I am considering using PSS/E for building node-breaker model of a small power system, for aiding the development of a substation level state estimation algorithm. In other words, I want to model the substations in greater detail, including their bus sections and breakers.
In doing so, I discovered that PSS/E models zero-impedance branches with zero resistance, but with a small reactance (0.0001 pu, cf. attached image of a breaker-and-a-half bus topology) under a given threshold (given as 0.0001 pu). Why does it do so, and what consequences does it have for a steady-state load flow?
My assumption here is that it does not have a separate modeling for zero-impedance branches, but simply treats them as low-impedance branches. But, why would it then have a zero-impedance threshold?
I greatly appreciate all inputs you might have on this subject, even if it is suggestion for another program for the same use.
I think one simple way is to model the zero impedance branch as equivalent as a closed breaker or switch. Possibly zero impedance may cause some divided by zero issues in the calculation, so it's better to avoid it. Plus, a small enough reactance won't make much change to the power flow.
PSS/E models correctly a zero-impedance branch. The user is allowed to set the threshold for the branch reactance that will define which branches to treat as zero-impedance branches or not.
So, you enter the branch data with R=0 and X=1e-5 pu (or some very small, non-zero, value for the branch reactance). As long as the provided value is below the selected threshold for minimum value for a reactance to be treated as a non-zero reactance, everything should be OK.
Now, I cannot remember how PSS/E handles loops of zero-reactance branches. Once upon a time, that was an error, I think. I know this has been modified over the years, but cannot tell you with certainty if any topology of zero-reactance branches can be entered/handled in PSS/E.
Feedback from me: You can using another programme, e.g.: MATLAB programme. Then how to run this programme that MATLAB to review PSS/E model of zero-impedance branch can be seen on the following page: https://www.mathworks.com/search/site_search.html?c%5B%5D=entire_site&q=PSS%2FE+model+of+zero-impedance+branch
Regards,
Syahrial Shaddiq
Thank you all! As Mr. Zhang pointed out, a very small reactance would not alter the load flow results, so modeling the zero-impedance branches/closed breakers as a small reactance should yield a good result.
Since I am implementing the state estimator in Matlab, I actually decided to use MATLAB Matpower, because the model is easily built and the power flow easily run within the MATLAB environment. Matpower does not really handle breakers and zero-impedance branches, but enforcing a very low reactance on closed breakers and a very high reactance on open breakers in the model, as you pointed out, results in a quite accurate load flow!
Now that I have implemented this, it seems to be working well to provide the state estimator with the necessary measurements (of voltages and power flows). The downside, I would imagine, of handling the zero-impedance breakers in this way is the risk of very ill-conditioned matrices. Therefore, closed breaker and open breaker impedance values should not be too extreme.
Again, thank you for your assistance!
Maybe I did not explain myself: PSS/E does represent as a zero impedance branch any branch with a reactance smaller than the given threshold. No branches with very small reactances are included in any matrices. Two nodes connected by a branch being handled as a zero impedance branch are considered as the same bus/node.
- Badges
- Science topic