Lighting units and efficiency

[Fanni]

Dear All,

We have two questions regarding modelling lighting in TIMES Starter.

1. We are a bit uncertain about the unit of lighting demand. Demand in the DEM file in the B-Y calibration data worksheet is indicated as being in bn-lum-yr, which is an activity unit. However, the formula in the relevant cells refers to useful energy in the calibration sheet, which is in bn-lum, a capacity unit. Is this a correct way of interpreting the file? 

2. A related point to the above: We have been trying to interpret the EFF parameter value for lighting in the RSD file. The heading states that the unit is PJ/PJ but this seems wrong as the numbers in some cases are above 100%. We assume that the correct unit is bn-lm/PJ, this would also make sense in terms of what we know about the efficiency of different lighting technologies. If this is the case, we are wondering why the RSD file does not use CEFF-I instead of EFF to define lighting efficiency.

Regards,
Fanni and Agnes

[Antti-L]

I'll leave it to Gary to answer 1), but concerning the efficiencies (EFF), for output-normalized technologies they are efficiencies from the inputs to the activity (defining how many units of activity are produced by one unit of input), and so the unit of EFF depends on the corresponding units of the inputs and the activity. If your input unit is PJ and the activity unit is bn-lum-yr, the unit of EFF will thus be bn-lum-yr / PJ, no matter what the heading states (however, I am not seeing PJ/PJ in the RSD file EFF heading).

Concerning the last part, "why the RSD file does not use CEFF-I instead of EFF to define lighting efficiency", I don't fully understand the question.  As the lighting technologies are output-normalized (PCG = output), EFF represents the efficiency from the inputs to the activity, as described above. The VEDA shortcut CEFF-I refers to a commodity-specific efficiency defined on a specific input commodity, and so it represents exactly the same relation as EFF, when you have only a single input.

[Fanni]

I'll leave it to Gary to answer 1), but concerning the efficiencies (EFF), for output-normalized technologies they are efficiencies from the inputs to the activity (defining how many units of activity are produced by one unit of input), and so the unit of EFF depends on the corresponding units of the inputs and the activity. If your input unit is PJ and the activity unit is bn-lum-yr, the unit of EFF will thus be bn-lum-yr / PJ, no matter what the heading states (however, I am not seeing PJ/PJ in the RSD file EFF heading).

Concerning the last part, "why the RSD file does not use CEFF-I instead of EFF to define lighting efficiency", I don't fully understand the question.  As the lighting technologies are output-normalized (PCG = output), EFF represents the efficiency from the inputs to the activity, as described above. The VEDA shortcut CEFF-I refers to a commodity-specific efficiency defined on a specific input commodity, and so it represents exactly the same relation as EFF, when you have only a single input.

Thank you Antti for the clarification of the Efficiency. I still cannot understand the units in the RSD file Calibration sheet in the case of lighting.

So the EFF is bn-lum-yr/PJ.
The Final energy is clearly in PJ as indicated.
The Useful energy is Final energy (PJ) multiplied by EFF (bn-lum-yr/PJ), which, I suppose should be bn-lum-yr but PJ is indicated.
Installed capacity is indicated as PJ/yr, in fact the same value as Useful energy, athough, on the Process sheet, the capacity unit is indicated as bn-lum, and on the ProcData sheet, Unit Act/Cap is bn-lum-yr/bn-lum.

Thank you very much if you can help me out here.

[Antti-L]

Yes, I see your confusion. The question should best be addressed to Gary (or Anna, or Pat) who have built the Starter model.

I can only see that there are efficiencies such as 1.9, 2.0 and 2.13 on the calibration sheet, which indicates that the unit of the "useful energy" simply cannot be PJ, but you should better trust that the units are as given on the Process sheet, i.e. the activity unit is bn-lum-yr, and therefore the unit of the output is also bn-lum-yr, and finally the unit of the useful energy demand is also bn-lum-yr (because I cannot see that there would be any unit conversion parameter between the process activities and the demand).

I am not sure the efficiencies for (bn-lum-yr / PJ) are actually correct; they look too low. For example, currently a standard LED lamp produces about 120 lumen/W, which I think converts into 3.8 bn-lum-yr/PJ. But because the final energy use and the demands have been calibrated, for the energy results it doesn't matter if the efficiencies are not quite correct in absolute terms, as long as they are correct in relative terms (proportional differences across technologies). For investment costs it of course does matter...