One of my readers (the only one?), Phillip Pare, sent this email: “In the June 15 edition of Engineering News, I read an article by Jeremy Wakeford, in which he talks about the energy return on investment which, he claims, averages about 18:1 for wind power.

“This would seem to be at variance with my understanding of one of your earlier comments that the energy harvested from a wind turbine would take a very long time (perhaps in the order of 23 years) to be greater than the energy used to manufacture the turbine. Is there, perhaps, something that I have misunderstood?”

The concept of ‘energy return on investment’, or EROI, is one of those terms which have popped up recently. The definition of EROI is: (usable energy acquired)/(energy expended to create the energy source). Simply put (oh, yes, for the Natal graduates), if you take a diesel fuel machine and use it to plant a field of sugar cane, then you will get a certain tonnage of cane. If you use another machine to harvest the cane and take it to a mill in a truck and, at the mill, use cane knives and presses and so on, then, potentially, you will end up with some ethanol.

The EROI for this process is then (diesel fuel energy for machine for planting and harvesting and transport)/(energy content of ethanol).

You will find that this figure is about 5 – so, it makes energy sense to plant cane to get ethanol. The EROI figure for photovoltaic is about 5.8 and for hydropower it is about 100. For wind, it is about 18. The EROI figure is misleading. There is the question of the energy input to manufacture the machines in the process and the timeline to which the EROI relates.

In our example, do we allow for the energy input to melt the steel to fabricate the sugar cane machine for planting and harvesting? If not, why not? Say, it takes two years to build a processing machine. Once built, it uses 5 ℓ of diesel an hour, takes in sunlight and water and garden waste and, 2 000 hours later, produces 20 000 ℓ of diesel. The EROI is, thus, 10. But, if the machine needed 10 MWh to build, then the whole sum is negative. For the EROI to make sense, the energy input (say, sugar cane) and output (say, ethanol) must be much greater than the energy cost of fabricating the equipment used in the process.

This brings us to the second point – the time-line. If I make a wind turbine out of a bicycle wheel and it produces 50 W when the wind blows and if my little turbine lasts for 50 years, then the EROI will be much different to it lasting, say, two years.

Stating that wind turbines have an EROI of 18 does not mean you get back the input energy in one year – it all depends on how long and often the wind blows.

However, a summary of all the reports and studies to date was compiled by Cutler Cleveland (Boston University Oil Drum October 19, 2006). He found that wind turbines typically pay for their energy content within the first year of operation. Which settles it, right? Well, Engineering News readers Dirk Heydorn and Ernst van Niekerk came up with energy requirements to melt 267 t of steel (the mass of a 2 500 kW wind turbine) and the lowest estimated ‘payback time’ based on energy produced by the turbine was about eight months.

I have my own (higher) estimate. The thing to note is that many estimates ignore the fact that there is an energy cost in transport from Europe, erection, manufacture of rare-earth magnets, power lines, cabling, alternators, and so on, and generation estimates are often based on the turbine producing 50% of rated power for 25% of the time when, in fact, these figures may not be reached.

In summary, the EROI is a figure which estimates energy return but, for a variable- energy supply, such as wind, the time taken for that return is also variable.

To subscribe to Engineering News's print magazine email subscriptions@creamermedia.co.za or buy now.