I just have a doubt on the accounting of emissions in TIMES.
I have a COMEMI table that specifies for each fuel the quantity of CO2 emitted per PJ of fuel combusted. I have a coal plant that consumes 2.5 PJ of coal and 1PJ of coal emits 100 ktCO2 according my COMEMI table. Let's assume that a new coal plant captures 100% of the emitted CO2 so that the process does not emit CO2 anymore. What is the right implementation if I maintain the COMEMI table activated? (see attachment)
To ask the question differently, does TIMES doublecount emissions of fuel combustion and process activity?
08-04-2021, 07:41 PM (This post was last modified: 08-04-2021, 07:55 PM by Antti-L.)
No, TIMES should not be double-counting emissions, I think it is the modeller who may make such mistake.
There was no COM_EMI table attached. But I assume you may be defining CO2 emissions for all processes consuming ELCCOA in such a table, is that correct? If that is so, for a CCS technology you must then either define the captured amount as a negative emission somewhere, or replace the gross emission factor by an appropriate net factor. I guess the most robust place for a accounting the captured emission would be the CO2 storage technology, because one should not be credited for it unless it is actually sequestered somewhere. But you can of course also override the COM_EMI emission factor by defining a FLO_EMIS corresponding to the uncaptured emissions on the ELCCOA input flow.
However, if you use ENV_ACT, you are defining another source of emissions for the same plant, caused by the process activity (amount of electricity generated?). Then you will have CO2 emissions from two sources: those caused by the carbon oxidised when coal is combusted, and another stream of CO2 (captured), from the process activity. It is up to you to decide whether you want to model it that way, but then you would need to embed the process efficiency in the (negative) emission factor corresponding to the captured amount.
Thank you Antti... I think this is not clear at all for every modelers regarding the model I retrieved... Has there been a change in TIMES in relation to this in the past years?
To answer your question, yes I have A COMEMI table declared in the BASE files of the model. So if I understand correctly, I need to define the residual CO2 emitted by my capture plant with a FLO_EMIS and not and ENV_ACT?
I have another question though. If I declare a new COMEMI table in a SubRES file, will this new table erase and replace the initial one for all the model, or will this table be activated only for the commodities that are declared in the dedicated SubRES?
No, there have been no changes in TIMES in relation to this matter in the past years.
Concerning COM_EMI, it is a VEDA feature and so not my expertise, but it is translated into the TIMES VDA_EMCB parameter. And the emission factors defined by it can be overridden in TIMES by defining a FLO_EMIS for the same input flow.
Anyway, as I tried to explain, you have many options, one of which is to define the residual CO2 emitted by the capture plant with a FLO_EMIS. But as I tried to explain, you can also use an ENV_ACT, but then, if you have already accounted the gross emissions by using VDA_EMCB, you must define the captured amount with a negative emission factor, and embed the process efficiency into that emission factor. I find myself defining the emission factors directly on the input flows much more transparent.
I think I'd better leave the COMEMI table question to the VEDA experts.
Thanks, but can you confirm whether the original issue "does TIMES doublecount emissions of fuel combustion and process activity?" has been resolved, or do you see some kind of double-counting?
I cannot confirm it now cause my machine is in maintenance but I provided a file containing 3 different manners to model CCS emissions (whether with COMEMI, FLO_EMIS or FLO_EFF). I will let you know the results.
I have made some tests with 3 fictive CCS processes so cheap that the model chooses them so I can observe the output flows (please see the FI tables attached). For each case, I maintained the COMEMI table activated as it is related to many other processes of the rest of the model.
For the first case with the residual CO2 amount fixed by FLO_EMIS, the aforementioned attribute does not "override" the CO2 emitted through the emission factor of the COMEMI table. It sums the CO2 from coal combustion and the residual CO2. Thus, I am not sure what you really meant by "overriding the COMEMI emission factor by defining a FLO_EMIS". Did you mean that the FLO_EMIS value replaces the COMEMI emission factor? If yes, my test does not confirm that.
For the second test with FLO_EFF defining the residual CO2 emissions of capture (instead of FLO_EMIS), same conclusions.
As expected for the third case, it works fine: the amount of CO2 captured is deducted from the CO2 emitted through combustion of coal as defined in the COMEMI table. However, I find this modeling really not transparent and would prefer the FLO_EMIS attribute to override the emissions factor of COMEMI.
If these COMEMI tables are so untrustworthy, I am thinking of removing them all from the VT files of my global model and replace them by creating a SubRES defining FLO_EMIS values for each region of each fuel like proposed in this post https://forum.kanors-emr.org/showthread.php?tid=753&highlight=discourage. Do you think it makes sense?
I ran a final case for which I declare a PRCCOMEMI table attributing zero CO2 emissions from coal combustion whilst declaring the residual CO2 emissions under the ENV_ACT attribute. This works fine and it may be the easiest way to proceed.
13-04-2021, 06:18 PM (This post was last modified: 13-04-2021, 06:20 PM by Antti-L.)
I said that a FLO_EMIS defined on the input fuel would override the COMEMI (VDA_EMCB) emission factor. I am sorry but I cannot see you defining any FLO_EMIS on the input fuel. In fact, it seems the FLO_EMIS you are defining in Case 1) is missing the source Commgrp altogether. Maybe you could post a picture from browse and show the FLO_EMIS, with all dimensions expanded? Or post the *.DD file (zipped)?
So, for the first case, I think your conclusion is invalid. You have not defined any FLO_EMIS on the input fuel, have you?
For the second case, again I think your conclusion is invalid. You have not defined any FLO_EFF on the input fuel.
Besides, I just tested it myself, and fully verified that a FLO_EMIS defined on the process input flow does indeed unequivocally override (replace) the VDA_EMCB emission factor. So, I am very puzzled: How could you possible arrive at your strange conclusions?
Defining FLO_EMIS requires that you specify the source flow of the emission. Amit defined it correctly in his example, but you don't define any source. FLO_EMIS parameter without specifying the source flow is useless and will be just ignored...
The initial problem is solved indeed, but generated a new one. Here attached is the TFM_INS I declared with null FLO_EMIS for processes named EZ* and for the ELCCO2N commodity. Do you have an idea why this rule is also applied to industrial processes whereas they are named in the form of I* and generate INDCO2N? And how can I avoid this?
It seems that the filters stated by Pset_PN, Pset_CI and Pset_CO do not work...
