VedaSupport Forum
Early Retirement Costs - Printable Version

+- VedaSupport Forum (https://forum.kanors-emr.org)
+-- Forum: VEDA Front End (https://forum.kanors-emr.org/forumdisplay.php?fid=2)
+--- Forum: Questions and Observations (https://forum.kanors-emr.org/forumdisplay.php?fid=7)
+--- Thread: Early Retirement Costs (/showthread.php?tid=845)



Early Retirement Costs - Felix - 25-05-2020

Hey everybody! I'm trying to implement early retirement for gas grids into my model to give the model the possibility to reduce the Fixcosts coming from the grid. I included early retirement by defining RCAP_BND_Up=0 for the Base Year for this process. As far as I can see, the model is retiring unneeded capacities so this part would work as intended. The problem now is that I additionally tried to implement Costs for this early retirement. Basically an Investment-Cost to reduce the capacity of a process. I tried to do this with the NCAP_DCOST Parameter, but as far as I can see, this parameter has no effect on the early retirement: I ran multiple scenarios with completely different NCAP_DCOST-values but the results (retired capacities, total discounted system costs) barely changed. I assume now that NCAP-DCOST only affects decomissioning after the end of the lifetime of a process not early retirement. At this point I'm wondering wether I made a mistake with modelling or NCAP_DCOST doesn't apply to early retirement. My questions now would be:
  • Does NCAP_DCOST apply to early retirement? Where would I see the resulting costs in VEDA_BE (which attribute)?
  • If NCAP_DCOST doesn't apply to early retirement: Is it possible to implement Costs for the early retirement of a process?
Best regards Felix


RE: Early Retirement Costs - Antti-L - 25-05-2020

Indeed, NCAP_DCOST is the decommissioning cost, and it is scheduled after the technical lifetime, regardless of early retirements. So, NCAP_DCOST applies to processes with early retirements, but the decommissioning is not re-scheduled. The early retirements feature was based on the MARKAL design, redesigned by Ken Noble for TIMES.  The design does not include any costs for early retirements, other than by losing the salvage value. The savings in fixed costs is the main incentive for the model choosing early retirements. If you wish to suggest a design for accounting additional costs from early retirements, I encourage you to do so. You can submit your proposal to the ETSAP project head.


RE: Early Retirement Costs - Felix - 27-05-2020

Hello Antti, that's what I thought (and feared). But at least I can stop trying to model it this way. Thank you anyway for your quick response!


RE: Early Retirement Costs - Antti-L - 27-05-2020

So, shall we receive your novel design for accounting additional costs from early retirements? However, I think there are many examples of plants where early retirement does not cause any notable additional costs. The plant is just mothballed, and the fixed costs are saved until the end of the lifetime, and then decommissioned according to original schedule. Therefore, I would think your design should not be based on unconditionally imposing re-scheduling of decommissioning.


RE: Early Retirement Costs - Antti-L - 30-05-2020

Felix Wrote:If NCAP_DCOST doesn't apply to early retirement: Is it possible to implement Costs for the early retirement of a process?
Ok Felix, as you don't seem interested in providing a design for accounting additional costs from early retirements, let's still return to your question "Is it possible to implement Costs for the early retirement of a process?" As already mentioned, NCAP_DCOST does indeed apply to processes with early retirements, in the same way as applied normally. However, no direct option for defining additional costs on early retirements have been implemented, simply because thus far no-one has considered it useful enough to merit designing such for TIMES. However, there is a work-around, by defining a retrofit option for the process, and requesting that the retrofitted capacity is equal to the retired capacity. When doing so, the investment cost defined for the retrofit will be effectively paid at the beginning of the period where the original process is retired.  And the retrofit process can be a dummy process without any other functionality than defining this investment cost for the amount of retired capacity. I guess this could be pretty much as you wanted? I just tested this work-around for a gas pipeline process, defining a retrofit option for it and requesting the retrofitted capacity to be equal to the retired capacity. And as far as I can see, it worked perfectly well.


RE: Early Retirement Costs - Felix - 02-06-2020

Hello Antti, yes you're right: Unfortunately I don't have the capacities (and maybe knowledge as well) to provide a new design... But I tried your suggested solution and it worked! That was exactly what I wanted to model. I actually also tried to implement the early retirement costs via retrofit before, but without coupling the capacities the model naturally didn't build the dummy-retrofit-process. This dummy retrofit process now has a low efficiency with high variable costs to avoid the usage of this "retired capacity". Thanks alot for your help!


RE: Early Retirement Costs - Antti-L - 02-06-2020

Good that you got it working, too.
(02-06-2020, 04:20 PM)Felix Wrote: This dummy retrofit process now has a low efficiency with high variable costs to avoid the usage of this "retired capacity".
Hmm... don't fully understand. A dummy process would not produce any energy (PCG=dummy commodity), and so there would be no need for either low efficiency nor high variable costs.


RE: Early Retirement Costs - Felix - 02-06-2020

I'm sorry I didn't make myself clear: I used the word "dummy" although it's not an actual dummy process. It has the same In- and Output as well as the same PCG as the retired process. It's bascially a copy of this process but with the described altered values. I found it personally more convenient this way. I hope my modelling is a bit clearer now. Best regards Felix


RE: Early Retirement Costs - Ceas19 - 21-06-2022

(30-05-2020, 07:35 PM)Antti-L Wrote:
Felix Wrote:If NCAP_DCOST doesn't apply to early retirement: Is it possible to implement Costs for the early retirement of a process?
Ok Felix, as you don't seem interested in providing a design for accounting additional costs from early retirements, let's still return to your question "Is it possible to implement Costs for the early retirement of a process?" As already mentioned, NCAP_DCOST does indeed apply to processes with early retirements, in the same way as applied normally. However, no direct option for defining additional costs on early retirements have been implemented, simply because thus far no-one has considered it useful enough to merit designing such for TIMES. However, there is a work-around, by defining a retrofit option for the process, and requesting that the retrofitted capacity is equal to the retired capacity. When doing so, the investment cost defined for the retrofit will be effectively paid at the beginning of the period where the original process is retired.  And the retrofit process can be a dummy process without any other functionality than defining this investment cost for the amount of retired capacity. I guess this could be pretty much as you wanted? I just tested this work-around for a gas pipeline process, defining a retrofit option for it and requesting the retrofitted capacity to be equal to the retired capacity. And as far as I can see, it worked perfectly well.
Dear Anti, I am representing a retrofit option in my model. I just would like to be sure how do I request the retrofitted capacity to be equal to the retired capacity. Do I just have to include a NCAP_BND for the retrofit option? Thanks in advance for your help.


RE: Early Retirement Costs - Antti-L - 21-06-2022

Assuming your host process is prc (which has some retrofit option p), just define also PRC_REFIT(r,prc,prc)=2, and then the retrofitted capacity will be forced equal to the retired capacity of prc (if the capacity of prc is retired early).


RE: Early Retirement Costs - Ceas19 - 22-06-2022

(21-06-2022, 10:38 PM)Antti-L Wrote: Assuming your host process is prc (which has some retrofit option p), just define also PRC_REFIT(r,prc,prc)=2, and then the retrofitted capacity will be forced equal to the retired capacity of prc (if the capacity of prc is retired early).
Got it, thank you very much.