September 18, 2011
Sharp Photovoltaic System Price Decline 2010-2011
Cheaper solar power installs for home or office.
Berkeley, CA — The installed cost of solar photovoltaic (PV) power systems in the United States fell substantially in 2010 and into the first half of 2011, according to the latest edition of an annual PV cost tracking report released by the Department of Energy’s Lawrence Berkeley National Laboratory (Berkeley Lab).
The average installed cost of residential and commercial PV systems completed in 2010 fell by roughly 17 percent from the year before, and by an additional 11 percent within the first six months of 2011. These recent installed cost reductions are attributable, in part, to dramatic reductions in the price of PV modules. Galen Barbose of Berkeley Lab’s Environmental Energy Technologies Division and co-author of the report explains: “Wholesale PV module prices have fallen precipitously since about 2008, and those upstream cost reductions have made their way through to consumers.”
This does not mean that manufacturing costs are falling as fast as prices. For several years in the run up to the 2008 financial crisis demand for solar power was growing so fast due to government incentives (especially in Germany) that declines in production costs did not translate into declines in market prices. The recession caused a drop in demand while capacity was still growing. So prices are catching up with previous production cost declines. Companies are feeling more pricing pressures and a shake-out is going on with weaker players failing or merging.
If you are going to build a home you can get PV installed more cheaply if you design for it and do it during your construction process.
The report also found that residential PV systems installed on new homes had significantly lower average installed costs than those installed as retrofits to existing homes.
The full report (PDF) has some interesting details.
Average installed costs vary widely across states; among ≤10 kW systems completed in
2010, average costs range from a low of $6.3/W in New Hampshire to a high of $8.4/W in
Utah. The country’s largest state PV markets, California and New Jersey, were near the
center of this range, suggesting that, in addition to absolute market size, other state and local factors (e.g., permitting requirements, labor rates, the extent of third party ownership, and sales tax exemptions) also strongly influence installed costs.
Utility-scale solar projects are much cheaper. So people installing solar on their roofs are actually driving up average costs of solar installs.
Since wholesale module costs are now a small fraction of total installed costs it seems like the big cost reductions needed to make solar competitive mostly fall in other areas such as inverters, mounting brackets, permits, labor, and installation.
The recent decline in installed costs is, in large part, attributable to falling wholesale module prices, which fell by $0.9/W from 2008 to 2009, by $0.5/W from 2009 to 2010, and which have fallen further still in 2011 (based on Navigant Consulting’s Global Power Module
The savings when installs are done as part of new home construction are sizable.
The new construction market offers cost advantages for small residential PV systems.
Among 2-3 kW residential systems (the size range typical for residential new construction)
installed in 2010 and funded through California’s incentive programs, new construction
systems cost $0.7/W less, on average, than comparably sized residential retrofit systems (or
$1.5/W less if comparing only rack-mounted systems).
With non-module costs now several times larger than module costs I am left wondering about the prospects for the decline of non-module costs. Will PV total installation costs decline much more slowly than we've seen in the last few years? What is needed to lower the non-module costs? Installer robots? Roof tile as PV? Inverters integrated into the modules? Anyone have insights into where the next big round of cost cutting will come from and when it will happen?
The important number in that report was $6.2/Watt, installed. That was the 2010 average, which was $1.7 below the 2009 average. When you figure that the price of the actual panels is less than $2 per watt, it's clear that the actual panels aren't the driving factor in cost. Any extreme advances is going to have to come from cheaper installs.
I don't see widespread adoption of PV until there is some way to store the power and use it at night and on cloudy days. Right now, the cost of a battery would double the installed cost and a minimal 3kW system would cost you over $35,000. That price is going to have to come down a long way before folks just "go solar".
It's helpful that solar comes "on peak", that is, solar peaks when grid demand peaks, more or less. But you still have the problem of cloudy days, which sometimes come several at a time in the winter.
I just don't see solar as anything more than a peaking resource, it just can't replace a 24x7 power plant, at least not yet. Storage is the key and it's a problem we just don't have a clue about.
Small inverter prices are also dropping quickly (possibly as well as quality) as manufacturing moves to China. A home scale inverter that might have cost $6k in 2009 (US or European made) can be had for just under $1k made in China now.
I think you will see installers lowering margins as the know how to do it spreads. Up through now there were a few oligopolistic installers in each area (who used to also arrange for the panel purchase). Now installing solar is becoming much more commoditized.
The other interesting developments are in the lease financing available. If the local incentives are high enough, in some places solar has genuinely reached grid parity pricing. NJ and CA are leading markets, NJ because of the very generous SRECs and CA because of their tiered pricing systems...shaving a little off your load can save you big bucks. There are a few utility areas in AZ that are also have very competitive incentives.
No one should consider the "incentive" price as the price.
Incentives are just money taken from productive taxpayers that could have been doing something else.
The point to remember is life-cycle capital cost.
PVWATTS tells us that a 1KW PV system will generate only about $100 per year in electricity in most of America.
That $100 has to pay for principle, interest, maintenance and installation for a PV that will last maybe 30 years.
The Chinese are loaning us $ at 4.26% for 30 year bonds.
We have to get down below $2/watt INSTALLED without subsidies to even speak of breaking even.
And we simply aren't there yet.
At present, solar consumes wealth, it doesn't create it.
Wholesale PV module prices have fallen precipitously since about 2008, and those upstream cost reductions have made their way through to consumers.
Not surprising. In 2001 polysilicon was selling for about $20 per kilogram and was up over $450 per kilogram in 2007 due to shortages. For the semiconductor industry the cost of the refined silicon is significantly less as a percentage of the finished product than is the case for solar panels where, it is my understanding, accounts for 45% of the cost of the final product. To put it in perspective in 2000 I was working on a project where we were considering buying a 200mm wafer (which would yield 40 to 80 thousand parts). These were pretty generic parts and essentially commodity items at a cost of $1 per part. A 200 mm wafer is about 50 grams. Last quote I saw earlier this year was for about $50 per kilogram. It would seem to me that price drop is simply a function of the drop in the price of the raw materials.
Looking into this in Los Angeles, you can spend as little as $5/W or as much as $8/W installed, depending on vendor and how picky you are about looks. On residential, a solid-black panel is much preferred to a standard blue-and-white pattern, especially if it can be seen from the street. Cost more and only a few vendors offer these as yet.
As far as batteries are concerned, off-grid solar will probably never be cheap, and isn't even all that useful unless there is no grid, or grid prices are extreme (something that will have other invasive consequences). How often does power go out, anyway? Is it worth adding $5 or 10 thousand to the system (not just batteries but isolation devices to protect the grid -- power company workers don't like it when the "dead" wires aren't dead).
The bottom line, though can be much cheaper. In LA, you get $2.20/W (effective) subsidy, plus the 30% from the feds, so a $20K 4KW system only costs you about $9K. You would also average $84/month in savings from your bill (counting all taxes and bumps), and be a permanent member of Tier 1. I think it works right now. Lower prices will help, but not all that much.
Why do I have the feeling that, however far PV prices drop, they will remain too expensive for general use.
PV prices seem immune to Moore's law.
In response to RS: Solar power doesn't have to 'create' wealth to make it a good investment. It only needs to beat the competition. Last I checked digging up coal, hauling it to a power plant, and transporting the energy to my house over power lines didn't 'create' any wealth either! I had to pay someone to make my refrigerator run!
But when I break even over the coal solution in six or seven years, and benefit from that financially for the next 23-24 years, then solar is a winner. Currently you still need incentives to get a six year break even point, but the prices will continue to fall.
By the time you can 'create wealth' from solar energy, it will be on every house in your neighborhood if not yours!
In response to PacRim Jim: PV prices don't follow Moore's law, but then please name some other advancement that has done so. Moore's law reflects an incredibly accelerated development rate that nothing else in history has come close to matching. However, I have seen calculations showing linear downward sloping PV prices as a function of total installed watts of PV. In other words, the more people devote their energy into producing solar power, and the higher the shipment and installation volumes, then the lower the price. No different than any other industry on the planet really.
Seriously, coal is cheap up front. But I'm not so sure that in the long run it's cheaper than solar.
Consider the environmental destruction of mountaintop removal. The health costs of increasing mercury poisoning, COPD, Asthma. I really don't think it's that cheap.
I'm sure there are toxins produced in the manufacture of solar panels - one of the reasons chinese panels are cheaper is because they can just go ahead and pollute with repercussions. We in the U.S. can do better than that.
First, I would seriously disagree with the statement "non-module costs now several times larger than module costs" -- if you dig into the report a little, you'll see that "module costs" as reported by the system installer were $3.2/W, representing about half the total cost. This is lower than a few years ago, but still very significant. For most residential installations, high-end, highly efficient modules get used because unshaded roof space can be limited, it is easier to limit the visibility of small systems, installation costs are inefficient at small scale, and so forth. Also most PV installers do not have the buying power or purchasing economy of scale of utility-scale installers, so they need to pay more and include a higher mark-up. I would expect these factors to become less important, but gradually.
As to where cost reductions might come from, I see three potential factors:
1) Shifting costs to home-builders. Most houses are not really built for solar power, everything from how far an out-of-view south-facing roof or patio is from the main electrical box to what kind of trees are planted and where. This one will take a long time to change, but particularly in places like southern CA I expect builders will pay more attention if and as the residential market expands.
2) Do-it-yourself labor. Right now a professional install and inspection is usually required to receive many of the available rebates and/or needed permits, and for very good reasons (including the voltages and current involved as well as the potential need for structural modifications, weight and wind loading, etc.) Technologies like micro-inverters, flexible lightweight modules, and easy mounting are emerging, but I'd guess it will take at least a few years for many of these to reach the residential market, and even longer before large DIY installations become a popular product category.
3) Installation speed. This factor I think is the most likely to show substantial improvement in the next few years if the residential market keeps growing. Already there is a large incentive to speed up the utility-scale installation process, and, based on what I've observed, plenty of room for improvement. I'd expect the bigger lease outfits to try and use this approach to get an edge on the competition, and in turn some of the smaller installers will use it to try and stay cost-competitive.
Overall my guess is that any rapid drops we see in installation costs will have more to do with supply and demand influencing an otherwise gradually declining set of installation costs.