Start-up costs calculated wrong if start-up fuel is not one of the inputs

The calculation of start-up fuel costs in 3c_inputsLoop sums over nu(node, unit) although nothing (anymore?) ensures that p_uStartupfuel nodes are linked to units in nu(node, unit).

In addition, p_unitEmissionCost, which is part of start-up cost calculation, is only calculated for nu(node, unit) and gnu_input(grid, node, unit) in 1e_inputs.

In the emission cap constraint, summing over nu(node, unit) is also used when calculating start-up emissions.

Perhaps it was originally assumed that start-up fuels would also be included as input nodes in p_gnu_io but this is not well documented.

Suggestion: Create a separate nu_startup(node, unit) set that can be used when calculating start-up fuel use, start-up emissions, start-up fuel costs, and start-up emission costs.

The calculation of p_unStartup could also take place before 3c_inputsLoop.