Merit order

[Deju]

Dear all,

I want to better understand the importance of considering multiple service demand levels (aggregated vs disaggregated) for a realistic demand producing technologies diffusion in my model. That is to say how is the merit order of a certain demand producing technologies (competing for a given demand) constructed at a particular time slice in TIMES modelling framework? Does the duration curve (COM_FR) has an impact on the utilization factor of the technologies?

Regards,
Dejene

[Antti-L]

It seems the question is not directly related to VEDA at all?

Anyway, as far as I can tell, it all depends on the modeller.  

Nevertheless, the common way of modelling the demand technologies is to use ANNUAL level technologies, which produce ANNUAL level demands that have "load curves" (COM_FR).  In that case, all the competing technologies will see exactly the same demand load profile, and so the load profile has no direct impact on the "merit order": They all produce the demand according to the same profile.  I think this corresponds reasonably well with reality, because, for example, the demands for lighting, appliances, space heat and cooling can be assumed to have a similar profile among the same type of households, and all the competing technologies for each end-use must thus be evaluated against the same demand profile. But of course the load profile affects technology competitiveness through the prices of the input fuels in different timeslices. Demand technologies with an integrated storage can also be modelled consistently within this approach: the storage output will also follow the demand profile.  For different types of households/houses, the modeller can define different demands with different load profiles to increase heterogeneity in the modelling of the end-uses.

But I know some modellers that prefer modeling the demand technologies at the DAYNITE level.  In that case the modeller should carefully make sure that the competing technologies will see the same demand profile (or if they are different, they are so only for a good reason). That could be accomplished by e.g. by using fixed NCAP_AFs, which would force the correct profile for each technology. As you can imagine, the model size will become considerably larger in this approach.

In the first approach, the duration curve (COM_FR) thus does not have any impact on the utilization factor of the technologies. The utilization factors are defined by the modeller independently of the demand profile.
In the second approach, the duration curve (COM_FR) may or may not have an impact on the utilization factor of the technologies, depending on the detailed modelling of the technologies.  But if the competing technologies are forced to see the same demand profile, I would say that neither in this case COM_FR would effectively have any impact on the utilization factor, but the factor would be determined by the utilization profile defined by NCAP_AF.

[Deju]

Dear Antti,

Thank you very much for your detailed explanation. Now I understand how it works.