By "no-load fuel cost" I assume you mean "Fixed generation cost". Is this value set by the Fuel consumption data in the Fuel curve calculator, specifically, the fuel consumption at 0 kW? Is this cost always incurred, even when the generator is not running (one would think the generator is off and not consuming fuel at 0 kW)? If my Fuel consumption data defines the generator to consume 0.93 L/hr at 0 kW, and cost of fuel is set at $1.04 per liter, how can Fixed generation cost be greater than $1.04 per hour? Further, if I input the Fuel consumption data to be 0 L/hr at 0 kW, I still get a fixed generation cost of $0.654 per hour. I guess I am just not understanding the Fixed generation cost.

The Fixed Generation Cost = Intercept Coefficient * kW capacity of the generator * Fuel price.  What may be confusing you is that HOMER uses an affine (linear with intercept) function that is creates based on the data you put into the Fuel Curve Calculator. We find this is a very good approximation of real diesel fuel curves.  I suspect that what you might have done is to take some real diesel fuel curve data and then just experiment with setting the zero load point to zero fuel consumption, but left the remaining data.  As a result HOMER could not come up with a good affine approximation.  The best it could do still had significant fuel consumption at zero load.  HOMER uses the affine function rather than the actual data because it creates a constant marginal consumption which simplifies the dispatch optimization.  Regarding your other questions, 1) there is not fuel cost when the diesel is off. 2) The only fuel consumption data that HOMER needs is the two parameters of intercept and slope that you see on the Fuel tab.