mcgowanmc , nopeak, RIGHT - ERoEI is everything. It is everything to the cheeta chasing the gazelle. It is everything to us. ERoEI seldom counts human energy but as we move closer to an ERoEI of 1 suddenly it will become clear just as it is for subsistance farmers or hunter-gatherers. For example the ERoEI of cane ethanol is better than corn. Partly this is because the cane is burned after extracting the syrup to provide some of the energy of processing (in other words no crop residues go back to the soil so the ERoEI is being supplemented by soil depletion) and partly because the energy of the workers (food, housing, transportation) is not counted. A cane cutter is expected to cut about 3 tons or more of cane a day. If machinery was used to cut the cane the fuel for the machine would be counted. Thus in that way of accounting if we used virtual slave labor to harvest corn and didn't count the food we fed them we too could have a better ERoEI on corn ethanol.

If a solar panel factory had to pay its workers in electricity produced by some of the panels it made - had to dedicate a portion of them to "pay" its workers, it would lower the ERoEI of the panels. Of course a human's pay is more than the value of lighting and heating a house, so we would have to subtract from the ERoEI of solar some amount to run electric tractors to grow the food for the workers, etc etc etc.

What level of ERoEI we can live with and still have industrial civilization is debatable. I suspect we are getting very close. I am sure that solar and wind will NEVER have enough to sustain the world as we know it now.

One suggestion for counting ERoEI but first the problem.

Problem:

When calculating ERoEI we have to calculate ALL energies that we put in. This is the part which is making trouble. The other part of calculating the energy that we get out is easy and straight-forward. The problem in the first part is that we don't know to what extend we must go. It get really complicated when we try to estimate energy needed to employ one person for one month. Should we estimate energy used for entertainment (show biz etc)? What about the wastes (do the person really need a car to drive to work? etc)? Another problem is to what extend we must calculate the energy (energy going to Morocco's miner of phosphate rocks to make fertilizer to be used in rice farm in china that grow rice for the factory worker in china that made that dolls that the daughter of ceo of shell get for her birthday etc).

Solution

Subtract the energy used for hiring humans. Only calculate the energy going in producing objects (foods, cars, buildings, electricity etc) assuming that humans are doing it for free or that the work is done automatically. Ofcourse there are non-human energy consumptions in a factory and farm and even in services, things like fuels, electricity etc.

Once this is done and we have a database of estimated energy used in making each of the objects, estimate where employees are spending their incomes. People use their incomes for buying two basic things: goods and services. When a service is bought money is given to another human who use some part of it in buying goods and some part in buying services. As you move forward in the life of the money eventually all the money is spent in buying some kind of goods. Once you have a database of estimated non-human energy used in production of goods AND database of goods usually bought by the people you can have a better (though not exact) estimation of ERoEI.

Now about the wastes. To what level people can use their money more efficiently (do you really need that car to go to work? is it really necessary to go abroad on vacations every year?) would be answered differently by different people so we don't go in that discussion.