Technocrat

Photovoltaics follow Moore's Law

David Leppik
Fri May 16 08:57:25 -0700 2008
About+Technocrat.net
The price per watt of a photovoltaic module has been decreasing at a remarkably steady 6% a year for the last 25 years-- despite wavering levels of R&D.
Put differently, watts per dollar has been increasing exponentially. Here's the graph, for those too lazy to follow the link:

Photovoltaic module price per watt, 2006 dollars
This is market price, not production cost. (This is from my brother's blog, but to my knowledge the data isn't the web anywhere else.)

I do a lot of data analysis, and this is a remarkably good fit to the curve. Up until 2003, that is, when prices start increasing due to capacity shortages. Not enough factories to handle the demand both for solar cells and for the silicon wafers used to produce them. As new capacity comes online, we can expect the prices to at least drop back down to the trend line.

If current trends continue, solar cells are going to become cheaper than $1 per watt around 2023. That's far more pessimistic than a lot of scenarios, especially since it doesn't take into consideration the super-cheap (to manufacture) thin-film technology that's now commercially available. Nor does it factor in the billions of dollars that are now being spent in solar R&D. But even in the 1980s, when investors were running away from solar as fast as they could, it appears that manufacturers were finding ways to make them cheaper.

As for when it will be cost effective for the average person to install solar panels on the roof, it depends on the installation cost as well as the lifetime cost of grid power and interest rates. But for Minnesota, the article estimates 2015.


 
Photovoltaics follow Moore's Law
President4242
Fri May 16 09:28:55 -0700 2008

It occurs to me that if you also plot Moore's law in terms of a vertical of processing power and a horizontal of wattage consumed, you've got an opposite slant.

I wonder at what date the two slants meet?  Seems to me that would be a MUCH better predictor of when cost effectiveness would hit for the average person in new construction- and based on my own experiementation in this arena, I'd suggest it's more like 2010 (in that between LED lamps, solid-state laptops, solid state refridgeration/heating tech, solar budgeting of insulation, and flat screen low power TV sets, most household uses of power will be able to be covered by then with a 12Volt DC system and ambient energy of various types).  I also note that "Extreme Home Makeover" on ABC has noticed this as well- and has added utility bills and local energy generation to some of their charitable rebuilds.

 

  

costs versus benefits

zogger
Fri May 16 11:08:15 -0700 2008

By far the cheapest way to get affordable solar is to drop your demand so the initial upfront cost isn't as high, ie, you won't need as much so you can get by with a cheaper rig. The two lines of supply versus demand cross much sooner then, and taking away from the demand side is cheap compared to anything else. There's also an intangible to quantify cost and benefit ratio, what is it worth to the homeowner to guarantee at least some supply of electricity that isn't depending on the grid and prices out of his control and geopolitics and etc, keeping in mind solar PV versus conventional grid supply is never an either/or proposition, it can be any percentage of one or the other for your total electricity needs? When things are running just ducky, grid supplied being affordable and there and working, the advantages of solar (with battery banks*, I wouldn't have solar without them) is not as apparent. Go through a week or more without power and be sweating in a heat wave without even a simple fan and watching the food rot in the freezer and be sitting in the dark, etc, and it starts to look like a bargain. Conversely in the winter a lot of modern furnaces can't function without some electrical input into the controller box, a simple solar rig will insure that works so your house won't freeze out and you lose your water pipes, etc. To *me* that is the overwhelming advantage to solar and why I made a point of affording at least "some" solar many years back, I have a guaranteed minimum supply for decades, all paid off, owned. Ut's not currently enough to replace all my day to day needs, but push comes to shove, heck ya I could get by with what I have, enough for lights, run a laptop, run a little TV and radio, etc and eventually some sort of at least local transportation of the electrical drive kind, keep the batteries charged up for a once a week run someplace.

* I have told people to actually start with a charger/inverter and the battery banks before they get solar panel 1. Keep them topped off from the grid, then you have a circuit or two on a heckuva UPS system, use it for your home office and the freezer for instance.

  
Photovoltaics follow Moore's Law
Thue
Fri May 16 11:09:24 -0700 2008

It seems silly to describe everything which follows an exponential law as following "Moore's Law".

In the present case, just say that the price decreases exponentially, 6 percent per year...

  
Photovoltaics follow Moore's Law
Adam Hazzlebank
Mon May 19 02:33:42 -0700 2008

Doesn't even look like it's decreasing exponentially to me, more like it's decreasing linearly (the y axis is dollars).

  
Photovoltaics follow Moore's Law
The Baron
Mon May 19 06:45:29 -0700 2008

The Y-axis is in Dollars, but the distance from $1 to $10 is the same as the distance from $10 to $100, thus a logarithmic chart, i.e. a straight line is exponential growth.
  
Photovoltaics follow Moore's Law
Uncle Entity
Sat May 17 15:44:37 -0700 2008

Might have to recalculate your chart...

I also don't think price/watt is an useful metric to measure efficiency gains (as the Moore's Law comparison would imply) since as you admit it has gone up from causes other than efficiency gains. Maybe watt/cm² or something.

correlation != causation and all that...

  
Photovoltaics follow Moore's Law
Uncle Entity
Sat May 17 16:05:15 -0700 2008

Also, I think it's reasonable to assume that the installation will still be serviceable in the long term — once the panels wear out the new ones could be bolted in place with minimal costs over the purchase price. With this you could possibly reach the break even point at some point earlier by retrofitting new modules when they become economically viable with virtually zero installation cost (a case of beer?) if you get a couple friends to help out

Or this at least seems reasonable to me given that you would get surplus watts out of the retrofit (assuming a replacement with equal surface area) to sell to the utility to offset some of the previous lost costs.

Might be worth calculating by someone with the math skillz.

  
Photovoltaics follow Moore's Law
Yves Quemener
Sun May 18 11:58:15 -0700 2008

Moore's law is about doubling every two years. If it had been a mere 6% per year, Wolfenstein 3D would still require a super-computer to have any chance of running at 5 fps

  
 

Do you have questions about the site? Write to bruce at perens.com . List of all articles we've ever published. Source Code Developer's Documentation RSS XML Sitemap XML News Sitemap Valid XHTML 1.0 Strict Valid CSS! [Valid RSS]