Soy for the Hummer

Several weeks ago I drew some comparisons between the energy needs of every day devices and the capability of a typical cyclist in “Watts in a Kilojoule.” I received much public and private feedback on that post. I’d like to expand on one of the themes in that post.

A couple of astute readers pointed out carelessness in my analysis. I compared energy needs of cars powered by internal combustion engines to fully electric vehicles on a one-to-one basis. This led me to overestimate the demands that would be put on our power grid if all vehicles on the road were converted to 100% electric. I've spent 10yrs of my career working with battery powered electric vehicles, so I should not have been so careless. Guess I tried to keep things simple, thinking nobody would notice.

Internal combustion engines are only about 20% efficient, interestingly, about the same as a trained cyclist. This means 80% of the consumed energy is wasted as heat. This is why cars have big radiators and humans sweat. An electric car, on the other hand, is about 80% efficient. Only 20% is wasted as heat. Grid to wheels isn't quite this good though, as charging the battery after each drive wastes energy in the form of heat in the charging equipment and the battery itself. Grid to wheels efficiency is closer to 60%. But still, this is three times more efficient than gas engines.

If all vehicles were converted to electric, we’d only need to build 200 new nuclear power plants, not 600 as I originally surmised. That’s only four nuke reactors per state. Relieved now? Note I’m not anti-nuclear, I’m actually strongly in favor of this form of energy over coal or foreign oil. I suspect many readers are not keen on nuclear power. My point here is to show the magnitude of how much energy it takes to power our mobile lifestyle.

I also pointed out previously that if we chose a renewable biofuel path to reduce CO2 emissions, we didn't have enough farmland to power our needs and would have nothing left to eat if we tried. I’d like to expand on this point a good deal further.

I don’t have anything against limited biofuel use. I know a couple people who run biodiesel in their cars. A teammate uses unprocessed McDonalds oil in his car. Smells like French-fries! Using biofuel that might otherwise be wasted is a brilliant and noble thing to do. I do, however, have a problem with large-scale biofuel production. There are ethical challenges here, in that our “need” to drive 2000kg SUVs around impacts world food supply when increasing percentages of the fuel we burn is mandated to come from food crops. The whole biofuel industry in this country is mired in politics at its ugliest. But I don’t want to, nor need to go into the messy side of biofuels here. It is much easier to show that biofuel production is very inefficient use of land.

Let’s take a step back and look at the bigger scene here. Fossil fuels are hundreds of millions of years old. These hydrocarbons were created with energy from the sun. It takes some energy to extract and refine crude oil, but over 90% of the energy extracted from the ground is made available in the various distillation products. Crude oil has very high energy density, and eons of solar energy are stored up in it (along with carbon that used to be in the air, topic for another post).

When we grow a crop with the intent of turning it into biofuel, what do you think the net efficiency is? 50%? 10%? 1%? You will be surprised. Many crops grown for biofuel are once a year harvest in the US. A good part of the year, like in winter, the land is unproductive and producing no energy. Even during most of the growing season, the plant is capturing solar energy to build the infrastructure to produce the fruit that is harvested to be fermented or pressed for oil.

Photosynthesis has a peak efficiency of about 6%. This is the process in the leaf itself, not a measure of all the sunlight illuminating a field. This is not likely to improve dramatically anytime soon with genetic engineering. Photosynthesis uses only a portion of the Sun’s spectrum, and not all of the energy that is absorbed is retained.  This already puts a pretty pessimistic upper bound on biofuel efficiency. The average efficiency of photosynthesis over a course of a year is no more 0.6%. This is total energy the plant captured from available sunlight. With current biofuel processes to make fuel for cars, we must next make ethanol or biodiesel. Only certain parts of the plant are useful for this. You can’t press oil out of cellulose, where some of the plants energy is stored. The efficiency now drops to less than 0.1%! This doesn't even factor in that a lot of outside energy must be put into this process to till the land, produce the fertilizers, and ferment or press the grain. (see Michel and ARPA-E for references)

Ok, so far we've learned that biofuels can at best capture 1 Joule out of every 1000 Joules of energy from the sun. Since this is a liquid biofuel, it will be burned by a combustion engine with about 20% efficiency. That 0.1% just dropped to 0.02% for a Sun to wheels efficiency. See how absurd this is getting?

So here’s where I’m going with this. State of the art photovoltaics can now convert about 30% of the Sun’s energy directly into electricity. Theoretical peak efficiency is about 66%, but we are still a ways away from that. The beauty with PVs is this. The energy is already in a form that can immediately be fed across the country with negligible additional processing or loss. Using the 60% grid-to-wheels efficiency established earlier for electric cars, photovoltaics gives 18% Sun to wheels efficiency. This is 900x more efficient than biofuel for combustion cars! Right now, today, you can get 900 times more energy from an acre of solar cells than you can from an acre of biofuel crop over the course of a year. Further, you can use unproductive land for photovoltaics. No farmland for gas tanks is needed at all.

Now some biofuel proponents may find this troubling. Even if these estimates are off by a factor of 10, or if we suddenly engineer a new biofuel plant that doesn't go viral and consume the planet, harnessing energy from the sun with photovoltaics will still be far more efficient from a land use perspective than growing crops.

The big problem with photovoltaics today is they are expensive to produce. They can’t compete with cheap oil and coal on a large scale. Payback periods are very long.  I believe we must end mandates and subsidies for biofuels and redouble our efforts on direct electricity production form the sun. If there ever was a need for government funded research or subsidies, solar energy and battery storage technology are it.