In honor of Earth Day, I thought I would post this piece I wrote in August of 2009 exploring the viability of solar as a baseline energy production method.

20110728_solarFarmShumlinOne friend of mine said to me, “I like that big battery in the sky.” If you’ve ever been underdressed on a really cold day and happen to walk under a cloud of shade and then into some blazing sunlight, you know firsthand the power of the sun. There’s a lot of potential energy that makes it to planet earth in the form of light from the sun. Harnessing it is a different story.

For the purposes of this discussion, I am not going to address how sunlight is used to produce electricity. If you are ignorant or curious, there’s a great article at The Scientific American. Actually, that article is a great explanation from Chemist Paul Alivisatos on how to generate electricity from sunlight. Of course, Google and Bing work wonders as well.

The great thing about solar is that the process of capturing the light and transforming it into electricity emits no pollutants. The sun will be around for a very long time and will provide enough energy to meet our needs many times over. Kinda makes it seem like a win/win, so why am I not singing solar’s praises over the hills and everywhere?

Because it’s expensive. Yeah, I know sunlight is free. It’s the harnessing that’s expensive. It’s so expensive in fact that the average solar panel will barely pay for itself during its average life expectancy.

North Carolina State’s Solar House serves as a good model. This house is small. NCSU estimates that the 3.6 kW PV (PV = photovoltaic) system supplies about half of the house’s energy. It doesn’t store power in batteries as it is connected to the grid.

Tracking the Sun: The Installed Costs of Photovoltaics in the US from 1998-2007, a study from the Environmental Energies Division of the Lawrence Berkley National Laboratory states that current costs of installed photovoltaics is $9 kW PV for small systems 750 kW, with no incentives. It is key to note that the study also found prices for installed photovoltaics remained steady from 2005-2007. Even assuming a slight drop in price to roughly $6.50 kW PV, the power panels on that house would cost $23,400. Take it off the grid and the price goes up to account for batteries. Add a single room AC unit and any installed price doubles.

Let’s assume that the average power bill for the NCSU Solar House without the assist of solar is $100/month. That’s an estimated savings of $50/month with the $23,400 solar panels installed. You would recoup your investment after 39 years of operation. That’s right kids, it would take you 39 years to get your investment back!

And that’s being conservative. If your house isn’t designed with solar in mind, you have a particularly rainy year, or your neighbor plants a tree that blocks the sun from your house you can expect your savings to be lower than 50%. Heck, you may not even see a return on your investment in your lifetime.

Some producers of solar powered hot water heaters suggest that your $5000 investment will pay for itself in 3-5 years. Seriously? Even if you get 100% water heater power from solar, you have to have started with an $80/month water heating bill alone! My entire electric bill is barely $100/month in the coldest months of the year!

“But it’s more environmentally friendly, Mike! You said yourself harnessing solar produces no pollutants.”

Yes, I did say that the act of capturing light and transforming it to electricity produces no pollutants. I did not address manufacturing processes required to produce the solar panels. Solar panels are made of gallium arsenide, which requires a chemically intensive manufacturing process, or silicon wafers, which require a thermodynamically intensive manufacturing process.

Gallium arsenide (GaAs) is a compound of metallic gallium and the notoriously poisonous metalloid arsenic. GaAs is produced using several different methods the first two of which are the crystal growth using a horizontal zone furnace (Bridgman-Stockbarger technique) where Ga and Arsenic vapor react and deposit on a seed crystal at the cooler end of the furnace and LEC (liquid encapsulated Czochralski) growth. A couple of other methods of producing GaAs include chemical vapor deposition reaction of gaseous gallium metal and arsenic trichloride which conveniently gives off chlorine gas and wet etching which gives off arsenic acid. But that stuff is used mostly in celestial applications like spacecraft.

It’s the silicon wafers that we’re concerned about as they’re more popular in earthly applications. Silicon is widely available normally bound in silica sand. Extracting it requires temperatures around 1700 degrees. Bonding it to form the panel can be done at a much cooler 1300 degrees. The average silicon solar panel has to make electricity for two years just to make up for what was used in production. Using our current environmentally evil coal to produce the power used to make the panels will more than negate the zero emissions solar advocates brag about.

“What about solar use on a large scale, Mike?”

Well, I’m glad you asked.

Solar panels take up space. Lots of space. I mean loads of space.

A typical solar panel produces 50 mW/sq inch. Let’s assume that you get peak production 5 hours per day. That’s 250 mWH/sq inch per day or .00025 kWH/sq inch per day.

I use 1000 kWH per month, or 33 kWh per day. That means to get all of my power needs from solar, I would require a panel 132,000 square inches large, roughly 916 sq ft or .02 acre. That doesn’t seem like a lot, but when you translate that to what you would need at a solar farm, it is quite a bit more.

There’s this phenomenon known as I2R losses. As electricity travels over power lines, power is lost. It is lost in heat through power line resistance. Operating power lines and grids at high voltages reduces these losses. That’s why a lot of power lines have a 1,000,000 volts running through them. Some lines even have up to 2,000,000 volts running through them. Anything higher than that and corona discharge negates the low resistance efficiency. Corona discharge is also dangerous and generally hazardous to human health.

The Climate Technology Program estimated in a study in 2003 that I2R losses across power lines from power station to end user was on average 7.2% in the US in 1995. For the sake of argument, let’s assume that technology has improved and average US losses are down to 5%. That means that now, I need a 965 sq ft power grid to power my house.

This doesn’t account for spikes in my electric use caused by starting appliances or by increased usage of certain appliances. To account for these spikes and safely provide my energy needs, the power company usually supplies me with 1.5 times what I normally use. That 965 square feet is now nearly 1500 square feet or .03 acre.

Oh and did I mention that I am single and am home very little. An average family of four can easily use up twice as much power as I do. Decision Data Resources reported that as of 2008, the Nashville, TN economic market had a population of 1.4 million. That’s 350,000 households. A solar farm would have to be 21,000 acres of solar panels alone just to power the residential needs of Nashville. Braidwood Nuclear Station, Exelon’s newest power generating station supplying power to Chicago, is capable of providing power to 2,000,000 homes and only takes up 4,500 acres. And that’s one of the largest in the country.

While solar may sound great, the technology has not progressed to a point yet where it can be a viable means of baseline energy production. It has a place in augmenting power supply needs and will work well on a residential basis for homes that are in remote areas far away from the grid and/or homes that see an abundance of sunlight. Furthermore, since the cost hasn’t decreased significantly in the past two to four years, I doubt that it will ever be a viable means of baseline energy production.