢ per kWh
Transportation: 11.0¢ per kWh
Commercial: 10.3¢ per kWh
Industrial: 6.8¢ per kWh
The three states with the highest average price of electricity in 2010 were:
Hawaii (25.12¢ per kWh)
Connecticut (17.39¢ per kWh)
New York (16.31¢ per kWh)
Those with the lowest average prices in 2010 were:
Wyoming (6.20¢ per kWh)
Idaho (6.54¢ per kWh)
Kentucky (6.75¢ per kWh)
On average, electricity prices are highest in Hawaii, mainly because most of the electricity there is generated with fuel oil. Idaho usually has the lowest prices mainly because of the availability of low-cost hydroelectric power from federal dams.
When claims are made about solar being price competitive in certain markets, just remember that they are cherry picking the most expensive markets, like Hawaii. No mention is ever made of places like Wyoming.
A lovely graph is included on page 13 showing projected LCOEs for a range of countries - Australia is shown as having a solar LCOE of around 21c/kWh. But in the footnotes to the graph, it assumes a weighted average cost of capital of 6%. This is laughably low - it should be at least 8-10%, which would have a major impact on the LCOE. You can of course fiddle your WACC by throwing in grants, subsidies and low interest loans.
I grabbed the following graph from a 2005 report (cited in the references) from the International Energy Agency. Don't worry about the costs so much - what it shows is the LCOE assuming a 5% discount rate and a 10% discount rate. See the big differences? What this shows is how you can get any result you want by fiddling with the financial assumptions.
Solar is price competitive in some markets:
"Data from IRENA now indicate that grid-connected PV in Africa has already become competitive with diesel-generated power, with an LCOE between $0.30 and $0.95/kWh, based on size, local diesel subsidies, and pilferage. BNEF concludes that falling costs in PV technology mean that solar power is already a viable option for electricity generation in the Persian Gulf Region, where it can generate good economic returns by replacing the burning of oil for electricity generation"
Yep, if you're paying $0.95/kWh for power in Africa, solar starts to make sense. Except you probably need to keep that diesel generator in order to provide power at night - unless you've also invested in a hideously expensive storage system. Or maybe Africans just don't need power at night.
And of course the footnotes provide this gem about their calculations from the Persian Gulf:
"As long as the unburnt oil is valued at the international selling price, rather than extraction cost."
Given that it costs the Arabs bugger all to extract the oil, I'd say they'll be burning a lot of oil to generate power for a long time to come. Unless of course the Arabs don't need power at night either. Plus I had a look at the paper it cited - they fail to mention that the calculations were done on a laughably small 1MW plant, and that the "cost of lowering CO2 emissions are accounted for" - weaselspeak for cranking up the LCOE of an oil fired plant by adding a hypothetical carbon price.
As for solar being "cheap" in India - it is cheaper than having your own diesel generator. A diesel costs 17 rupees per kWh, whilst solar can do it for 8.78 rupees per kWh. And what about coal fired power? 4 rupees per kWh. However, why do so many Indians have diesel generators? Because they don't have enough coal fired power plants, so their grid is really unreliable. The solution for a poor country like India would clearly be to build a lot of cheap, coal fired generation capacity - but they've been suckered into expensive solar like so many others. "India’s solar industry has benefited from tax breaks and a guaranteed government buyer of its cleaner power."
And solar doesn't make those diesel generators disappear entirely:
Acme Telepower Ltd., a Gurgaon- based company converting sites for Viom Networks Ltd. and Bharti, estimates the panels can cut the diesel running time of a rural tower to eight hours a day from 22.
I had a look at the references too - I bet no one ever bothers doing that. One of the first references is:
Baillie, R., 2011. Solar closes in on grid parity. Renewable Energy World.
I'd never heard of Renewable Energy World, so I had a peek. It turns out the authors were a bit selective in what they chose to quote from Mr Baillie (I'm guilty of the same thing, but at least I'm open about it):
When grid parity is going to be achieved is another tricky question, and varies greatly from place to place. Again according to Ernst & Young, retail grid parity may be reached generally between 2012 and 2015 with, for example, the US to the fore and the U.K. having the prospect of parity in 2015, if retail electricity prices continue to rise.
Get that last bit - solar will become competitive if retail electricity prices continue to rise! A huge proportion of the UK population is currently in a state of fuel poverty, and it's only going to get worse. Cold kills more people in the UK than car crashes - old poor people who can't afford to heat their homes. And this is seen as a good thing? Only the completely demented and fanatical could fail to see the problem here - that freezing pensioners to death is not a good policy outcome.
However, if solar is judged by the harsher test of wholesale price parity, then it is not expected to be achievable until about 2030 in Italy — with concentrating solar power (CSP) achieving parity a few years earlier, between 2025 and 2027 in California and Spain, the company believes.
Unless those countries and California go broke in the meantime, and all those lovely subsidies evaporate. In the comments, it's noted that electricity prices in California are now around 30c/kWh. Ouch. And California is in the grip of a terrible recession and companies and residents are fleeing the state.
Other forecasters offer different estimates that would result in grid parity happening sooner. UN expert Sven Teske, a contributing author to the Intergovernmental Panel on Climate Change (IPCC)’s recent report on renewable energies as well as renewables director at Greenpeace, says the EU is on track for solar grid parity as early as 2017. Teske says that on current trends he expects Spain, Italy, France and Germany to reach grid parity by 2015, but that progress could be endangered by market uncertainty over the future of these nations’ FiTs.
Yes, there is good reason for uncertainty over their FiTs (feed in tarrifs) - going broke generally concentrates the mind wonderfully. And note who he quoted as the expert here - the renewables director at Greenpeace.
And I don't hold out much hope for Germany continuing to pour money into its solar "money pit" - again, from the References, I found this paper by Bjorn Lomburg - Germany's sunshine daydream:
One of the world’s biggest green-energy public-policy experiments is coming to a bitter end in Germany, with important lessons for policymakers elsewhere.
Germany once prided itself on being the “photovoltaic world champion”, doling out generous subsidies – totalling more than $130 billion, according to research from Germany’s Ruhr University – to citizens to invest in solar energy. But now the German government is vowing to cut the subsidies sooner than planned, and to phase out support over the next five years. What went wrong?
There is a fundamental problem with subsidizing inefficient green technology: it is affordable only if it is done in tiny, tokenistic amounts. Using the government’s generous subsidies, Germans installed 7.5 gigawatts of photovoltaic (PV) capacity last year, more than double what the government had deemed “acceptable.” It is estimated that this increase alone will lead to a $260 hike in the average consumer’s annual power bill.
According to Der Spiegel, even members of Chancellor Angela Merkel’s staff are now describing the policy as a massive money pit. Philipp Rösler, Germany’s minister of economics and technology, has called the spiraling solar subsidies a “threat to the economy.”
Unfortunately, Germany – like most of the world – is not as sunny as the Sahara. And, while sunlight is free, panels and installation are not. Solar power is at least four times more costly than energy produced by fossil fuels. It also has the distinct disadvantage of not working at night, when much electricity is consumed.
In the words of the German Association of Physicists, “solar energy cannot replace any additional power plants.” On short, overcast winter days, Germany’s 1.1 million solar-power systems can generate no electricity at all. The country is then forced to import considerable amounts of electricity from nuclear power plants in France and the Czech Republic. When the sun failed to shine last winter, one emergency back-up plan powered up an Austrian oil-fired plant to fill the supply gap.
Oh dear. Read the whole thing.
Anyway, that was fun.
Badly proofread by Boy on a bike at Monday, June 04, 2012