Michigan's solar industry has a dirty secret: It needs a lot of coal-fired power.
The process of manufacturing base materials and panels to capture electricity from the sun is energy-intensive, utility officials say.

And that energy comes mostly from fossil fuels in Michigan, where up to eight new coal-fired power plants are on the drawing board.

That includes a $2 billion-plus, 800-megawatt plant proposed by Consumers Energy in Bay County's Hampton Township.

Hemlock Semiconductor Corp. in Saginaw County, a major world manufacturer of solar base materials, is Consumers' largest single electricity user.

But solar panels pay back the energy used to produce them in just a few years, says Vasilis M. Fthenakis in New York, head of the National Photovoltaic Environmental Research Center at Brookhaven National Laboratory and director of the Center for Life Cycle Analysis at Columbia University.

"It's about a year to 2.5 years, depending on the technology and where the photovoltaics are placed," said Fthenakis, who co-authored a study funded by the European Commission and U.S. Department of Energy on the environmental footprint of solar energy.

Solar panels have a life expectancy of about 30 years, he said.

But making sun power is not pollution-free.

Shocking, no? Pie-in-the-sky crashing to Earth and making a gigantic (and possibly delicious) mess. Actually, I disagree with stated number of 2.5 years, albeit I have not read through the study. The 2.5 years may be the the point where the energy generated by the solar technology surpassed the energy required, but that number certainly won't be true in Michigan. Why? Here's why: this is a solar map of the U.S.:
Guess where large-scale solar plants are going? From Wired Science:You might as well ignore the two one in New Jersey and the one in New York as they are more publicity stunt than anything else. Now overlay the maps and you get a good sense of why solar will simply not work in Michigan. It won't be 2.5 years. I'd estimate 10. Now you may say that that is still good considering that the panels last 30 years, but there is deception in this as well.

You see, when the calculation was made of energy generated versus energy used, the total well-to-use analysis was missing. For example, let's say that it takes 1000 kW-hr of energy to produce the solar panel and related materials that are required for construction. That 1000 kW-hrs was generated by plants on the grid. What is not taken into account is that those coal power plants are about 33% efficient, which means 3000 kW-hr of coal was burned to generate that 1000 kW-hrs of electrical energy. So triple the number. We're now at a 30-year payback period on just the energy. But even that is not all. That coal had to be mined and transported, and that took energy too. Plus the solar panels had to be transported and installed. That took energy too. Without specifics, the payback period here in Michigan would be closer to 50 years!

That is just the energy side. The analysis apparently doesn't take into account the chemical waste from the process of constructing solar panels or the raw material necessary. Some of the gases in particular are 10,000 times more potent than CO2 as a greenhouse gas. If you want to calculate the carbon footprint, or environmental footprint, of solar technology, it really does not look good.

If you compare technologies for generating electrical power, a fair basis would be to look at the ration of waste generated per unit energy generated over the entire life-cycle of the technology. Any guesses on which technology wins  - and I mean by a lot? Not solar. Not wind. Not hydro. Still thinking? I'll give you a hint - it's the other "n" word. That's right - nuclear. The cleanest energy we have ever had.

By the way, all of the above is engineering, not economics. I can say a few things on the economics as well, and by example to boot. At Oakland University, we installed solar shingles made by UniSolar of Auburn Hills. The 10kW installation, about the size you want for an average home, was installed on the student apartment clubhouse. Here's a pic where you can clearly see the shingles:The shingles each have two leads that have to be wired all together. The DC power then goes through a couple of inverters and integrates into the buildings power grid. During summer months when excess electricity is produces, it goes right back into the campus grid. So how much did the installation cost? About $130,000, most of that coming from demonstration grants. (that's just for the shingles, inverter and grid tie-in, nothing else) At the payback rate, this installation might pay for itself after I retire from teaching altogether (I'm still in my 30s). That's if they last that long. So therein lies the problem with photovoltaics - economic feasibility. Should such an array come down in price to about $1 per Watt of installed generation capacity (it's about $10 currently), I will buy these shingles myself. Until then, I'll stick with DTE..