NCAP_AFC and NCAP_AFA

[Antti-L]

> I am also confused about this point, if I define the NCAP_AFA=0.5, and NCAP_AFC=0.5 for one specific timeslice (like FD), does that mean the activity level <STOCK*CAP2ACT*0.25?

No, because 1) ANNUAL # FD, and 2) your relation is ambiguous concerning timeslices.

> Besides, if we remove the NCAP_AFA defination, does that mean the activity level <STOCK*CAP2ACT*0.5?

If you remove the NCAP_AFA, then you will no longer have an ANNUAL utilization factor, unless you define NCAP_AFS(ANNUAL) or NCAP_AFC(ANNUAL) or NCAP_AFAC.

[[email protected]]

> I am also confused about this point, if I define the NCAP_AFA=0.5, and NCAP_AFC=0.5 for one specific timeslice (like FD), does that mean the activity level

No, because 1) ANNUAL # FD, and 2) your relation is ambiguous concerning timeslices.

> Besides, if we remove the NCAP_AFA defination, does that mean the activity level

If you remove the NCAP_AFA, then you will no longer have an ANNUAL utilization factor, unless you define NCAP_AFS(ANNUAL) or NCAP_AFC(ANNUAL) or NCAP_AFAC.

Hi Antti,

Thank you for the reply. My question is: 


I have 96 timeslices (4 seasons × 24 hours) and want to model the variability of wind energy.

For the 0 a.m. timeslice in winter (which I define as WH0), if the availability factor (AF) is 0.25, with a capacity (Stock) of 1 GW and a CAP2ACT conversion of 31.54 PJ/GW, does this mean the activity level for WH0 (in PJ) should be constrained by:

• 1 GW×31.54 PJ/GW×0.25
  

or

• 1 GW×31.54 PJ/GW×0.25/96 (assuming the time fraction is evenly distributed across all timeslices)?
  

I want to limit it as the second option (because the whole year generation is around 1 GW×31.54 PJ/GW×0.25 rather than one time-slice), but my results indicate that the VAR_ACT far exceeds 1×31.54×0.25/96 and thus the annual sum is beyond 1×31.54×0.25.

How can I set the up limit for this timeslice (just as an example)? Should the AF be adjusted to 0.25/96?


Best,
Xiao

[Antti-L]

> How can I set the up limit for this timeslice (just as an example)?

If NCAP_AF(ANNUAL)=0.25, PRC_RESID=1 and PRC_ACTFLO=31.54 then the limit would indeed be 1×31.54×0.25/96 for each timeslice (assuming the G_YRFR fraction is evenly distributed across all 96 timeslices).

> Should the AF be adjusted to 0.25/96?

No.

> the annual sum is beyond 1×31.54×0.25.

Surely it shouldn't be so. Do you have any evidence supporting that statement? Maybe you could put the model and the results demonstrating your issue available on Veda Online? Are they already there?

[[email protected]]

> How can I set the up limit for this timeslice (just as an example)?

If NCAP_AF(ANNUAL)=0.25, PRC_RESID=1 and PRC_ACTFLO=31.54 then the limit would indeed be 1×31.54×0.25/96 for each timeslice (assuming the G_YRFR fraction is evenly distributed across all 96 timeslices).

> Should the AF be adjusted to 0.25/96?

No.

> the annual sum is beyond 1×31.54×0.25.

Surely it shouldn't be so. Do you have any evidence supporting that statement?  Maybe you could put the model and the results demonstrating your issue available on Veda Online? Are they already there?

Hi Antti,

It is a bit confused but I put my questions here.

We have imposed AFA=0.6, CAP2ACT=31.54, and PRC_RESID=0.011GW for the heat pump energy technology (R_ES-SH-AP_HET) in the AL region, which limits the activity level to a maximum of 0.6 × 31.54 × 0.011 = 0.21 PJ per year. The results show the VAR_Act for AL is 0.09 PJ per year, aligning with this constraint. However:

1. The system tends to use IMPDMZ to meet the output demand (R_ES-AP-SpHeat), even though there is remaining capacity for those technologies (like R_ES-SH-AP_HET).
   


2. I defined input shares for RSDAHT and RSDELC as 0.67 and 0.33, respectively (upper limit type). Despite this, the results indicate that only RSDAHT was used. Why is this the case?
   


3. Although the activity level is DAYNITE, it seems activity level during all winter timeslice (because there are demands only in winter) are same, Why?
   

I opened the model in VO, 
CAN_TIMES_v1[XIAO-VACANT]

created by: seanli12354

Can you help me?

Best,
Xiao

[Antti-L]

> 1. The system tends to use IMPDMZ to meet the output demand (R_ES-AP-SpHeat), even though there is remaining capacity for those technologies (like R_ES-SH-AP_HET).

Not true.  There is no remaining capacity for R_ES-SH-AP_HET, the capacity is fully utilized (100%) over the winter timeslices, even though the resulting ANNUAL utilization factor is only 0.2486, i.e. much lower than your limit.  In addition, I can see that the price of e.g. ELC and RSDELC is astronomical in 2020, and so there would be no remaining capacity for any technology using RSDELC either.

> 2. I defined input shares for RSDAHT and RSDELC as 0.67 and 0.33, respectively (upper limit type). Despite this, the results indicate that only RSDAHT was used. Why is this the case?

You also defined the process vintaged, and you defined an interpolation option 5 for the FLO_SHAR(AL,2020, R_ES-SH-AP_HET,...,UP) parameters. Therefore, the share bound is not imposed for this process at all (because its vintage is 2019). As an easy remedy, I would simply suggest to remove the Vintaged qualification for all those existing technologies.

> 3. Although the activity level is DAYNITE, it seems activity level during all winter timeslice (because there are demands only in winter) are same, Why?

We have discussed this matter before (see Availability about Space heating tech).  You have deliberately chosen to model the process at the DAYNITE level, and thus you should know doing so requires that you enforce an appropriate load profile for the process, unless you wish to let the model freely optimize the load profiles of demand technologies, like with power plants. Now you have allowed such free optimization, and the optimal solution appears to be a flat level during the winter timeslices.

[[email protected]]

> 1. The system tends to use IMPDMZ to meet the output demand (R_ES-AP-SpHeat), even though there is remaining capacity for those technologies (like R_ES-SH-AP_HET).

Not true.  There is no remaining capacity for R_ES-SH-AP_HET, the capacity is fully utilized (100%) over the winter timeslices, even though the resulting ANNUAL utilization factor is only 0.2486, i.e. much lower than your limit.  In addition, I can see that the price of e.g. ELC and RSDELC is astronomical in 2020, and so there would be no remaining capacity for any technology using RSDELC either.

> 2. I defined input shares for RSDAHT and RSDELC as 0.67 and 0.33, respectively (upper limit type). Despite this, the results indicate that only RSDAHT was used. Why is this the case?

You also defined the process vintaged, and you defined an interpolation option 5 for the FLO_SHAR(AL,2020, R_ES-SH-AP_HET,...,UP) parameters. Therefore, the share bound is not imposed for this process at all (because its vintage is 2019). As an easy remedy, I would simply suggest to remove the Vintaged qualification for all those existing technologies.

> 3. Although the activity level is DAYNITE, it seems activity level during all winter timeslice (because there are demands only in winter) are same, Why?

We have discussed this matter before (see Availability about Space heating tech).  You have deliberately chosen to model the process at the DAYNITE level, and thus you should know doing so requires that you enforce an appropriate load profile for the process, unless you wish to let the model freely optimize the load profiles of demand technologies, like with power plants. Now you have allowed such free optimization, and the optimal solution appears to be a flat level during the winter timeslices.

Hi Antti,

So nice to have your help. For the third point:

Yes, I defined it as a DAYNITE tech, and I imposed the demand fractions for all timeslices. However, I only defined AFA (to reduce the computational burden, I guess it could). 

Now I want to change the flat output, thus to address more output in higher demand time while lower output in lower demand period, therefore to utilize it 100% and also meet the demand requirements. so I need to define the AF for each timeslice? I am still confused. [I just want to model the day nite variations and also avoid IMPDMD]

Best,
Xiao

[Antti-L]

> I defined it as a DAYNITE tech, and I imposed the demand fractions for all timeslices.

But as I already explained, doing that will not impose the load profile for the technologies, because you chose the DAYNITE level approach. As I explained, it would impose it automatically when using the ANNUAL approach. So, in your case your COM_FR only defines the profile for the total demand of R_ES-AP-SpHeat.

> However, I only defined AFA (to reduce the computational burden, I guess it could).

It does not reduce the computational burden, because the DAYNITE level EQ_CAPACT equations are always generated for any regular DAYNITE level processes.  So, it is in fact slightly increasing the burden, by adding one additional CAPACT equation (ANNUAL) for each technology, each vintage, and each period.

> so I need to define the AF for each timeslice?

That would be one way of imposing a load profile in your DAYNITE case, and undoubtedly with the least additional burden (because in your case the EQ_CAPACT equations are there anyway for each timeslice).

> I just want to model the day nite variations and also avoid IMPDMD

I think you would also need to improve your model calibration, because as far as I can see (looking just at R_ES-AP-SpHeat), the existing capacity in 2020 is not sufficient for satisfying the peak demand, even if every existing technology would operate at full load (if I got it right, the deficit seems about 11%).

[[email protected]]

Hi Antti,

Thank you, I adjusted all of this tech to be ANNUAL and only AFA, DEM_FRA imposed without AF.

One more question: is it necessary to impose the O&M cost for the existing tech in BY?

Best,
Xiao

[[email protected]]

Hi Antti,

Thank you, I adjusted all of this tech to be ANNUAL and only AFA, DEM_FRA imposed without AF.

One more question: is it necessary to impose the O&M cost for the existing tech in BY?

Best,
Xiao

Sorry if any confusion. I have this additional question because I noticed DEMO12 only impose O&M for existing mining techs, electricity generation techs, and transportation techs. So in line with the TIMES-Demo12 setup, we applied O&M costs to the mining, electricity generation, and transport sectors, excluding navigation, aviation, and rail. Does that make sense? 

Best,
Xiao