CPF - issue with PQ/PV conversion with multiple generators at PV bus #26
Comments
|
Once again, thank you for raising the issue and providing a concrete example to test it.
I'll have to take a look at what's causing this issue. I believe all the generators should hit their reactive power limits simultaneously. CPF uses the reactive power calculated by one of the core MATPOWER routines (I think it's pfsoln). MATPOWER uses a proportional sharing logic based on the reactive power range for dispatching generator reactive power. This results in all the generators incident at a bus hitting their Q limits simultaneously.
I will check and report back.
On Sep 25, 2017, at 3:32 AM, ElisNycander <notifications@github.com<mailto:notifications@github.com>> wrote:
I think there is a problem with converting PV buses with multiple generators to PQ buses in the CPF. What happens is that if one generator reaches its QMAX, the bus will be converted to a PQ bus even if there are other generators at the bus that still have more Q capacity.
It happens in the following example using case4gs, where I have added an extra generator to bus 4. In the CPF, in the very first few steps, the QMAX = 167.2 of the new generator is reached, and bus 4 is converted to PQ. However, generator 1 is only at Q =-2.26 MVAr, i.e. not at its q limit.
If instead an ordinary PF with the target case is run, the q limits are handled correctly, and both generators at bus 4 are at their maximum q limits, for a total of 367.8 MVAr.
clear;
define_constants;
% options
mpopt = mpoption();
mpopt.out.all = 1;
mpopt.pf.enforce_q_lims = 1;
mpopt.cpf.enforce_q_lims = 1;
mpopt.cpf.parameterization = 3;
mpopt.adapt_step = 0;
mpopt.step = 0.1;
mpopt.cpf.enforce_v_lims = 0;
mpopt.cpf.enforce_flow_lims = 0;
mpopt.cpf.plot.level = 2;
mpopt.cpf.stop_at = 'FULL';
mpc = case4gs();
mpc.gen(1,[PG QMAX QMIN]) = [300 200 -1e9];
mpc.gen(2,QMAX) = 1e9;
% add extra generator
mpc.gen = [mpc.gen;
4 200 0 167.8199 -167.8199 1.02 100 1 200 zeros(1,12)
];
mpcb = mpc;
mpct = mpc;
% increase load
mpct.bus(:,[PD QD]) = 2*mpct.bus(:,[PD QD]);
% run cpf
mpcc = runcpf(mpcb,mpct,mpopt);
% run pf
mpc2 = runpf(mpct,mpopt);
Matpower version:
MATPOWER Version 6.1-dev 23-Aug-2017
MATLAB Version 8.6 13-Aug-2015 Release: (R2015b)
Optimization Toolbox Version 7.3 13-Aug-2015 Release: (R2015b)
MIPS Version 1.3-dev 16-Mar-2017
MOST Version 1.0.1-dev 07-Aug-2017
SDP_PF -- not installed --
YALMIP -- not installed --
BPMPD_MEX -- not installed --
CLP -- not installed --
CPLEX -- not installed --
GLPK -- not installed --
Gurobi -- not installed --
IPOPT -- not installed --
KNITRO -- not installed --
MINOPF -- not installed --
MOSEK -- not installed --
PARDISO -- not installed --
PDIPMOPF -- not installed --
SCPDIPMOPF -- not installed --
SDPT3 -- not installed --
SeDuMi -- not installed --
TRALMOPF -- not installed --
Architecture: PCWIN64
—
You are receiving this because you are subscribed to this thread.
Reply to this email directly, view it on GitHub<#26>, or mute the thread<https://github.com/notifications/unsubscribe-auth/ABHuMIolYH2oofV-Cks_P4elW4J2CAUEks5sl2UKgaJpZM4PiX4e>.
|
|
The issue here is related to generators at the same bus with wildly different Q ranges. The In the example provided, we have ... So, when the total Q is 185.32, alpha is 0.9999998175, resulting in ... So the upper bound on 4 appears to be binding, even though the upper bound on 1 is clearly not yet and won't be until alpha gets extremely close to 1. The solution (which I haven't yet implemented) is to check the The temporary workaround is to avoid using generator Q limits with such wildly different ranges. |
|
Has the generator reactive distribution logic changed recently? I thought the generator reactive powers are distributed proportional to their reactive power range such that all the generators hit their reactive power limit simultaneously.
On Dec 5, 2017, at 8:27 AM, Ray Zimmerman <notifications@github.com<mailto:notifications@github.com>> wrote:
The issue here is related to generators at the same bus with wildly different Q ranges. The pfsoln() function distributes the total Q between generators at the same bus by choosing a factor alpha, between 0 and 1, such that the the sum of the Qg across the generators at the bus is the total Q, and for each generator Qg = Qmin + alpha * (Qmax - Qmin).
In the example provided, we have ...
Qmin1 = -1e-9, Qmax = 200
Qmin4 = -167.8199, Qmax4 = 167.8199
So, when the total Q is 185.32, alpha is 0.9999998175, resulting in ...
Qg1 = 17.5025
Qg4 = 167.8199
So the upper bound on 4 appears to be binding, even though the upper bound on 1 is clearly not yet and won't be until alpha gets extremely close to 1.
The solution (which I haven't yet implemented) is to check the cpf.q_lims_tol on the difference between the dispatch and the limit for the totals at the bus, not for the individual generators.
The temporary workaround is to avoid using generator Q limits with such wildly different ranges.
—
You are receiving this because you commented.
Reply to this email directly, view it on GitHub<#26 (comment)>, or mute the thread<https://github.com/notifications/unsubscribe-auth/ABHuMFUntUAHNe7zOpcXz8UCYrfzoT1Mks5s9VKJgaJpZM4PiX4e>.
|
|
Well, it has changed recently, but only in the handling of gens with infinite ranges, which doesn't affect this case. What I described above is proportional and, if you have infinite precision, they do reach their limits simultaneously. It's just that with wildly different ranges, numerically one falls within the For example, in this case, we have two generators that are 99.99998175% of the way to Qmax, for one (with a range of ~300MVAr) that's easily within |
- Triggered by multiple generators at the bus with wildly different reactive ranges.
I think there is a problem with converting PV buses with multiple generators to PQ buses in the CPF. What happens is that if one generator reaches its QMAX, the bus will be converted to a PQ bus even if there are other generators at the bus that still have more Q capacity.
It happens in the following example using case4gs, where I have added an extra generator to bus 4. In the CPF, in the very first few steps, the QMAX = 167.2 of the new generator is reached, and bus 4 is converted to PQ. However, generator 1 is only at Q =-2.26 MVAr, i.e. not at its q limit.
If instead an ordinary PF with the target case is run, the q limits are handled correctly, and both generators at bus 4 are at their maximum q limits, for a total of 367.8 MVAr.
Matpower version:
The text was updated successfully, but these errors were encountered: