NPCC and WECC networks are unstable?

[federicagiannunzio]

Hi everyone,
I'm trying to analyze the eigenvalues of some networks, included wecc and npcc.
This is how I generate the eigenvalues report:
andes.config_logger(stream_level=20) print(andes.utils.paths.cases_root()) case_path = andes.get_case('npcc/npcc.xlsx') ss = andes.load(case_path) ss.PFlow.run() ss.EIG.run()

and this is the output I obtain:
**Input file parsed in 0.4421 seconds.
System internal structure set up in 0.0302 seconds.
-> System connectivity check results:
No islanded bus detected.
System is interconnected.
Each island has a slack bus correctly defined and enabled.

-> Power flow calculation
Numba: Off
Sparse solver: KLU
Solution method: NR method
Power flow initialized in 0.0056 seconds.
0: |F(x)| = 0.004215993603
1: |F(x)| = 3.030635298e-08
Converged in 2 iterations in 0.0023 seconds.
Report saved to "npcc_out.txt" in 0.0030 seconds.
GENCLS (vf range) out of typical lower limit.

idx | values | limit
-----------+--------+------
GENCLS_3 | 0.990 | 1
GENCLS_4 | 0.926 | 1
GENCLS_16 | 0.933 | 1

Initialization for dynamics completed in 0.0313 seconds.
Initialization was successful.

-> Eigenvalue Analysis:
48 states are associated with zero time constants.
Positive 41
Zeros 1
Negative 293
Eigenvalue analysis finished in 0.1205 seconds.**

Similar output for WECC network.

I have noticed that there are some positive poles, meaning that the system it's not stable.
Can someone explain me why those two systems are unstable by default?
Thank you in advance,
Federica

[cuihantao]

We are checking on it. Thanks for reporting!

[zilinzhuang]

Hi, Federica! You're right—our previous version indeed had several real roots. I reviewed the results of the eigenvalue analysis and found that the occurrence of many positive real parts was linked to the IEEEX1 model. Consequently, I tried adjusting a few parameters; feel free to give them a try. We set the parameters for all IEEEX1 instances as follows: TR=0.01, TC=1, and TB=1 (which aligns with the default values ​​specified in the ANDES documentation). Additionally, I adjusted the following gains (I made this change because, despite tweaking numerous other parameters, the variable W_x IEEEX1 4 remained unstable; upon consulting the ANDES documentation, I noted that this variable is correlated with KE): for IEEEX1_4, I set KE=1, VRMAX=5, and VRMIN=-5. Give it a shot!

[cuihantao]

It would be helpful if you could upload the fixed .xlsx file here.

[zilinzhuang]

Sure!
npcc.xlsx

[federicagiannunzio]

Hello, thank you so much for your help.
I noticed default positive poles also for WECC network (the script I'm trying to run is the same one I have reported at the beginning), maybe the issue is the same?
By the way, I tried to run the simulation again with the fixed npcc.slsx file and now the poles are ok, but this is the omega plot I obtain now. Does it sounds right to you?
Thank you in advance,
Federica

[zilinzhuang]

Regarding the generator's speeds, I believe it is correct.

[zilinzhuang]

test_wecc.ipynb
I suspect this issue is very likely caused by certain "zero time constants" within the XLSX file. Therefore, you might want to try changing them to non-zero values. Regarding the frequency verification for NPCC, I will get back to you later!