In my KTM-TIMES model, I tried to do modal choice based on time and cost. I used fuel cost as COMM-IN and time input (inverse of speed hr/km) as auxiliary input COMM-IN-A (screenshot attached). But the result in VEDA-BE shows modal choice based on fuel cost only and totally ignored auxiliary input. I expected some changes in modal share after introducing auxiliary input, but the result remain unchanged. Bus being cheaper technology is selected as the preferred mode but after providing time input, faster mode should also compete to give optimal modal mix. But time input is totally ignored in my model. Please help.
Have you also defined some value or budget for TIME? If time has zero value and is an inexhaustible resource, consuming time will not affect the results. Is TIME in your model a completely free resource, or is it produced by some process? If it is produced (it has a commodity balance), you can put a cost or a tax on TIME e.g. by defining COM_CSTPRD / COM_TAXPRD. You can also set a bound on TIME consumption e.g. by defining COM_BNDPRD.
I tried to limit minining of TIME commodity (MINTTB) by defining UP BOUND of time commodity (COM_BNDPRD) in scenario files. But there is still no change in output. I guess speed of the vehicles is not considered during optimizing the modal mix. I am trying to get an optimal modal mix of passenger vehicles for Kathmandu valley, Nepal based on fuel cost and an additional attribute i.e. travel speed of the vehicles. Even if the vehicle is cheaper, due to faster speed private cars and two wheelers should also come to compete with Bus. But there is only bus as desired modal despite of time input. I am trying to do it without time cost (Rs/hr) and with time cost. In both cases, the optimal mix that VEDA-BE gave seems to be the result of optimal fuel cost alone. Why time input (km/hr) is not considered in optimizing the modal mix? For reference, it will be of great help if I can have sample model related to time input.
24-10-2019, 01:21 PM (This post was last modified: 24-10-2019, 02:31 PM by Antti-L.)
I recalled there having been an ETSAP project addressing some of these Transport Modal Choice Modelling issues. I looked at their report now, and it seems you have actually been following their methodology, is that correct?
The authors are very experienced modelers: Hannah Daly, Kalai Ramea, Alessandro Chiodi, Sonia Yeh, and Maurizio Gargiulo. Therefore, I am sure the methodology is sound. Thus far, I have myself never tried to model travel time budgets in a TIMES model. But I can see the that the authors of the report make the following conclusions:
"The modal shift modelling methodology, implemented in a simple TIMES model presented here, is a significant and novel step towards incorporating behaviour into energy systems models.", but they also say that:
"The important next step is the calibration of the Travel Time Investment variable. The cost of TTI is critical to the choice of modes: At low TTI costs, the model only favors low-cost public transport, which is an unrealistic result. The TTI cost must be calibrated for the region in question to ensure the results are reflect a realistic baseline."
So, I guess your problem may be poor calibration? As I have no experience with this methodology, I am afraid I cannot help much with it. But some of the authors are also members in this Forum, and so let's hope they can offer some help.
