You are correct that the OPF assumes given fixed values for the shunt
parameters and they are not variables in the optimization.
You could do an OPF with temporary reactive-only generators placed at the buses
where you will have shunts and see what level of reactive output is needed from
them in order to get a feasible voltage profile. Then you could set your shunt
parameters accordingly.
Just an idea,
Ray
> On Apr 2, 2015, at 10:47 AM, Eser Patrick <[email protected]> wrote:
>
> Ray,
>
> Thank you very much for your reply. To go through your points step-by-step:
> Yes, I am sure that my p.u. parameters are correct. I calculate them with the
> same methodology as for smaller European countries – and there, the values
> match the expectations nicely. So I think this is not the problem here.
> The voltage limits I am imposing are even less restrictive than the real
> voltage limits. So I have already relaxed that constraint, yet I can not
> achieve convergence.
> The shunt capacitors are the critical point in my view. I am not using any in
> the current network, since I have no information about placement and
> dimensioning of the capacitors in the real system. So effectively, I am not
> doing any voltage support, which should explain the voltage drops. But my
> main problem with this issue is the following: If I understand correctly, the
> inserted shunt conductance and susceptance are imposed on the system. Meaning
> the MW and MVAr demanded/injected by the shunt capacitor are not being
> optimised during the OPF run. Is that correct?
> If my understanding about the shunt conductors is correct, how can I know
> beforehand, how much voltage support I will need at a certain substation?
> Would it make sense to do an iteration where I re-run the runopf() several
> times while adjusting the shunt values as a function of the voltages after
> each run of runopf()? Or do you recommend an alternative strategy here?
>
> I greatly appreciate your inputs. Any hints would help me.
>
> Best Regards,
>
> Patrick Eser
>
> From: Ray Zimmerman <[email protected] <mailto:[email protected]>>
> Reply-To: MATPOWER discussion forum <[email protected]
> <mailto:[email protected]>>
> Date: Monday 30 March 2015 15:21
> To: MATPOWER discussion forum <[email protected]
> <mailto:[email protected]>>
> Subject: Re: Voltage drops of very long transmission lines
>
> Here are few questions that might be helpful in steering you toward a
> solution ...
>
> - If you are convinced that you have the correct p.u. parameters for the
> transmission lines? Errors in the p.u. conversion could result in such
> problems?
> - If you are attempting to model an existing system, currently in operation,
> are you sure you are using voltage limits that include the actual normal
> operating point?
> - Are you missing any shunt capacitors used for voltage support?
> - Have you tried a simple power flow to see how far out of range the voltages
> are?
> - Have you tried using a continuation power flow to see if your operating
> point is beyond the point of voltage collapse?
> - Have you looked at FAQ #5
> <http://www.pserc.cornell.edu/matpower/#pfconvergence>?
>
> Best,
>
> Ray
>
>
>
>> On Mar 30, 2015, at 5:12 AM, Eser Patrick <[email protected]
>> <mailto:[email protected]>> wrote:
>>
>> Dear MATPOWER community,
>>
>> I am currently building a transmission system database for an emerging
>> nation. When running an OPF simulation for this system, I encounter the
>> problem of massive voltage drops over particularly long lines. This is
>> hardly surprising, since some lines of the system are longer than 400 km.
>> Combined with realistic reactances per km, this yields high per unit X (and
>> B) values, and hence voltage drops. When I artificially reduce the reactance
>> and susceptance values, I am able to achieve convergence.
>>
>> My questions are: Is there a way of solving this issue, without tampering
>> with the X and B values?
>>
>> I have played around with the tap ratios, but unfortunately, these are not
>> within the parameters of the optimisation (i.e. they are fixed a priori). I
>> have also tried adding a fictitious “reactive power generator” at each
>> substation, without success.
>>
>> Any help would be appreciated. Thanks!
>>
>> Best Regards,
>>
>> Patrick Eser
>
