Well, it is possible that it is simply the solution with least expected cost. 
If, for instance, other generators have large ramping or redispatch costs, it 
can be optimal to simply not use a generator that may fail in order to avoid 
redispatch costs. On the other hand, physical ramp limits (RAMP_10 in the gen 
matrix) can also prevent the dispatch from differing in the contingency and 
base states. Large redispatch or reserve costs on the generator in question can 
also be a possible cause (I know you said you tried zero reserve costs, so that 
shouldn’t be the cause in this case). One more idea … is the generator so 
expensive that it isn’t used at all, even if NOT modeling contingencies?

    Ray



> On Sep 20, 2016, at 12:01 PM, Rusch Maurus <rus...@student.ethz.ch> wrote:
> 
> Dear all,
> 
> I am currently trying to model generator contingencies in MOST. However, my 
> problem is, that there is no variation of generator dispatch over all the 
> modelled states. The generators with a contingency assigned have therefore 
> zero dispatch in the base and contingency state. Consequently I get no 
> differences between the base case and the states in which contingencies do 
> occur.
> Moreover I have applied the same set up to model transmission line 
> contingencies. This setup doesn’t cause any problems and I am getting 
> different power flows in the base state and the states with contingencies.
> 
> In my setup, I have assigned a big amount of positive/negative reserve 
> quantities with zero reserve costs to the generators with modelled 
> contingencies. I have already tried to define contingencies for all kind of 
> generators (different maximum capacity, different marginal costs). The system 
> I am simulating consists of 940 buses, 4500 generators and 1350 transmission 
> lines.
> 
> Did someone have the same problem? What could be the reason that I don’t get 
> any differences in generator dispatch among the modelled states?
> 
> Thank you very much for any kind of inputs and advices
> 
> Maurus Rusch



Reply via email to