Ken Novak's Weblog
Purpose of this blog: to retain annotated bookmarks for my future reference, and to offer others my filter technology and other news. Note that this blog is categorized. Use the category links to find items that match your interests.
Subscribe to get this blog by e-mail.
New: Read what I'm reading on Bloglines.

Projections on solar photovoltaic cell prices

I sometimes see the question, "when will solar cells arrive?", as in "when will they be economical enough to make a dent in our energy system (and ultimately global warming)?"  Here's my take.

The cheapest PV cells today cost about $3 per watt of peak power.  At every price level there are some new markets; e.g., at $3, a variety of traffic and other outdoor lighting can be more cheaply installed with PV+battery than with digging into conduit to find a tap into the grid.  When prices hit $1.50/watt, PVs become attractive for many buildings (especially in states with high energy prices and with net metering on the grid), or off-grid locations where gensets are used today, or developing countries where grids have yet to be built).  At $1, it's competitive with grid power, leading to ubiquitous use.  So that's pretty big applications at one more halving of cost, and PVs really "arrive" at, say, 1.5 halvings.

Incremental advances in chrystalline and amorphous silicon cells have had an average 8-year cost-halving time.  (This also correlates with experience curves based on doubling total production).  So that puts at 2011 and 2015 for the "big" and "bigger" applications.

Disruptive technologies could change this.  Different substrates or radically different processes can jump us to new experience curves at lower initial prices.  Several nano-based technologies are mature enough now to have received substantial venture funding in 2002-2003, and in these post-bubble times, that means they are very close to product.  For example, Konarka claims they will sell organic dye PVs in 2004 for $2/watt.  STMicroelectronics recently announced similar organic dye plans, aiming for $.20/watt eventually.  Nanosys is planning building-integrated systems around $1 for 2007.  Nanosolar makes similar claims.  There are also startups pursing solar thermal options, like Energy Innovations.  Plus, there are several distinct approaches that are pre-VC funding, or are inside larger corporations, that could bust out as well.  The failure rate on these will be high, but there are many more that are active now than there have been in the past.

My gut feeling is that we're looking at $1.50 in 2007, and $1 in 2010. 

Market adoption comes later.  Judging by other solid-state innovations, there's a lag of at least 5 years to mass market levels.  When presidential candidates talk about a federal program for renewables, I think it's in market adoption where federal involvement can really help.  (This is in addition to the current substantial federal funding for basic nanotech research.)  Market-building activities could include, for example, an auction-based PV purchase program for federl facilities, and/or it could institute a production-credit program (based loosely on the wind production tax credit, perhaps).  These would accelerate travel down the experience curve, especially for the "balance of system" components of a solar installation (the parts other the cells themselves, which aren't likely to change very much).  In the near term, these programs would embolden investors and cut perhaps a year off the arrival date of the technology; but after that, it would make a much bigger dent in the market adoption lag.

-- Ken Novak, 25 Sep 2003

Public Bloglines
Technorati Profile
news search
blog search
Last update: 11/25/2005; 12:18:12 AM.
0 page reads.