Cao Huy's profile - activity

Firstly, changing Voltage set point doesn't change generator's Qmax, but Qgen. Secondly, there are many ways to control reactive power. You could try some of them:

1. Optimal power flow
2. Implementing of compensators (fixed shunt, SVC...)
3. Changing parametters manually. In your case, I suggest to raise Vset of not only one generator, but all generators in the surrounding area.

Here are some reasons:

1. Supply-demand unbalance
2. Too much power transfer through an interface
3. Voltage control problems

I suggest:

1. balance load and generators manually, for each area.
2. run loadflow with flatstart, ignore reactive power limits.

In case there is not sensitive information in your file, I could lend you a hand.

How about make a new source subsystem with only generations you want to change?

I haven't used GE models before. Usually, I check if the wind farm is modeled well in loadflow condition, whether it checks in Wind Machine Control Mode, any problem with the reactive power and voltage in loadflow. Then, I check INITIAL CONDITION CHECKING if it's available. Hope that helps.

It's also happened to me.

My bads. Thank you. I have fixed the answer.

Which one do you mean? About the RoCoF, I need to clarify that the output is in 1/Sec. If you want to have ouput in Hz/Sec, you need to multiply the output with the rated frequency: f.

I think you should check following assumptions of your simulations:

1. Did your RoCoFs have the same time frame? Usually, people don't compare RoCoF of the different time frame.
2. Did you simulate the quick frequency response components? That may affect the RoCoF.
3. Check your primary frequency control? Is there any different in the droop of simulations? How many generations taking part in the frequency control in each case? How many active power reserve?

Theoretically, RoCoF at the instant of losing a generator is calculated by the formula:

RoCoF = f0Delta P/(2Hi*Si) with Delta P is the active power imbalance and Hi and Si is inertia and apparent power rating of generator i. Given the Hi of SPVs are zero. RoCoF are increase with increasing of SPV penetration rate.

Please check your viewing option and make sure that you have save it as a *.dyr file. Sometimes, you don't show extension so that you mistake .dyr.txt file with a genuie dyr file.

I have tried and it didn't work. I guess because PSSE neglect loads component when conduct short circuit analysis.

I have a few steps you might want to follows:

1. Looking for dynamic models. Each controls has one or more than one dynamic models, for example gorvernor,turbine and exciter for a generator. You can see the block diagram, syntax and typical parameters of built in models in MODELS documents of PSSE. Or you can search in the internet for some available benmark dyr files.
2. Paste those parameters to a txt file and save as a .dyr file.

That's a practical approach. I thought of using a constant load with ILOAD 1.2pu. What do you think about it?

Hello, guys.

I am conducting a short circuit calculation with the existance of type 4 wind turbines. The manufacturer recommended:

The WTG should be modelled as a current source for short circuit studies. The current source should contribute a maximum of 1.2 p.u during the fault. In PSS/E the value of Zsource is obtained from the maximum short circuit current contribution. The recommended values are provided separately.

Anyone here has experienced with short circuit studies of WTG? How did you do it? I have no idea to add a current source to SC study in PSSE. I thought of using Zsource =1/ Isc_max. Is this possible?

I usually use filename.raw. You should check the API document of PSSE. Unless you declare the new path in python, it is default that your files and your macro in the same folder. "psspy.rawd_2()" functions is for writing raw file; psspy.read() for reding raw file. My bad. I also fixed the answer.

Thanks for your reply. By creating fi-cases first instead of for each iteration will reduce a lot of time. I will also check for multiprocessing, havent done a parallel running before. Of course I'll give it a try.

Here are few steps to do the task:

Firstly, you have to record the macro you want to apply, using the reocding button in the GUI of PSSE.

Secondly, fill in these below code, which will do the task for you.

import psspy
File_List = [your list of sav formatted file names you want to excecute]
for file in File_List:
    psspy.case(file) #or load a raw formated file by psspy.read()
    #insert your macro here
    #save sav file if you want by psspy.sav(); save raw file by psspy.rawd_2()

Thirdly, run it.

Above are just some simple stuffs, for more infomation you should check out the documents of PSSE.

Hello everyone,

My question here lies on the method to change the parameters corresponding to the interested frequency: 1. Export to raw then change parameters corresponding to the interested frequency in raw. Then import to PSSE and get Thevenin impedance. 2. Change the parameters in PSSE, then get Thevenin impedance.

I have tried the 1st one, but it took about 20s to execute a scan (one N-0 loadflow case of about 500 buses, scan in range of 5 to 50, step 0.5). It seems forever when you want to do N-1 and N-1-1 analysis.

Which way is better, in your opinion? Do you have another way to do the job? Anyone has a benmark or test system to validate the results?

Hello everyone,

My question here lies on the method to change the parameters corresponding to the interested frequency: 1. Export to raw then change parameters corresponding to the interested frequency in raw. Then import to PSSE and get Thevenin impedance. 2. Change the parameters in PSSE, then get Thevenin impedance.

I have tried the 1st one, but it took about 20s to execute a scan (one N-0 loadflow case of about 500 buses, scan in range of 5 to 50, step 0.5). It seems forever when you want to do N-1 and N-1-1 analysis.

Which way is better, in your opinion? Do you have another way to do the job? Anyone has a benmark or test system to validate the results?

Haven't used PSSE v 34, but PSSE 33 respond well with print.

I don't think it work when you applied short-circuit impedance in to reach LF solution. I recommend you change the transient parameter of generator to do dynamic simulation instead of putting a abnormal value of impedance into system.

Because in case Pgen is a float, you can't expect it remain the same after the conversion. Normaly, when doing float comparision, we compare difference with a small epsilon.

For more information, read:

float comparision

In the 2 winding transfomer data, you can change the winding connection angle and turn ratio to make an 2 winding transformer - phase shifting.

Detail is described in the following document:

Phase-shifting

IMHO, when you applied a fault, it's normally to see "network not converged" message, cause you changed a impedance to zero (or close to zero). So I think it's ok to see one or a few NoC message in the process.

Love to hear another idea.

I am trying to run dynamic simulation in psse 33 with data files in version 29, those I had ran sucessfully in the past. Power flow parts and preparing steps like convert generators, convert loads, factorizing and so on worked well, but when I started using command run, psse33 said something about complex load not converged, NaN and hanged forever. Of course, I couldn't get result file I want.

There's no other warning from PSSE when I run simulation on psse 33. I also have checked complex load model in Psse 33 Documents Models.pdf and the syntax haven't changed at all.

If anybody know about this problem, please give me a hint. Thank in advance

Hi, all. I'm trying to analyse angle stability of machines in the system, which costs me a lot of time because I have to run dynamic simulation over and over again to find CCT (critical clearing time).

That's so I came up with an idea to run it automatically. The script isn't so hard, I guess I can deal with it, but the problem is how to identify if a system stability or not.

Manually, I could say if a system is instability by some signs, such as that machines' speeds and angles are seperated largely, or voltages drop below 0.5 (voltage stability), etc. However, defining a quantitative criteria which works for every machines in the system is not as easy.

I came some ideas for the criteria, but not so sure about them:

1. the difference between any 2 machine's angles isn't bigger than 100 degree.
2. The difference between any machine's speed and the average speed of the system must be smaller than 0.1. Average speed is calculated like this: Sum((Machine speed)*(machine inertia constant))/ (Sum Inertia constants)
3. d Delta/dt turn sign (from negative to positive or vice visa) more than once, in which Delta is the difference between angle of interested machine and reference machine.

I want to know is there any available function in pss/e that can distinguish instability system? If not, is there any other idea to make it?

I've tried but... In the case my DLL file is corrupted, could any of you please send me your PSSplotDLL.dll file. I am using PSSE 33.4.

Is it sound similiar to you?