how to define the capacity or stock for storage?

[guozhi1305]

Dear all, 
    In my model, there are 98 (4*24 hours)timeslice and two kinds of storage, one is the pumped hydro energy storage(PHS) with unit GW, and the other battery with unit GW. Now in the base year, I want to define the capacity of these two storage as shown the attached file,  while the VAR_CAP of PHS from veda-be is 454,not 31.49 GW as set in the FI_T. what is wrong？I have tried many many times, it still occurs something wrong .  If available hours is 1300 for PHS, how to set the NCAP_AFC? and if the battery storage discharges two hours per day, how to set the them correctly? Is there some demos for the capacity of the storage?  If it is possible, could you please correct the setting in the attached file. Looking forward to your reply.
    Thank you very much.
     Best regards
Guo Zhi

[Antti-L]

Well, for me at least it is not quite clear what you want, and what you mean by "VAR_CAP of PHS from veda-be is 454,not 31.49 GW as set in the FI_T". Some questions below to start with:

1) Could you post the full VAR_CAP results of the ELCSTSPHS01 process from VEDA-BE, with all dimensions expanded, to see where that 454 is coming from?
2) The file is a Subres file, and therefore primarily meant for new technologies. Do you want to prohibit new investments into these storage processes?
3) You have defined PRC_RESID = 31.49, but no values for any other years. That would result in the existing capacity reducing linearly to zero in 30 years, is that what you want?
4) You have defined ELCS and ELCW as the input commodities to the storage processes (electricity from solar and wind power). And the output is ELC. But you have defined all these three commodities to be tracked only at the ANNUAL level.  Is that indeed what you want, that ELCS, ELCW and ELC are modeled on the ANNUAL level?
5) You have defined NCAP_AFC(ELC,ANNUAL)=0.1369863, corresponding to 1200 hours' peak utilization time. Capacity factor would be the peak utilization time divided by 8760. But in your question you talk about "available hours" being 1300 hours, what does that mean?  Are you referring to the peak utilization time there as well? And if so, why 1300 hours and not the 1200 hours as in your Excel?

[guozhi1305]

Well, for me at least it is not quite clear what you want, and what you mean by "VAR_CAP of PHS from veda-be is 454,not 31.49 GW as set in the FI_T". Some questions below to start with:

1) Could you post the full VAR_CAP results of the ELCSTSPHS01 process from VEDA-BE, with all dimensions expanded, to see where that 454 is coming from?
2) The file is a Subres file, and therefore primarily meant for new technologies. Do you want to prohibit new investments into these storage processes?
3) You have defined PRC_RESID = 31.49, but no values for any other years. That would result in the existing capacity reducing linearly to zero in 30 years, is that what you want?
4) You have defined ELCS and ELCW as the input commodities to the storage processes (electricity from solar and wind power). And the output is ELC. But you have defined all these three commodities to be tracked only at the ANNUAL level.  Is that indeed what you want, that ELCS, ELCW and ELC are modeled on the ANNUAL level?
5) You have defined NCAP_AFC(ELC,ANNUAL)=0.1369863, corresponding to 1200 hours' peak utilization time. Capacity factor would be the peak utilization time divided by 8760. But in your question you talk about "available hours" being 1300 hours, what does that mean?  Are you referring to the peak utilization time there as well? And if so, why 1300 hours and not the 1200 hours as in your Excel?

Dear Antti-L,
    Thanks very much for your detailed reply. 

1) Could you post the full VAR_CAP results of the ELCSTSPHS01 process from VEDA-BE, with all dimensions expanded, to see where that 454 is coming from?
--Thanks for your remind. I checked the VEDA-BE results, 454 is the VAR_CAP(453.52)=PRC_RESID(31.49)+VAR_NCAP(422.03)

2) The file is a Subres file, and therefore primarily meant for new technologies. Do you want to prohibit new investments into these storage processes？
---2020 is the base year, my original intention was to define energy storage technology for the base year in Subres file, and then I will  define another new process for PHS with capacity, such as ELCSTSPHS02.

3) You have defined PRC_RESID = 31.49, but no values for any other years. That would result in the existing capacity reducing linearly to zero in 30 years, is that what you want?
--For this process, yes. And I will define another process ELCSTSPHS02 with values for other years.

4) You have defined ELCS and ELCW as the input commodities to the storage processes (electricity from solar and wind power). And the output is ELC. But you have defined all these three commodities to be tracked only at the ANNUAL level.  Is that indeed what you want, that ELCS, ELCW and ELC are modeled on the ANNUAL level?
--No, I want to defineall these commodities to be tracked at the daynite level, maybe the definition method I used is not correct. Also, I want the ELCS,ELCW and ELC are modelled on the daynite level. 

5) You have defined NCAP_AFC(ELC,ANNUAL)=0.1369863, corresponding to 1200 hours' peak utilization time. Capacity factor would be the peak utilization time divided by 8760. But in your question you talk about "available hours" being 1300 hours, what does that mean?  Are you referring to the peak utilization time there as well? And if so, why 1300 hours and not the 1200 hours as in your Excel?
---Sorry, I made the mistake, it should be 1200 hours. Here, NCAP_AFC(ELC,ANNUAL)=0.1369863 can only be defined on the annual level?

   Thanks very much for your reply, I tried another way, ELCSTSPHS01 is defined in the BY, and ELCSTSPHS02 is defined in the Subres file as shown in the attached file. And then, the VAR_cap for ELCSTSPHS01 and ELCSTSPHS02 from veda-be equal to the PRC_RESID as I set in the BY and Subres file. If it is possible, could you please help me to check the setting of the stroage, especially for the NCAP_AFC? I want to use the available factor( maybe the NCAP_AFC) to limit the output of the storage, that is VAR_CAP*AF*CAP2ACT > the total output, and the output of each timeslice is lower than VAR_CAP*CAP2ACT/96. And I also want to know more about how to model ELCS, ELCW and ELC on the daynite level.
    Thank you very much.
     Best regards.
Guo Zhi

[Antti-L]

1) > I checked the VEDA-BE results, 454 is the VAR_CAP(453.52)=PRC_RESID(31.49)+VAR_NCAP(422.03)

Ok, so the existing capacity was correct, and the model just wanted to invest into additional capacity.  So, what was the problem here?  I already asked you whether you would wanted to prohibit new investments, but I could not see you giving any answer. If, however, that is the case, you can just change your START year to 2100 (or whatever), to prohibit those new investments until that year.

2) I could not see any answer to my question. See above.

3) Ok, fine.

4) Ok, if you want the commodities to be modeled on the DAYNITE level, please put DAYNITE into the CTSLvl column in the FI_Comm table.

5) Ok, so then your NCAP_AFC(ELC,ANNUAL) = 0.1369863 is fine (for 1200 hours). But you would need also NCAP_AFC(ELC,DAYNITE) = 1  to limit the limit the output of the storage, that is VAR_CAP*AF*CAP2ACT > the total output, and the output of each timeslice is lower than VAR_CAP*CAP2ACT/96. You could add a column NCAP_AFC~DAYNITE for this.

The comments above in general apply also to the battery storage, but to model a battery with a 2 hour storage energy capacity would require defining NCAP_AFC(ACT,DAYNITE) = 2/24 / STG_EFF. You can add a CommGrp column for specifying the 'ACT' for it. And in that case it does not make sense to define the storage as 'STS', but 'STG' would make more sense (there is no basically no seasonal storage capability with a 2h storage).

[guozhi1305]

1) > I checked the VEDA-BE results, 454 is the VAR_CAP(453.52)=PRC_RESID(31.49)+VAR_NCAP(422.03)

Ok, so the existing capacity was correct, and the model just wanted to invest into additional capacity.  So, what was the problem here?  I already asked you whether you would wanted to prohibit new investments, but I could not see you giving any answer. If, however, that is the case, you can just change your START year to 2100 (or whatever), to prohibit those new investments until that year.

2) I could not see any answer to my question. See above.

3) Ok, fine.

4) Ok, if you want the commodities to be modeled on the DAYNITE level, please put DAYNITE into the CTSLvl column in the FI_Comm table.

5) Ok, so then your NCAP_AFC(ELC,ANNUAL) = 0.1369863 is fine (for 1200 hours). But you would need also NCAP_AFC(ELC,DAYNITE) = 1  to limit the limit the output of the storage, that is VAR_CAP*AF*CAP2ACT > the total output, and the output of each timeslice is lower than VAR_CAP*CAP2ACT/96. You could add a column NCAP_AFC~DAYNITE for this.

The comments above in general apply also to the battery storage, but to model a battery with a 2 hour storage energy capacity would require defining NCAP_AFC(ACT,DAYNITE) = 2/24 / STG_EFF. You can add a CommGrp column for specifying the 'ACT' for it. And in that case it does not make sense to define the storage as 'STS', but 'STG' would make more sense (there is no basically no seasonal storage capability with a 2h storage).

Dear Antti-L,
       Thanks for your reply, Your help is very important to me. Thank you very much. And I am sorry for the second one because I am unclear about "prohibit new investment". What does it mean? Does it conflict with "invest into additional capacity"? And now, I have other problems needing your help. In China, coal power will undertake the task of peak shaving, meaning that the existing coal power requires flexibility transformation. Therefore, I want to model the timeslice level for coal power SEASON in the base year, and DAYNITE after transformed with additional cost(shown in INVCOST~2030) in 2030. I tried this way as shown in the attached file, TSLVL(Coal power, 2020)=SEASON, and TSLVL(Coal power, 2030)=DAYNITE, while the outcome shows that the timeslice in 2030 is still SEASON, not DAYNITE. (1) Is there another way to define the timeslice with different level in different year? (2) And in my model, there are four coal power generation technologies, how to limit the capacity of coal power with flexibility transformation to be a certain level, such as 2 GW. And the same for CCS technology, in 2030, the existing coal power will install CCS technology with additional cost (as shown in FIXOM~2030 and VAROM~2030), (3) Can I set the cost and emission of coal power with CCS as in the attached file? Looking forward to your reply.
     Best regards~
Guo Zhi

[Antti-L]

> I am unclear about "prohibit new investment". What does it mean?

It means to prevent investments into new capacity. I was unclear about why you said that there was some problem with the capacity, even though the existing capacity turned out to be exactly correct.

> (1) Is there another way to define the timeslice with different level in different year?

In your example, you are defining fixed NCAP_AF factors for the coal technologies, which means that they would already be effectively operating on a SEASON basis (flat activity in each season), even if defined to operate on the DAYNITE level.  So, in the simplest case, you could just define the technologies on the DAYNITE level, and use interpolation option 4 for the NCAP_AF factors (meaning that in 2030 they would become flexible). And the CCS capability could also be added starting from 2030, simply by adjusting the emission factors, efficiencies and variable costs between 2025 and 2030. In such a forced retrofit case, the additional investment costs and fixed O&M costs would not affect the solution and could therefore basically be ignored. Operation on a SEASON basis could be also well simulated with e.g. ramping rate constraints or user constraints, and so there are several options for modeling such a flexibility change for individual technologies. However, assuming you want a more elaborate solution, see my additional remarks about using endogenous retrofits and/or lifetime extensions below.

> (2) there are four coal power generation technologies, how to limit the capacity of coal power with flexibility transformation to be a certain level, such as 2 GW.

That would be easy with a simple UC constraint.

> in 2030, the existing coal power will install CCS technology with additional cost (as shown in FIXOM~2030 and VAROM~2030)

I assume you mean by this that the existing capacity should not operate without CCS after 2029. However, it is still unclear what you exactly mean by this:
A) do you want to force all existing coal power to be unconditionally retrofitted with CCS
B) do you want to give them the option to be either retrofitted with CCS or retired
C) do you want to give them the option to be either refurbished with CCS (with a life-time extension) or retired.

Nonetheless, looking at the investment costs you have specified in the file for 2030, I can see they are even higher than the original full investment costs. That would seem strange if you mean only investing into a retrofit for the remaining lifetime of the existing capacity, with the linear phase-out of you have defined.  Therefore, it seems to me that a lifetime extension option (with flexible DAYNITE operation and CCS) would probably suit best for your purpose. You could force retirement of the existing capacity in 2030 (by using CAP_BND), but allow it to be replaced by the lifetime extension option (either optionally or unconditionally), with any desired investment costs, fixed O&M costs, variable costs and other process characteristics.

The documentation for the TIMES Retrofits and Lifetime extensions facility is found here.

There are some examples linked on the VEDA Forum, see e.g. the following thread: CCS retrofit file documentation

[guozhi1305]

 Dear Antti,
        
    Thank you for your detailed reply. I am sorry I just noticed it.

>  I was unclear about why you said that there was some problem with the capacity, even though the existing capacity turned out to be exactly correct.
 ----Yes, there is no problem about the capacity, thank you very much.

> (1)   So, in the simplest case, you could just define the technologies on the DAYNITE level, and use interpolation option 4 for the NCAP_AF factors (meaning that in 2030 they would become flexible). And the CCS capability could also be added starting from 2030, simply by adjusting the emission factors, efficiencies and variable costs between 2025 and 2030. In such a forced retrofit case, the additional investment costs and fixed O&M costs would not affect the solution and could therefore basically be ignored. Operation on a SEASON basis could be also well simulated with e.g. ramping rate constraints or user constraints, and so there are several options for modeling such a flexibility change for individual technologies. However, assuming you want a more elaborate solution, see my additional remarks about using endogenous retrofits and/or lifetime extensions below.
 
-----Yes, these days, I tried to define the coal power on the DAYNITE level, and use the AF~UP and AF~LO in 2030 to adjust the flexibility. " In such a forced retrofit case, the additional investment costs and fixed O&M costs would not affect the solution and could therefore basically be ignored."  which means the additional cost due to CCS will not work?

> (2) there are four coal power generation technologies, how to limit the capacity of coal power with flexibility transformation to be a certain level, such as 2 GW. ---That would be easy with a simple UC constraint.
 
----Thank you very much, I will try.

> in 2030, the existing coal power will install CCS technology with additional cost (as shown in FIXOM~2030 and VAROM~2030)
I assume you mean by this that the existing capacity should not operate without CCS after 2029. However, it is still unclear what you exactly mean by this:
A) do you want to force all existing coal power to be unconditionally retrofitted with CCS
B) do you want to give them the option to be either retrofitted with CCS or retired
C) do you want to give them the option to be either refurbished with CCS (with a life-time extension) or retired.

----Actually,  I want to express that the existing coal power in their lifetime will install CCS technology or not depending on the emission reduction target. It also means that how much coal-fired power needs to be installed with CCS technology to achieve carbon reduction targets? and the remaining units will operate normally throughout their lifetime.


Nonetheless, looking at the investment costs you have specified in the file for 2030, I can see they are even higher than the original full investment costs. That would seem strange if you mean only investing into a retrofit for the remaining lifetime of the existing capacity, with the linear phase-out of you have defined.  Therefore, it seems to me that a lifetime extension option (with flexible DAYNITE operation and CCS) would probably suit best for your purpose. You could force retirement of the existing capacity in 2030 (by using CAP_BND), but allow it to be replaced by the lifetime extension option (either optionally or unconditionally), with any desired investment costs, fixed O&M costs, variable costs and other process characteristics.

The documentation for the TIMES Retrofits and Lifetime extensions facility is found here.

There are some examples linked on the VEDA Forum, see e.g. the following thread: CCS retrofit file documentation

----Thank you.