Time-slice

[Samaneh]

Hello,

I have a very basic question. I have 4 time-slices (AM, PM, Peak, and Night) and 9 time periods in my model. The availability factor (AF) of wind power plants is the same for AM, PM, and Peak from 2010 to 2020 (e.g. AF(AM, PM, and Peak)=0.3 for 2010, 2015, and 2020). Is there anyway to show the AF for these 3 time-slices and 3 time periods in one column as one parameter? 

Should it be like AF~AM,PM,Peak~2010,2015,2020=0.3?

Thanks,

Samaneh

[AKanudia]

What you want can be done via a TFM_INS table in one of the transformation files, or a regular scenario file. You can have comma-separated entries in the TimeSlice and Year columns. Make sure that you put a leading <'> character when writing 2010,2015,2020 as it will be interpreted as a single number by default. You can also introduce Year and TimeSlice columns in an FI_T table and use comma-separated entries. But column headers in FI_T don't allow multiple elements for any dimension.

However, you should consider using the very powerful and efficient inheritance and interpolation properties/options of TIMES.  For example, I would make the following declarations:

AF(ANNUAL,2010) = .3

AF(Night,2010) = xyz

If these values change after 2020, then I would make the same declaration for 2020, and again for the new values in the year they apply. AF is interpolated and extrapolated by default, so you don't even have to specify an interpolation rule if you want to hold it constant. This is indicated by the green color in the Years column for NCAP_AF in Attributes Master.

Antti, please correct me if I am wrong, but the ANNUAL values will be inherited by all timeslices that don't have an implicit or explicit value specified. By implicit, I mean if one of the SEASONAL timeslices have a specification, then the WEEKLY and DAYNITE slices of that season will get that value, rather than the ANNUAL one.

[Antti-L]

Amit: You are right, ANNUAL values will, in general, be inherited by all timeslices that don't have an implicit or explicit value specified.

However, according to the original design of TIMES, specifying AF(Night,2010) = xyz would prevent the ANNUAL values from being inherited, assuming that Night is a timeslice at the process timeslice level. But if the nightly AF is smaller than 0.3 you could accomplish the same by specifying 

AF(ANNUAL,2010) = .3
AFS(Night,2010) = xyz

The exception to the inheritance rule applies only to the the NCAP_AF parameter. For other levelized timeslice-dependent parameters, ANNUAL values are inherited by all timeslices that don't have an implicit or explicit value specified, even if some values are specified at the process timeslice level.

I don't know what the motivation has been behind this exception in the inheritance rules, and I think I would agree to remove it if such a change would be proposed.

[AKanudia]

Hmmm... I don't remember being a party to this exception (explicitly or implicitly). So, it would be unavailable in the other process timeslices, if we made the declarations that I had proposed?

I would vote for removing this exception.

[Antti-L]

Yes, with your example it would be unavailable in the other process timeslices (assuming xyz>0).

Of course (as usual), there is also an exception to the exception:

As mentioned, if xyz > 0, any inheritance will be prevented. However, if all such AF values specified at the process timeslice level are zero, inheritance is nonetheless activated. This latter exception should be understandable, because if inheritance would be prevented also in that case, the process would not be able to operate in any timeslice, which would hardly be the user intention.

[Samaneh]

Thank you so  much for your help. As you mentioned, I introduced time slices in FI_T table as comma-separated entries and It worked