There is a myth out there that says solar panels take more energy to make than what they generate. Wrong! http://solarbus.org/documents/pvpayback.pdf
Energy Returned On Energy Invested…
Solar panels will take between 6 months and 15 years depending on efficiency of source energy, type and location to generate the amount of energy it takes to make them and they last much longer than that. Conventional silicon panels are the most energy intensive but I do believe they last much longer than thinfilm amorphous (which is less efficient and degrades over time). Copper indium gallium selenium (CIGS) takes about a year and a half to reach EROEI and is almost as efficient as silicon (and doesn’t degrade). Gallium arsenide (GaAs, the most efficient and robust) requires a bit more energy per cell, but should use concentrating optics (to about a thousand suns) which would greatly shorten the time to reach EROEI to about 6 months (since much less energy is required to make the optics). Realize that aluminum and glass (included in these figures) require energy to make too.
It is generally assumed that wind turbines take only 6 months to reach EROEI. With a lifespan of 20 years, that would be an EROEI of 40 which is better than oil!
I figure that average solar panels generate 10 x the electricity than the equivalent amount of energy used to make them (or an EROEI of 10). This is assuming it takes 3 years to reach EROEI and that it lasts 30 years. However, I have stumbled upon a very serious question… The above listed links state x amount of kWh required to make a certain kind of panel. Does that include the 2/3rds energy wasted in the conversion process to electricity in the first place? Or is that just an equivalent used to measure the amount of fossil fuels used (assuming very little electricity used)? Worst case, it could take up to TEN years (if the bulk of the energy required is electricity without regards to all the extra fossil fuels wasted in the conversion process in its generation). With a lifetime of 30 years, that is an EROEI of only 3! Still positive, but not good. After further searching, I believe it IS mostly electricity since an electric arc furnace is used to make silicon ingots. Hmmm. http://www.electrochem.org/dl/interface/wtr/wtr08/wtr08_p30-35.pdf . Electricity does not always have to be “dirty” however 🙂
For comparison, back in the early days of oil in Pennsylvania, it only took 1 barrel of oil to recover 100! And now, during the peak (when global recoverable oil is already half way consumed) the EROEI is still about 33 and 10 for imported and domestic oil, respectively http://www.theoildrum.com/node/8625 . This is where the excitement over renewable energy quickly fades. Even though wind power has an EROEI of nearly 40, it takes much longer to regain said energy, than say, a barrel of oil from the tar sands with EROEI of only 2 to 8 (also listed in the link to the oil drum)! Thus, the quickness of which fossil fuels can be extracted can justify a much lower EROEI in our world of monetary concerns and ambition, especially since interest rates charged for building generating capacity IS based on time.
Is it possible for solar energy to make itself? At first glance, it seems so. However, in the overall manufacturing process, employees still use gasoline powered cars and parts and tools (everywhere) made from fossil fueled processes. Ok, let’s pretend that all the parts and transpo are magically already made and powered from clean energy sources. How long can it take for solar and wind to “energy make itself”? First, we need to make a solar panel (or concentrated array) with the greatest EROEI. Second, we need to “invest” some of the electricity it generates into its continued mass production. And third, we need to account for a decreasing supply of fossil fuel which suggests that we need to mass produce an even greater amount of solar capacity. Fourth, let’s not count on any fossil fuels or nuclear for sake of (this) argument except for initial start up.
I will assume that GaAs Fresnel concentrated PV with two axis tracking has an average EROEI of 9 months and that it will last 30 years placed primarily in the desert regions. Ok, we use fossil fuels and or nuclear to build a gigantic robotic factory that can mass produce 10 square miles of GaAs solar collection per year. I will further guess that the building of the factory itself will entail an additional 30% energy input to the final product.
To add even further, let’s assume that it requires an additional 50% of all that to mine raw materials, make parts and factories for a decent storage system (be it the LiFePO4 battery, pumped hydro or whatever). Thus, it takes 18 months (in this imaginary example) for a GaAs Fresnel array and its storage to reach EROEI. If we use all of that energy to power air conditioners and watch fashion shows, liquid fuels are now (30 years later) too expensive, parts break down, we have no energy left to build another factory… end of story!
But if we use 20% of that energy to build additional robotic GaAs plants, in almost 8 years, we would have built another such factory. Now we have double the production capacity. In 15 years, four times, and in 30 years, 16 times the production minus the first one that would then be breaking down and being recycled. At the turn of the century, we would be building some 30,000 square miles of solar collection… every year. Surely, this would be enough electricity to power 20 billion people (and their electric cars) at near today’s western standard assuming better insulation, efficient led lighting (which is already almost twice as so than florescent) and slightly less overall transportation miles per person.
I based the math on a factory doubling rate every 7.5 years… After 90 years, that’s about 4,100 factories minus all the start up ones recycled after their 30 year lifespan.
Now, that amount of land seems preposterous, however America alone has paved 50,000 square miles just for roads and highways! Note that it will be impossible to grade the deserts for such an ambition, as that would cause environmental problems. Thus the mandate to post mount whatever best collection media without need for bulldozing. 1,000,000 square miles is 2% of the Earth’s land area. I would imagine that solar collection should be kept at a maximum of about 1%… Thus the need to look into other forms of clean energy and the need to halt population growth.
This is the concept of the
It is indeed wishful, as the basic premise excludes the current political and short term financial motivations. It is, however, imperative (if we exclude advanced nuclear) in order to sustain civilization!