I think this is the logical flaw in your argument: > But if they got paid less, they would do less. So whoever pays them is the one responsible for their energ

I think this is the logical flaw in your argument: > But if they got paid less, they would do less. So whoever pays them is the one responsible for their energy usage. Production (and thus energy usage) is not independent of cost (in this you are right), but not identical / interchangable in the sense that if cost goes to zero, nothing is produced and no energy is used. People produced things before there was money - as do animals. You can't seriously say that a pre-money society didn't use energy (or a bird that builds a nest doesn't use energy because no one pays for the bird). > On a previous post you excluded materials and wages. I can sort of understand wages, but why exclude energy cost of materials? ok, that was inprecise. Energy cost of materials should be included, but the rest of the material cost excluded (I know, you say, all cost of materials is energy...). > Let me shift a question to you: Which energy costs of manufacturing will you include, and which will you exclude? All measurable energy to produce something is included - except that I measure the energy in Joule which is more practical to analyze a wind turbine in terms of energy yield. Why do I think it is more practical? Let me also rephrase my questions to you. 1. Let's look at two identical wind turbines T1 and T2, identical environment conditions over 20 years (wind, ...). Over the 20 year span, they produce the same amount of energy. Do you agree? 2. Let's say wind turbine T1 is built in a world where average energy prices remain more or less equal over the 20 years. Wind turbine T2 is built in a world where the energy price is the same at first, but doubles one month after installation for some reason (invent your own) and stays the same afterwards. Now return on investment of T2 is nearly twice as high as the ROI of T1. Do you agree? But in your model (ROI = energy yield) you would say that T2 produces twice as much energy. Do you understand why I think this doesn't make much sense? I mean, I understand you, yes, but in your model you cannot predict the energy yield of a wind turbine without assuming a certain future price for energy - which makes this model impractical. It's good to be able to say "this wind turbine will have an expected energy output of x joules - independent of future energy cost fluctuations".