We have modeled a H2-powered fuel cell technology as a combined heat and power (CHP) process in a TIMES case. We want it to produce both- heat and power and specified the attributes as in the attached scenario file.
This technology should exhibit an electrical efficiency of ~52% and heat efficiency of ~37%. Hence a combined efficiency of ~89%. However, we are only getting an efficiency of 76% (see figure). I feel there is a problem with the NCAP_CHPR~FX bound but not sure how to correct it (or if it is even the correct attribute to use in this case; this specification was created by another modeler). I went through TIMES Part II documentation (Section 4 on CHP) and not sure what value to use for CEH (if at all).
Could you please help me with deciding the correct attributes and values to use? Thank you.
I made the changes suggested in both tabs of your file and additionally changed the CAP2ACT to 1 because we changed Tcap from GW to PJ/a. After making the changes suggested by you, I re-ran the NetZero case and TIMES is not deploying the technology at all. Increasing efficiency from 76% to 89% should make the technology more attractive, not less, right?
Hence, I would need some more explanation on the file you shared.
1. I was meaning to define efficiencies in terms of the input itself - 1PJ input gives 0.52 ELC output and 0.37 PJ HET output so I don't understand the meaning of your text bubble. However, I did copy the exact line of ECHPSOFC-OK technology that you defined.
2. What is LTHEAT Ctype?
(As an aside: I made all the COMM tab changes to my SysSettings file and H2 base scenario file where the commodities were originally defined.)
22-03-2024, 10:42 PM (This post was last modified: 22-03-2024, 11:36 PM by Antti-L.)
>After making the changes suggested by you, I re-ran the NetZero case and TIMES is not deploying the technology at all. Increasing efficiency from 76% to 89% should make the technology more attractive, not less, right?
Yes, increasing the efficiency from 76% to 89% should make the technology more attractive. Indeed, I checked that the efficiency was correct after my changes.
>additionally changed the CAP2ACT to 1
Ok, then your costs per activity would be much higher. Would that explain your zero deployment?
>I was meaning to define efficiencies in terms of the input itself - 1PJ input gives 0.52 ELC output and 0.37 PJ HET output so I don't understand the meaning of your text bubble.
Sorry, but that makes no sense for CHP ACT_EFF. Have you read the documentation? ACT_EFF for an output flow would be a multiplier for the efficiency, when producing that output flow. The default eff is 1 if no input efficiency is specified. So, your specification would mean that the efficiency is 37% when producing heat and 52% when producing electricity. Can you see that the combined total energy efficiency would not make sense (it would in general be between 37% and 52%, and not 52%+37%)? ACT_EFF for an output would make sense e.g. for a CAR technology producing short and long distance travel. For such a car technology, the fuel consumption can be differentiated for the travel types (short/long), and the combined energy efficiency would make perfect sense. For CHP differentiating the output efficiencies would never work right (because of the internal processing of CHP parameters). As you saw, it did not work as you expected. So, I simply corrected your modeling error, following the documentation. As stated there, only the electrical efficiency should be specified by ACT_EFF for CHP (from the input fuel(s) to the activity).
>What is LTHEAT Ctype?
Low temperature heat. I added the FI_Comm table just for my own testing, you can ignore what was there.
1. Is CAP2ACT=1 not the only correct value to use when Tact = PJ and Tcap=PJ/a?
2. It is proving difficult for me to understand the efficiency concept that you are describing (I will read more about it). However, I ran the specification that you sent (with only electrical efficiency listed).
I made a mistake in interpreting the results before. The run is yielding me an 89% efficiency now, so thanks for that.
Now the only question that remains is using CAP2ACT=1 vs. 31.536. My understanding is that CAP2ACT=31.536 is only used for cases where we go from GW to PJ. And here we go from PJ/a to PJ. So why shouldn't we use CAP2ACT=1?
23-03-2024, 04:33 AM (This post was last modified: 23-03-2024, 04:38 AM by Antti-L.)
Ok, maybe I see now the cause for your confusion, some elaboration about it below:
When I changed your process specification, I also wanted to make a quick test with it. And therefore, I needed to add an entry to the FI_Process table for the new technology to be tested (plus I also needed to add the commodities to a FI_Comm table). At first, I was mistakenly assuming you had the FI_Process table below your FI_T table (thinking it was an active table), and just copied the row for "CHP ECHPSOFC SOFC mimicking MARKAL description", which was there, to a new row for my test process. The units were thus simply copied from that table below your FI_T table, which I however soon realized was a spurious leftover table (apparently not relevant to your question at all).
After this, I realized that the active FI_Process table was actually above the FI_T table, and I moved the new line I had added there. What I did not notice was that you had specified different units there! If you had not included that spurious second FI_Process table, I would not have made that mistake. I am so sorry that I mistakenly copied the PJ/a unit from your table below.
However, note that the units specified in FI_Process tables have no impact on the model: They are just documentation entries. PRC_CAPACT defines the unit conversion between the capacity and the activity, and I did not change that (I didn't even consider changing anything about your units).
Again, I am very sorry if this is the cause for your confusion (caused by your second FI_Process table).
Thank you for the clarification. I understand now. I was unaware that the ~Fl_Process entry is just for documentation - will keep it in mind from now on. Will also make it a practice to grey-out or hide unused tables to avoid confusion in the future.
(23-03-2024, 04:46 AM)Antti-L Wrote: It is only the units (and process description) in ~Fl_Process tables that are for documentation. Otherwise the fields in it are important.
