Time-stepped model

[Bryce]

Hello,

I have a question relating to how the time-stepped model variant in TIMES works with regard to the overlapping years. I am running some time-stepped models with 10 year blocks and 5 year overlaps. In the description of how to use the time-stepped model variant in in the 'TIMES version 3.3. user note - User control switches in TIMES' document section 6.2 (I have attached the diagram as well), there is no mention on what the effect is of these overlaps is, but the way I interpret it is that it is similar to the description given by IIkka Keppo's paper in that the model optimises for the block of 10 years but only the results from the first 5 years is taken whereas the last 5 years is then subject to the next optimisation block (of 10 years). So you are planning for 10 year blocks but the planning is 'revised' after 5 years.

Is this the correct interpretation for how TIMES works in the time-stepped model variant?

The paper:

IIkka Keppo & Manfred Strubegger, "Implications of limited foresight and sequential decision making for long term energy systems planning: an applicaion of the myopic MESSAGE model"

The diagram from the 'User note - User control switches in TIMES' section 6.2:

Thanks for your help.

[Antti-L]

Yes, after each step the period(s) not overlapping with the next step are fixed to the current solution, while the overlapping period(s) remain subject to optimization in the next step.

[Bryce]

Thanks for the quick response!