Wednesday, July 06, 2011

Solar at midyear

The last 3 months have been an awful "year" to be invested in solar. The price of solar panels has fallen 25-30% since April. (May was particularly rough.) And the stock prices of many solar panel manufacturers have collapsed (along with their p/e multiples).

The good news is that demand finally appears to be picking up and panel prices are stabilizing in the 1 euro/watt range ($1.40/W). My guess is that 2H 2011 will see a record amount of solar installed in Europe, which is still the key market for solar.

I was too optimistic about the prospects for the solar industry as 2011 started. At the time I believed $1.50/W would be the lowest price for panels in the second half. Mid-tier vendors (mostly driven by China) are already selling panels for $1.35/W or under. It could be that prices rebound by year end if demand picks up fast enough to clear channel inventory, but I think it is more likely prices will hang around the 1 euro/watt range (+/- 10%) for the next ~12 months. Current prices appear low enough to drive many western solar manufacturers out of business--I am especially concerned about thin film manufacturers who were just getting back on their feet after the 2009 price collapse (First Solar the low cost leader excepted). Still I expect even FSLR to experience a significant margin contraction. FSLR's impressive record of cost reduction appears to be stagnating just as the cost reductions of its silicon based competitors is accelerating. It is possible that FSLR has "captured" enough installation pipeline to manage a margin "soft landing" as opposed to a margin crash...but either way I don't see a route for them to increase margins near or mid term.

Based on recent history (early 2009) it seems that project developers require several months to react to a major price adjustment--although many project developers have started to "forward price" large projects. Forward pricing means they project what next year's panel costs will be for a project they intend to install next year based on recent history/experience of pricing. It can be dangerous/expensive if you make a substantial error, but for now market dynamics appear to be benefiting the developers using the forward pricing model.

Part of what prompted me to post today was an article I ran across trumpeting the 17GW and growing US project pipeline from Solarbuzz. In the article they note that the average utility scale installation cost (i.e. installations over 1MW) has reached $4.50/W in the US with a substantial fraction (1/3) of the pipeline booked at sub $4/W pricing. (see the paragraph just under the pie chart).

Since I doubted that solar would hit $4/W installed in 2011 in my end of 2010 posting, this confirms the terrible year solar has experienced (in terms of price declines) in the 1st half of 2011. But it is also worth remembering that with a 30% federal tax credit/grant a $4/W installed cost yields an out of pocket cost of $2.80/watt for the utility not counting any other (i.e. state or tax) incentives. It is important to note that $2.80/watt is not much more than ~$2.50/watt of capacity that a new coal plant costs. The comparison is rough and incomplete, subsidized solar vs. subsidized coal; coal has much higher O&M costs (here coal pollution is equated to a pure subsidy) offset by many more hours of operation per year (3-4x) as solar, but half those coal production hours occur at night when prices (& demand) for power are much lower than during the day.

In the simplest terms $2.80/20yr leads to $0.14/kwh for any location receiving 1000 hours of sun/year which happens to include 90% of the lower 48. Since many solar projects will be installed in the US southwest where 1500 hours of sun/year is common (the very best locations get 1800+ hours of sun/yr), it is clear that solar projects are being planned now that will cost utilities less than $0.10/kwh for the next 20 years of peak (shaving) power. The last I checked $0.10/kwh was a very competitive price for peak power in the US southwest.

The treasury grant is set to expire at the end of the 2011 and is unlikely to be extended given the cost cutting mood of the country, but the 30% tax credit is good until 2016, meaning any project currently being planned is likely to benefit from it.

Granted the US is currently a minor player on the global stage (8-10%), but I think every country will look more closely at the competitive dynamics of solar power going forward. Returning to the unsubsidized $4/W installed price of solar and translating that into euros (conveniently) yields ~2.8 euros/W. Again using simple math 2.8/20 yields a cost of 14 eurocents/kwh in a country like Germany (currently paying 22-25 eurocents/watt for solar power for 20 years) and ~10 eurocents/kwh in Italy (currently paying somewhat more than 25 eurocents/watt for 20 years--but with some nebulous "soft" limit on how much solar Italy will allow to be connected each year). These two markets represented ~12GW of solar panel (maybe a smidgen less) or 75% of the world market in 2010. Given the available returns to investors outlined above, I expect both countries to install at least 12GW in 2011 or 50% of the world market. My guess would be 8GW go into Germany and 4GW go into Italy for the year. Both markets had a slow 1st half, meaning that installs will need to significantly increase starting immediately for this prediction to hold.

Another thing that I expect to see in the near future is very strong growth in solar installed in developing countries. As the cost of solar falls, more and more "applications" will be sold with solar "pre-installed". For example, solar powered street lighting will quickly become the norm in any country (if it is not already) that has not already built out their grid. Three factors will combine (in a virtuous circle) to make this a reality: 1) efficient LED lighting costs will fall as they reach industrial scale manufacturing--recall how the cost of CFLs dropped by 75% as their production ramped up over the past decade, 2) worldwide development of electric cars is leading to major investment in and the development of many new battery technologies--this will lead to better price/performance for all battery technologies over the next decade, 3) anytime the solar, or lighting, or battery technology improves in performance or cost the complete "packaged solution" will improve in performance or cost at an equal or faster rate. Therefore once a pre-installed solution becomes competitive for an application, it will progressively disrupt the previous solution in the market.

As solar becomes more competitive in price, new coal plants will not be built, not because solar is cheaper on a $/kwh basis but because the "applications" that currently drive demand for coal supplied power will simply come with solar built in.