Conditional branching based on penetration rate

[Ryo Ishida]

Hello all

We are often grateful for the help we receive on this forum. In this case, we would like to ask about the possibility of implementing the following:

”Varying the variable costs placed on the entire power system according to the penetration rate of renewable energy.”
For example, If the penetration of renewable energy is 20%, add 3 JPY/kWh to the electricity commodity (ELC). If the penetration rate of renewable energy is 40%, 4 JPY/kWh will be added to the electricity commodity (ELC). If the penetration rate of renewable energy is 60%, 5 JPY/kWh will be added to the electricity commodity (ELC).

Is it possible to conditionally bifurcate the variable cost depending on the penetration rate of renewable energy in this way?

Thank you in advance！

Ryo

[Antti-L]

Varying variable costs according to the penetration rate of renewable energy is certainly possible, but basically only with convex cost curves in TIMES. Convexity implies that the marginal costs should be monotonically increasing (or non-decreasing) along the increasing penetration rate.

Your description could be interpreted both in a non-convex and in a convex way, although it may in any case be considered a bit ambiguous. For example, one might assume that the costs you have stated should represent average additional variable costs for the total power generation at the stated penetration points (20%, 40%, 60%), and developing linearily between those points. That would indeed result in monotonically non-decreasing marginal costs along with increasing rates, and can thus be modelled. There is no automated mechanism for defining such, but it would be relatively easy by using some dummy variables. In your case, simply a base cost of 3 JPY/kWh for all power plus a cost of 5 JPY/kWh for all renewable generation in excess of a 20% share in total generation would seem to accomplish it, because then (assuming a total generation of 1 GWh) the total variable costs would be 3×1 + 5×0.2 = 4 MJPY at 40%, which is indeed 4 JPY/kWh, and similarly 3×1 + 5×0.4 = 5 MJPY for 60%.

[VictorG]

@Antti-L
 
I am interested in this topic but I am not sure to understand correctly and see how it can be implemented in TIMES. Are the penetration points (eg 80% of variable renewables)  linked to TIMES variables or fixed by the user ?

This reminds me a discussion with colleagues:

Let's say we have for a given renewable technology within one region different categories of installation sites expressed in terms of cost and maximal capacities. A way to see it would be to have NCAP_COST that depend of the installed capacity.

The solution found at that moment was to declare different process that share the same parameters except NCAP_COST.

Have you ever encountered this need and found another solution ?

[Antti-L]

> I am interested in this topic but I am not sure to understand correctly and see how it can be implemented in TIMES.

The example I described is easy to implement in TIMES. Just create a dummy flow somewhere in the model, and then define a user constraint saying that this dummy flow should be equal to the renewable generation in excess of 20% of total generation. And then define the variable cost (5 JPY/kWh in the example) on that flow.

> Are the penetration points (eg 80% of variable renewables)  linked to TIMES variables or fixed by the user ?

In the example, the penetration points were user-defined, but as you can see, only one point is in fact needed for that example (because the additional costs remains constant for any renewable generation in excess of 20%, while the additional costs per total generation is then increasing as described). The user constraint links the dummy flow variable to the 20% point, by giving the renewable generation in excess of 20% of total generation.

> Have you ever encountered this need and found another solution ?

Sorry, I could not quite follow the description on the kind of a need you are talking about. But yes, adding processes with different capital costs and constraints on the installed capacities seems a reasonable solution when the costs depend on the type of site.

[VictorG]

Alright, thanks, now I see. Sorry for the late reply