Question about using COM_FR

[Xin Wang]

Hello,

In my TIMES model, 4 timeslices have been specified, SD/SN/WD/WN, and their default YRFR are simply set as 0.25/0.25/0.25/0.25. When I use the COM_FR attribute to specify the fraction of CEMENT demand over each timeslice, for example, 0.27/0.23/0.26/0.24. It turns out that the demand of CEMENT are modified according to the COM_FR, as well as the CLINKER as FLOW_IN in the process of producing CEMENT. However, in the process that CLINKER are produced as FLOW_OUT, the output over each timeslice is consistent with the default length to be 0.25/0.25/0.25/0.25. Similar problems also happen when we add COM_FR for STEEL, which is another sector with more than one step in its production. 

Does anyone know what I should do with this problem? Thanks in advance!

Best,

WX

[Antti-L]

There are many possible ways to correct that problem. Perhaps the most straightforward solution is to define COM_FR for those intermediate products as well. COM_FR can be defined for any commodity, not just demands. But then you must also make sure that the clinker process has energy as its shadow group.

Another solution is to track the clinker demand at the SEASON or even DAYNITE level, and define the timeslice level of the clinker process accordingly.

[Xin Wang]

Thank you, Antti!

I'm still confused about the model result that the flow_in clinker defined according to the COM_FR attribute, while the proportion of flow_out clinker in its production process remains consistent with default length of timeslices. Does this mean that the model fails to realize the mass equilibrium in each timeslice or that the flow_out clinker specified at 'DAYNITE' level in result file makes no sense as the timeslice of clinker and its production process is defined as 'ANNUAL'? 

I tried the method of defining COM_FR for clinker, and it worked. In the process of producing clinker, coal will be consumed along with other materials. If we look at the coal mining process, the proportion in each timeslice is still the default length. Is this why you recommend that I make sure the cliker process has energy as its shadow group? And can you provide some details about the 'shadow group'?

About the solution of tracking clinker demand at 'SEASON' or 'DAYNITE' level, my understanding is that I should define the timeslice level of clinker commodity and clinker production process to be SEASON or DAYNITE. And if so, will the output of clinker be accordance with the COM_FR or the default length of timeslice? I'm really not sure about priority of COM_FR and default settings of timeslice in determining the proportion of output.

Your quick reply helps a lot and also inspires me to think more deeply about the issue of setting timeslices in our model. Thanks again!

Best,

WX

[Antti-L]

Hmm...  I don't understand how the flow_out of the clinker process could be DAYNITE, if the timeslice of clinker and its production process are defined as 'ANNUAL', as you stated. Maybe you can clarify that?

Concerning the shadow group: TIMES sets the timeslice levels of the input flows according to the highest level among the flows in the shadow group, or the process timeslice level, whichever is higher.  Therefore, if the process level is ANNUAL, the shadow group should have a higher level in order to have the COM_FR be reflected in those flows.  The shadow group is set internally by TIMES according to the commodity type of the PG.  If the PG is of type MAT, and you have MAT inputs as well, the shadow group is also set to MAT. And therefore, the inputs would normally be tracked on the ANNUAL level. But you could define some dummy MAT input at the DAYNITE level to rectify that.

About the solution of tracking clinker demand at 'SEASON' or 'DAYNITE' level: Yes, you should define the timeslice level of the clinker commodity and clinker production process to be SEASON or DAYNITE. And if so, clinker will be produced in accordance with the load curve of the demand for clinker. I think that's probably what you would like to have, or am I mistaken?

[Xin Wang]

Sorry, I didn't make it clear. In my model, the electricity 'ELC' generated is defined as 'DAYNITE'. And in cement and clinker sector, the 'ELC' is transformed to 'CEMELC', which is also at DAYNITE level. In the result, the flow_out of the clinker process is ANNUAL. However, when I track the flow of CEMELC, I find that the CEMELC consumed in cement process is bases on COM_FR, while the CEMELC consumed in clinker process is bases on default setting of timesilces.

For example, in the figure below, 'IND_CEMGRD' is the cement process, 'TFE-CEMELC' is the conversion from ELC to CEMELC, and the other processes are the clinker process. Here we adopt an extreme COM_FR of 0.54/0.06/0/36/0.04, just for an example.

[Xin Wang]

About the clinker production, I think its demand should be decided by the cement demand, because clinker is the commodity-in of the cement production process. However, it seems that the link between cement demand and clinker demand is missing. Do I have to set COM_FR for clinker production? Sorry to trouble you so much.

[Antti-L]

Ok, you say that it seems that the link between cement demand and clinker demand is missing.

Have you tried setting both the COM_TSL of clinker and the PRC_TSL of the clinker processes to SEASON or DAYNITE?  Doing so will establish the mass balance at that level (SEASON/DAYNITE).

Q: "Do I have to set COM_FR for clinker production?"  

A: No, if you link the clinker production to its demand on the  SEASON/DAYNITE level.  Only if you link them only at the ANNUAL level you would need to define COM_FR for clinker, as explained before.

[Xin Wang]

Really appreciate your help, and I will try that! Thank you!