PSS2C
PSS2C model
This article is incomplete, some sections must be written.
Context
This power system stabilizer model first appeared in the IEEE Std 421.5-2016 (of Electrical & Engineers, 2016). It has been reproduced identically in the IEC 61970-302:2024 version (Commission, 2024). In the previous standard versions, its predecessor models were called PSS2A (1992, 2005) and PSS2B (2005). Compared to PSS2A, PSS2B has a third lead-lag compensation block. Compared to PSS2B, PSS2C has a fourth lead-lag compensation block and an output logic.
Model use, assumptions, validity domain and limitations
To be completed
Model inputs and output
The input variables are :
Variable | Description | Units |
---|---|---|
omegaPu | measured angular frequency | pu (base omegaNom) |
omegaRefPu | reference angular frequency | pu (base omegaNom) |
PGenPu | active power | pu (base SnRef) |
The output signal is VPssPu in pu (base UNom).
Model parameters
Parameter | Description | Units |
---|---|---|
KOmega | Coefficient applied to angular frequency | - |
KOmegaRef | Coefficient applied to reference angular frequency | - |
Ks1 | Gain of power system stabilizer | pu |
Ks2 | Gain of transducer (active power branch) | pu |
Ks3 | Washouts coupling factor | pu |
M | Lag order of ramp-tracking filter | - |
N | Order of ramp-tracking filter | - |
OmegaMaxPu | Maximum angular frequency input of power system stabilizer | pu (base omegaNom) |
OmegaMinPu | Minimum angular frequency input of power system stabilizer | pu (base omegaNom) |
PGenMaxPu | Maximum active power input of power system stabilizer | pu (base SNom) (generator convention) |
PGenMinPu | Minimum active power input of power system stabilizer | pu (base SNom) (generator convention) |
PPssOffPu | Active power threshold for PSS deactivation | pu (base SNom) (generator convention) |
PPssOnPu | Active power threshold for PSS activation | pu (base SNom) (generator convention) |
t1 | First lead time constant | s |
t2 | First lag time constant | s |
t3 | Second lead time constant | s |
t4 | Second lag time constant | s |
t6 | Transducer time constant of angular frequency branch | s |
t7 | Transducer time constant of active power branch | s |
t8 | Ramp-tracking filter lead time constant | s |
t9 | Ramp-tracking filter lag time constant | s |
t10 | Third lead time constant | s |
t11 | Third lag time constant | s |
t12 | Fourth lead time constant | s |
t13 | Fourth lag time constant | s |
tW1 | First washout time constant (angular frequency branch) | s |
tW2 | Second washout time constant (angular frequency branch) | s |
tW3 | First washout time constant (active power branch) | s |
tW4 | Second washout time constant (active power branch) | s |
VPssMaxPu | Maximum voltage output of power system stabilizer | pu (base UNom) |
VPssMinPu | Minimum voltage output of power system stabilizer | pu (base UNom) |
SNom | Nominal apparent power | MVA |
Model diagram
Model variants
In the PSS2A and PSS2B models :
- the PSS deactivation for low active power values is absent (PPssOffPu = -1000, PPssOnPu = -999)
- the final lead-lag filter is ignored (t12 = t13 = 0)
Moreover, in the PSS2A model, the second to last lead-lag filter is ignored (t10 = t11 = 0).
Open source implementations
This model has been successfully implemented in :
Software | URL | Language | Open-Source License | Last consulted date | Comments |
---|---|---|---|---|---|
Dynawo | Link | Modelica | MPL v2.0 | 24/05/2024 |
References
- of Electrical, T. I., & Engineers, E. (2016). IEEE recommended practice for excitation system models for power system stability studies . IEEE Std 421.5-2016. https://home.engineering.iastate.edu/ jdm/ee554/IEEEstd421.5-2016RecPracExSysModsPwrSysStabStudies.pdf
- Commission, I. E. (2024). Energy management system application program interface (EMS-API) Part 302: Common information model (CIM) dynamics. IEC 61970-302. https://webstore.iec.ch/preview/info_iec61970-302%7Bed2.0%7Db.pdf