Parameter NCAP_FSUB

[janis]

I am trying to understand how the parameter NCAP_FSUB functions. According to the manual, it represents the subsidy per unit of installed capacity (or a fixed subsidy associated with the total installed capacity, VAR_NCAP + NCAP_PASTI). Essentially, it acts like a capacity payment for power plants.
I assume that by specifying the NCAP_FSUB parameter for certain years, the results should reflect corresponding values for the Cost_Fixx (CST_FIXX) parameter in those same years — calculated as the total installed capacity multiplied by the subsidy.
However, in the results, I only observe values for the Cost_Fixx (CST_FIXX) parameter starting from the year when new capacity investments are made. These values remain constant until the end of the technology's lifetime.

Thanks!

[Antti-L]

Cost_Fixx indeed reports the annual fixed taxes and subsidies, whereas Cost_Invx reports investment taxes and subsidies.  All the Cost* reporting attributes are nonetheless annualized costs/taxes/subsidies, and therefore also the investment taxes and subsidies are reported in terms of annualized amounts until the end of the technology's lifetime, quite in the same ways as Cost_Inv is reported in terms of annualized costs.

> I assume that by specifying the NCAP_FSUB parameter for certain years, the results should reflect corresponding values for the Cost_Fixx (CST_FIXX) parameter in those same years — calculated as the total installed capacity multiplied by the subsidy.

Indeed, NCAP_FSUB defines a fixed annual subsidy (currency units per unit of installed capacity). It is applied to all vintages that have such a parameter defined. And because it is a fixed subsidy (i.e. an annual subsidy per unit of capacity), the results reflect those parameters in the Cost_Fixx reporting parameter for all those vintages, over their full lifetime.

> However, in the results, I only observe values for the Cost_Fixx (CST_FIXX) parameter starting from the year when new capacity investments are made. These values remain constant until the end of the technology's lifetime.

I am not sure what you think is wrong there, can you clarify?  Do you mean that you have defined NCAP_FSUB also for existing capacity vintages, but you don't see the subsidies reported?  Are you then perhaps using an interpolation option 5?  Anyway, for each capacity vintage, the subsidy should indeed remain constant until the end of the lifetime, unless you apply a SHAPE or MULTI for it, because NCAP_FSUB represents a fixed subsidy. So, although the term "fixed" in general refers to expenses or subsidies that remain constant over time, in TIMES you can also shape them by age or period.

[janis]

Thank you, Antti, for the explanation! I conducted the test and obtained results exactly as you described: "Essentially, for each capacity vintage, the subsidy should remain constant until the end of its lifetime unless a SHAPE is applied, as NCAP_FSUB represents a fixed subsidy."

In other words, each capacity vintage receives the subsidy specified in the year the capacity appears in the results, and this subsidy remains in place until the end of the vintage’s lifetime. Please refer to the example below.

However, I’m not entirely sure what you mean by, "the investment taxes and subsidies are reported in terms of annualized amounts until the end of the technology's lifetime, similar to how Cost_Inv is reported as annualized costs." From my results, it seems that Cost_Fixx is simply the product of the capacity vintage and the subsidy.


Specified subsidy:

Results:

Actually, I would like to incorporate capacity payments for the power plant into the model. This includes payments for the total available capacity of the plant in specific year, with the payment amount differing from year to year. Thanks a lot for the suggestions!

[Antti-L]

> From my results, it seems that Cost_Fixx is simply the product of the capacity vintage and the subsidy.

Yes, correct.  Cost_Fixx reports the fixed taxes and subsidies, not investment taxes or subsidies.

> I’m not entirely sure what you mean by, "the investment taxes and subsidies are reported in terms of annualized amounts until the end of the technology's lifetime, similar to how Cost_Inv is reported as annualized costs.

As I mentioned, Cost_Invx reports the investment taxes and subsidies, in annualized terms, so that's what I meant.

> I would like to incorporate capacity payments for the power plant into the model. This includes payments for the total available capacity of the plant in specific year, with the payment amount differing from year to year.

You would have at least the following options (of which I think the first one would be easiest):

  • Define NCAP_COM(r,v,p,c,'OUT')=1 (using a dummy commodity c) and define the subsidy on that flow using FLO_SUB.
  • Define NCAP_FSUB = X, but with a MULTI multiplier corresponding to how the payment X is differing from year to year.

While NCAP_FSUBX + SHAPE shapes NCAP_FSUB by age, NCAP_FSUBM + MULTI can be used for shaping it by period.

[Antti-L]

Note that fixed costs, taxes and subsidies are vintage-specific, because fixed costs are often strongly tied to the physical characteristics of the installation, and those characteristics are not changing by period, like they are changing by the vintage of the installations, due to technology development and learning.

I understand, however, that in Markal fixed costs may have been period-specific instead of being vintage-specific.  While such a handling of fixed costs neither corresponds to what is usually meant by fixed costs nor reflects the reality well, I can see that for certain types of capacity-related expenditures or revenues it can be very useful.  Therefore, it could be useful to dedicate one dummy commodity (say 'CAPCOST', with LimType 'N') for these types of costs/taxes/subsidies. You could then always use NCAP_COM(r,v,p,'CAPCOST','OUT') with a FLO_COST/TAX/SUB for any process that should have capacity-related costs/taxes/subsidies period-specific instead of vintage-specidfic.

[janis]

Thanks to Antti for the suggestion to use NCAP_COM in the capacity payment representation in model. I implemented it, and it's working as expected!

Also, I appreciate the clarification on the distinction between period-specific and vintage-specific costs.

A quick question: regarding the dummy commodity with LimType 'N'—should it have been defined in the case of capacity payment as well?

[Antti-L]

LimType 'N' is just my preference for most dummy commodities, when I want to make sure they don't have unnecessary or possibly even detrimental commodity balances. But you can of course do as you like, and even define the capacity payments for the total commodity production instead of the flow, and thereby using 'LO' would be just fine.