Veda2.0 Released!


Changing the objective function in TIMES
#2
Yes, you can use almost any user-defined objective functions by using the Sensitivity and Tradeoff Analysis facility.
See the documentation: https://iea-etsap.org/docs/TIMES-Stochastic-Final2016.pdf

The user-defined components of the objective function must be defined as non-binding User Constraints.  Then you can combine them as you like in the objective function, where you can also use the standard objective function as a component.

However, in my opinion the amount of emissions (e.g. cumulative total GHG) is not a good objective function as such, because of several reasons.  First, the results from minimizing emissions does not lead to any meaningful prices of energy commodities in the results.  Second, the physical results may be in many respects arbitrary, because costs would not matter at all under a pure emission minimizing objective.  Third, many models have an unlimited backstop technology for GHG emissions, which can reduce annual emissions even to large negative amounts. The amount of CO2 in the atmosphere is currently less than 900 GtC.  And so, just by having negative emissions of -30 GtC for about 40 years, the model would end up to having close to zero CO2 in the atmosphere (taking into account the biosphere slowing down the decrease), which would be destructive for plants and climate.

In my view a more reasonable approach would be to the following, and the scenario set-up is illustrated below:
 1) first use the standard objective function for a Baseline
 2) then minimize GHG emissions subject to some limit in the loss in consumers' and producers' surplus compared to Baseline
 3) finally, minimize the costs again, by imposing the emission level obtained in step 2

   
This is a multiphase type of a trade-off analysis, where the final step is needed for getting meaningful energy prices and physical results. I just tested it and it appears to work as designed. The results from each phase are reported (suppressing the reporting of results from phases 1 and 2 does not appear to work now, but will be fixed in the next version of TIMES).
Reply


Messages In This Thread
RE: Changing the objective function in TIMES - by Antti-L - 09-11-2018, 09:54 PM

Forum Jump:


Users browsing this thread: 1 Guest(s)