Ask Your Question
0

Playback Model

asked 2024-09-26 12:58:21 -0500

RedOne gravatar image

updated 2024-09-26 13:22:32 -0500

Hi everyone,

I am attempting to run a dynamic simulation with the PSSE model. Specifically, I am simulating the frequency response of the model and comparing it with the measured data.

I have used the Playback model PLBVFU1 to apply the measured voltage at the HV level, and I adjusted the frequency to match the one forced on the PPC.

I am sharing the results with you: https://*ibb.co/ctqsnHp (delete *)

the model performs well for active power, but the reactive power does not match the measurements and varies significantly. Additionally, the voltage at the PCC has changed a little bit.

Could you assist me with this and provide any insights or suggestions for improvement?

edit retag flag offensive close merge delete

2 answers

Sort by » oldest newest most voted
0

answered 2024-10-02 11:55:13 -0500

RedOne gravatar image

updated 2024-10-02 11:58:08 -0500

Thank you, Jconto, for your answer.

First, I’d like to explain what I’m trying to simulate with my model. It’s a solar park modeled using generic models: REPCAU1 for the PPC and REECAU1/REGCAU1 for the converters. I’ve also modeled the equivalent collector system (cables, padmount transformer, and the main power transformer). The voltage is controlled on the secondary of the 138/34.5 kV transformer.

I’m measuring the frequency on the high voltage side. I’ve modeled a generator connected to the POI and used a playback PLBVFU1. I used a plb file with the average value of the three phases from meausrements and the forced frequency. No transmission line has been modeled.

The Qg measurement is not completely flat; it does respond to disturbances. In another simulation, i adjusted the voltage reference, and the plant reacted by absorbing or injecting reactive power. All recorded data comes from the PMU. Pg and Qg remain flat when there’s no disturbance.

I tried a few things to address the issue:

I deactivated the network frequency dependence (NETFRQ), which I had previously set to 1 in my Python script. I think this has an impact especially when there's is only renewables models and lines are not modeled...

I used an acceleration factor of 1.0 instead of 0.5.

The new results are somewhat closer to what I’m expecting from the simulation : https://*ibb.co/hmRng7x (delete *)

I believe the model still needs some fine-tuning to get even more accurate results. I’m not sure how to do that yet, but if you have any suggestions, they’d be greatly appreciated.

By the way, what is an ERUN test? I have no idea what that is.

edit flag offensive delete link more

Comments

Disregard my comment about erun since your model is reecau1. (More info about erun can be found by searching the API or the Program Application Guide vol2 manuals for 'erun').

jconto gravatar imagejconto ( 2024-10-02 13:05:44 -0500 )edit

Thank you, Jconto, for your help. I've successfully adjusted the model's parameters to achieve results that closely align with the measured data

RedOne gravatar imageRedOne ( 2024-10-07 07:45:04 -0500 )edit
0

answered 2024-09-26 18:13:44 -0500

jconto gravatar image

For a frequency event playback outcome, the goal is to get network simulated frequency response matching measurements. Get the simulated frequency from a bus close to the station where frequency was recorded.

In your plots, Qg-measurement is flat, while Qg-simulated is not. In general, a flat Qg output is not good, because generators should provide voltage support to any disturbance in the network. Homework: check the metering system for that machine and confirm that it provide Q support for other events.

Recorded P & Q of a generator (usually scada, better if pmu data is available) can be consider to be the 'reference' value and the simulated P & Q of that machine be the variable value. You can tune the generator's exciter model by changing parameter values to match the Qg-simulated to the Qg-measurement response. You can use the erun test for this tuning exercise.

Lastly, the flat period in the simulation (0-5sec) for Pg is flat as expected but not for Qg. Perform a 10-sec flat run (no disturbance) to get flat lines on all variables. Tune models to get these flat lines as much as possible!

edit flag offensive delete link more

Your Answer

Please start posting anonymously - your entry will be published after you log in or create a new account.

Add Answer

[hide preview]

Question Tools

2 followers

Stats

Asked: 2024-09-26 12:58:21 -0500

Seen: 133 times

Last updated: Oct 02