GaryB's profile - activity

1. Solve the case and record the MW transfer across a line.
2. Add 1 MW of load (or generation) on a bus and solve the case again.
3. Record the MW transfer again.

Shift factor is the difference in flow divided by the load/gen change.

sf = (MW1 - MW2) / (load [or gen] change)

The signs matter as shift factor may be negative or positive.

Example: If you have 100 MW flowing on a line, and you add 1 MW of generation at a particular bus. You now have 100.23 MW flowing on the line, the shift factor is (100.23 - 100) / 1 = 23%.

Have you checked the log file for warnings on the CMLD models? These types of messages may exist: At bus 4, CMLD load bus voltage of 0.9471 pu is below 0.95 minimum. Original feeder Z of 0.2295 +j 0.2295 pu on SBASE will be reduced in an attempt to bring V up above 0.95 pu

Start with a problem and try to isolate it. You can disable the other models to get it down to one DSTATE issue.

In the GGOV1 model, find state 2 in a diagram. That’s the differential control. Check RSELECT and other parameters upstream of State 2.

Change RSELECT to zero to try and isolate that circuit.

Manipulate r.

Check State 8. Plot it. If it’s stable, the problem is somewhere else.

I hope you know some control theory.

Move on to the next model. Some DSTATE issues can be tolerated.

It’s not easy. Good luck.

Can you please select the answer that worked for others to know how to fix this?

It should work; the speed circuit should always be active and working regardless of rselect. The only other thing I would try is to set R=0 (or something small) in the psas file if possible. This should logically do the same thing. Please post if you figure it out.

Does it work if it's initialized in ISOC mode?

I can't say I know much about GGOV1, but you can change the Rselect or any other ICON with BATCHANGEPLMODICON.

PASSTHRU
BAT_CHANGE_PLMOD_ICON,<busno>,'<unitID>','GGOV1',1,0
FIN

will change RSELECT to '0' inside a PSAS file.

ChatGPT is better suited to help you with your homework.

It’s good to use Prony Analysis if your angle damping rate is difficult or impossible to read.

Some areas can have multiple generators swinging together with different damping rates and natural frequencies. The superposition of these swings can yield rotor angles that don’t resemble a damped sine wave. Prony is a method to decompose the angles into individual damped (ideally) waves with a damping rate for each each component.

There should be some documentation in this. 1. Load the sav case. 2. Convert the loads and generators. 4. You can save the converted case. 5. Load the dyr file. 6. Initialize the case. 7. You can now save the .snp file. This saves the initialized snapshot before any simulation has occurred.

This is brief. There are tutorials on YouTube. I would look for a tutorial walking you through using the savenw file provided as an example with PSSE.

Another option is to pass psspy into the function:

def TS1_S1_LOADRELIEF(psspy):

I would look at doing an EMTP model with PSCAD or something similar.

The dynamic models are positive sequence only. It's possible to run unbalanced fault (at least SLG), but PSSE still only simulates the positive sequence. You cannot read negative sequence values because they aren't simulated.

I assume short-circuit analysis simulates negative and zero sequences. The detrimental effects of negative sequence happen during a fault. I would use short-circuit analysis and known breaker timings.

Try this:

import multiprocessing as mp
def load_out(out_file,q):
    try:
        chnfobj = dyntools.CHNF(out_file,outvrsn=0)
    except TypeError:
        chnfobj = None
    q.put(chnfobj)

q = mp.Queue() #load the .out in a separate process 
p = mp.Process(target=load_out, args = (out_file,q,))
p.start()
chnfobj = q.get()

It runs in a subprocess and doesn't collide with psspy.

Note that there is no p.join() function; it hangs indefinitely. There is probably a way to garbage collect or pass data rather than the chnf obj back. But this works for me.

The WECC has pretty good resources on the internet. Also googling these models will help find other sources. Dynamics and controls books can help with fundamentals.

0.01s seems long. We run 1/4 cycle time steps.

Parsing the log file is probably the best way. Also you will end up with zero length .out files that may also be used to detect such a condition.

Be advised, some simulations may halt and not terminate causing your script to hang up. I use python multiprocessing to launch each dynamics run in its own process. While running, I monitor the file sizes of the .out and .log files. If they stop growing larger and the case is still running, it gets terminated. An ever increasing size of these two files means it’s still working.

Try this:

import subprocess

runstr = r'"C:\Windows\SysWOW64\cmd.exe" /k '\
    r'"C:\Program Files (x86)\PTI\PSSE32\SET_PSSE_PATH.BAT"'\
    r'& compile.bat'

def open_app():
    p=subprocess.call(runstr)
open_app()

This is how I do it:

Assuming you can run Python programs currently and you want to run from the command line.

I can't post links, so I had to change this to be more complicated. There is a clever way to put the bulky path stuff in the beginning into a library. This should work with PSSE 33 or 34.

Modify with filenames and try and run this code, preferably from one of the Command Line file links provided by PTI with PSSE (in the PSSE34 Folder on the Start Menu):

import _winreg
import sys
import os
# Add the PSS/E directories to our Windows path variable so we can load in the PSS/E Python libraries that are available there


if 'PSSE34' in os.getenv('PATH'):
    print('PSSE 34 detected')
    try:
        pssepath, regtype = _winreg.QueryValueEx(_winreg.OpenKey(_winreg.HKEY_LOCAL_MACHINE, "SOFTWARE\\PTI\\PSSE 34\\Product Paths", 0, _winreg.KEY_READ), "PsseInstallPath")
    except WindowsError:
        pssepath, regtype = _winreg.QueryValueEx(_winreg.OpenKey(_winreg.HKEY_LOCAL_MACHINE, "SOFTWARE\\\Wow6432Node\\PTI\\PSSE 34\\Product Paths", 0, _winreg.KEY_READ), "PsseInstallPath")
    sys.path.append(os.path.join(pssepath, 'PSSBIN'))
    sys.path.append(os.path.join(pssepath, 'PSSLIB'))
    sys.path.append(os.path.join(pssepath, 'PSSPY27'))
    os.environ['PATH'] += ';' + os.path.join(pssepath, 'PSSBIN')
    os.environ['PATH'] += ';' + os.path.join(pssepath, 'PSSLIB')
    os.environ['PATH'] += ';' + os.path.join(pssepath, 'PSSPY27')
else:
    try:
        pssepath, regtype = _winreg.QueryValueEx(_winreg.OpenKey(_winreg.HKEY_LOCAL_MACHINE, "SOFTWARE\\PTI\\PSSE 33\\Product Paths", 0, _winreg.KEY_READ), "PsseInstallPath")
    except WindowsError:
        pssepath, regtype = _winreg.QueryValueEx(_winreg.OpenKey(_winreg.HKEY_LOCAL_MACHINE, "SOFTWARE\\\Wow6432Node\\PTI\\PSSE 33\\Product Paths", 0, _winreg.KEY_READ), "PsseInstallPath")
    sys.path.append(os.path.join(pssepath, 'PSSBIN'))
    sys.path.append(os.path.join(pssepath, 'PSSLIB'))
    os.environ['PATH'] += ';' + os.path.join(pssepath, 'PSSBIN')
    os.environ['PATH'] += ';' + os.path.join(pssepath, 'PSSLIB')

import psspy
import redirect

redirect.psse2py()

psspy.psseinit(150000)

case = r'your_save_case.sav'
resp = r'your_resp_file.idv'

psspy.case(case) #load the case
psspy.runrspnsfile(resp) #run an idev 
psspy.asys(0,1,[330]) #build a subsystem
psspy.area_2(0,0,1) #load/gen report on that subsystem

If it works, use the API in the documentation to build your own code.

You will likely get an error if you don't have certain Python packages installed. You will need to install these using PIP. (Google 'pip install'.)

You can also use the Recorder in the GUI to build .py code like you would .idv files.

You should know at a minimum the system contribution to three-phase faults (magnitude and angle) and line-to-ground faults. With these, you can use what's in a power system analysis book to derive what you need. You also should consider transient and sub-transient responses and some way to optimize other model parameters to reflect that of the system you're modelling.

There should be a documentation folder with PSSE or in the Start Menu. What you are looking for is in there.

Pro tip: Match the colors so the same bus has the same colors for different plots such as angle and voltage. It may be difficult.

1. Save an uninitialized snapshot then code the channel additions with chsb after you load the snapshot. When complete initialize then run your case. 2. Use matplotlib manual and automatic plot functions to get it like you want it. From experience, about 10 plots per page is reasonable.

By the way, when things to do with file access abruptly stop working, it’s often a file permissions issue. I don’t know about your situation but just a thought.

It says sequence item 0 which may be the first argument. I think there’s an issue with setting outfile higher in the code. Check the variable type of it, should be string like it says. I think outvrsn looks fine for 33.12.

Since the steam source is not modeled in PSSE, use typical steam turbine/governor models. If your unit is a cross compound type it will be a little trickier, and you’ll need to research or post this question separately. If your unit runs with no governor or a valves-wide-open mode, then you may remove the governor model in some cases.