I recently received sales material for solar power panels. According to the literature, I can buy a solar power system that will pay for itself and then continue generating returns for the next 30 to 40 years without the risks associated with investing in stocks. Something about this pitch smacks of overlooking areas of risk when adopting an immature technology. Perhaps I am merely a pessimist on the current state of technology for solar energy, but I think there are significant adoption risks that are analogous to ones that I have had to consider with other technologies.

According to the literature supplied to me: I can buy solar panels with a 30% tax credit and a generous rebate program. Despite the steep discount – these panels will take at least 5 years to reach a breakeven point – and that point assumes I can choose the perfectly sized system for my house. The system comes with a 10 year bumper-to-bumper warranty and a 25 year manufacturer’s warranty. This sounds like a great deal right?

To reach a breakeven point in 5 years, the solar panels need to provide a better than 14.4% annual rate of return on my initial investment. That is quite a healthy rate of return for any static installation to sustain for many years, but I am willing to consider that it might be realistic. I suspect that that rate of return does not include the cost for dismantling, maintaining, and replacing the solar panels – but for this scenario – I am willing to consider those as free activities – no future costs.

I expect that solar power technology will continue to improve each year – in fact, I expect that the rate of improvement of the solar conversion efficiency might mirror Moore’s law about transistor density in some analogous fashion. If this is true – and the organizations providing the rebates and tax credit subsidies are counting on it – solar power technology will be somewhere between four and eight times more efficient in 5 years than a system I would install today.

This scenario strongly reminds me of the opportunities to buy desktop computers in the 1980’s that were future proofed by having the system motherboard be able to accept the next generation processor by merely dropping in the new processor when you wanted it. My experience with those systems was that the extra expense for the extra complexity on the motherboard was not worth the extra cost.

Additionally, there were other things in the system that also changed that made using the motherboard beyond a few years a bad idea – namely, the operating system kept evolving, the device drivers kept evolving, and both of these provided no support for “old and obsolete” peripherals and modules. It was much cheaper, easier, and safer to buy what you needed when you needed it – and then replace it with the next round of available devices when you needed to upgrade.

Am I inappropriately applying this lesson from the past to solar power? According to my lessons learned, I should wait a few more years and realize the resulting improvement in cost and efficiency of solar power – In other words, I might come out ahead if I wait and do nothing today. I believe this is the condition of a classic early adopter. Have you experienced this type of trade-off when choosing components and features for your embedded designs?