AC Stalling Simulations (1)

7/28/2019 AC Stalling Simulations (1)

1/31

AC Stalling Computer

Simulations usingEMTDC/PSCAD


2/31

* EMTDC/PSCAD Model of the SCE 500 kV System andValley 115 kV System

* Computer model development and validation againstShort Circuit Program, Digital Fault Recorders (DFR)captured disturbances and PMU events

* SCE 500 kV System bus voltages after clearing faults onat SCE 500 kV buses computer simulations


3/31

THE EMTDC/PSCAD MODEL


4/31

SCE 500kV PSCAD SYSTEM MODEL


5/31

VALLEY (500 and 115 kV) SYSTEM PSCAD MODEL


6/31

THE PSCAD MOTOR LOAD AT VALLEY 115/12kV

* Air Conditioning Motor electrical characteristics from reference [1]


7/31

THE 500kV SYSTEM MOTOR LOAD


8/31

MODEL DEVELOPMENT

Generate a short circuit model reduction from CAPE software usingSCE protection engineering data base

Setting up initial conditions using PSLF load flow by adjusting thesource equivalent voltage and angle

System validation by comparing short circuit results between CAPEand PSCAD models

Comparing actual captured faults from DFR records and PMUevents


9/31

SHORT CIRCUIT SOURCE EQUIVALENT USING CAPE

MODEL REDUCTION

BUS X+ R+ Z+ Angle+ X0 R0 Z0 Angle0

(Ohms) (Ohms) (Ohms) (deg) (Ohms) (Ohms) (Ohms) (deg)

Vincent 2.94 51.92 51.999 86.759 0.83 24.80 24.816 88.090

Lugo 0.84 17.63 17.647 87.275 2.11 24.10 24.193 84.993

Mira Loma 1.14 39.87 39.888 88.356 0.27 22.48 22.477 89.315

Serrano 1.69 48.44 48.471 88.004 1.15 26.21 26.235 87.492

Valley 23.84 1207.79 1208.022 88.869 1.90 52.39 52.421 87.927

Devers 6.51 96.74 96.963 86.151 0.07 44.92 44.918 89.908

Palo Verde 0.32 6.91 6.920 87.385 0.24 5.10 5.107 87.346

The PSCAD model is designed based on a reduced network model fromCAPE (Protection Engineering Short Circuit Program).


10/31

INITIAL CONDITIONS (STEADY STATE) FROM LOAD

FLOW PROGRAM PSLF

Voltage Voltage Angle

BUS P (MW) Q(MVAR) P Q (p.u.)* (kV) (degrees)

(p.u. on 100MVA) (p.u. on 100 MVA)

VINCENT 1618.2 83.6 16.182 0.836 1.044 522 -13.982

LUGO 3736.1 338.4 37.361 3.384 1.035 517.5 -18.792

MIRA LOMA -3071.2 -195.6 -30.712 -1.956 1.024 512 -26.297

SERRANO -1556.7 263 -15.567 2.63 1.029 514.5 -29.219

VALLEY -1465.6 -79 -14.656 -0.79 1.028 514 -32.836

DEVERS -754.1 283.8 -7.541 2.838 1.038 519 -28.687

PALO VERDE 1587.5 266.1 15.875 2.661 1.07 535 -3.843

The initial conditions maybe take from a load flow program. In our casethe PSLF load flow is used as the initial condition for the PSCADsimulations.


11/31

THE VALIDATION PROCESS


12/31

SHORT CIRCUIT VALIDATION CAPE VR. PSCAD

A well know method to validate EMTP-type model is to compare theShort Circuit results between the EMTP type program (PSCAD) andShort Circuit program (CAPE)


13/31

VALIDATION AGAINST STATION DFR DISTURBANCE

(faults in the Valley 115 kV system)

SLG Fault on the secondary of Valley 2AA bank cleared inapproximately 3 cycles

-3

-2

-1

0

1

2

3

60 80 100 120 140 160 180

Valley 500/115 kV Tranformer Bank Current

Electrotek Concepts TOP, The Output Processor

agntue

ag

Time (ms)

DFR Record PSCAD Simulation

SLG Fault on 115kV Valley system at 115kV Skylard-Newcomb line

cleared in 6 cycles

-1.5

-1.0

-0.5

0.0

0.5

1.0

1.5

50 100 150 200

Valley 2AA Tranformer Bank Current


MagnitueMag

Time (ms)

DFR PSCAD


14/31

JULY 1, 2006 DISTURBANCE FROM PSCAD AND DFR

(fault at the 500 kV System)

Fault on the Devers-Palo Verde 500 kV transmission line 20 milesapproximately from California Caps on the Arizona side captured byDevers DFR.

-15

-10

-5

0

5

10

15

50 100 150 200

DEVERS - PALO VERDE PHASE A


Magnitude(kMag

Time (ms)

DEVERS DFR PSCAD

-20

-10

0

10

20

60 80 100 120 140 160 180

DEVERS PALO VERDE PHASE C CURRENT


Magnitude(kMag

Time (ms)

DFR PSCAD

To Devers To Palo Verde


15/31

HOW DO THESE FAULTS IMPACT THE REST OF THE500 kV SYSTEM?


16/31

FAULTS AND THE IMPACT OF THE 500 kV SYSTEM

The question may be answered by comparing the magnitude of the voltages fromthe PMU record before and after the fault is cleared. The in between points (duringfault) maybe more of a challenge but we will discuss it in a few slides later.

Pre-Fault

Post-Fault

During the fault


17/31

LOAD FLOW FROM PMU DATA

In order to setup the initial conditions to the PMU load flow the PMU BusVoltages and Angles are utilized instead of the PSLF


18/31

PRE-FAULT BUS VOLTAGES PSCAD VR. PMU(J uly 1, 2006 Disturbance)

Pre-Fault PSCAD vr. PMU Bus Voltages

400

420

440

460

480

500

520

540

560

Vincent 500kV Valley 500kV Lugo 500 kV Mira Loma 500 kV Serrano 500 kV

Voltage (kV)

PMU

PSCAD

Snap shot (voltage profile) of the SCE 500 kV system bus voltages before thedisturbance from the PMU and PSCAD


19/31

POST-FAULT BUS VOLTAGES PMU VR. PSCAD

(July 1, 2006 Disturbance)

Post-Fault B us Voltage PMU vr. PSCAD

300

350

400

450

500

550

Vincent 500kV Valley 500kV Lugo 500 kV Mira Loma 500 kV Serrano 500 kV

Voltage (kV)

PMU

PSCAD

Snap shot (voltage profile) of the SCE 500 kV system right after the fault is clearedfrom the PMU and PSCAD

3.5% error difference


20/31

WHAT ABOUT DURING THE FAULT?

During the fault as before mentioned we have to be careful when comparing thecomputer simulation to the actual PMU record. The PMU samples at 30Hz and inorder to look at the PSCAD simulation the PSCAD recorder has to be set of sampleat the same rate or the result may show a great difference. We must understand thatthe main purpose of the PMU is to look at slow time frame disturbance and not fast

transients.

Simulation is kept at small step

Plot at smaller step

PSCAD Recordersame as PMUSampling rate

COMPARING DATA PSCAD AND PMU AT SAME


21/31

COMPARING DATA PSCAD AND PMU AT SAMESAMPLE RATE

As show on the plots below the bus voltage magnitudes are closely correlatedbefore, during and after the fault is cleared


22/31

REMEMBER THE PMU BUS MAGNITUDE AT DEVERSDURING THE FAULT WENT TO ZERO?

CONVERTING DFR RECORD INSTANTANEOUS BUS VOLTAGE


23/31

WAVEFORMS TO POSITIVE SEQUENCE MAGNITUDE

Using a Matlab-Simulink developed tool the instantaneous values from the DFRrecord are converted to positive sequence magnitude (similar to PMU but fartersampling rate)


24/31

PSCAD SIMULATION VR. DFR RECORD

DFR Record

PSCAD Simulation

So lets compare it now to what our PSCAD simulation results show. On the plotbelow we compare the PSCAD simulation with a sample rate of 50 us to the DFRrecord positive sequence from Devers magnitude.


25/31

System Fault versus Motor Stalling


26/31

1 2 3 4 56 7 8 9 10 11 12 13 14 15 16 17

18S1

S3

S5

0.00

0.20

0.40

0.60

0.80

1.00

1.20

Voltage (Per Unit)

System Voltage After Clearing Fault

1.00-1.20

0.80-1.00

0.60-0.80

0.40-0.60

0.20-0.40

0.00-0.20


27/31

Motor Speed on All 500 kV Buses

R l d R ti P


28/31

LUGO : Motor Load LUGO

0.40 0.60 0.80 1.00 1.20 1.40 1.60 1.80

0

25

50

75

100125

150

175

200

RealPower(M

W)

Pmot_lug

-100

0

100

200

300

400

500

600

700

ReactivePower(MVAR

)

Qmot_lug

Real and Reactive Power


29/31

FUTURE WORK


30/31

ANY QUESTIONS?


31/31

* IEEE Transactions in Power Systems Air ConditioningResponse to Transmission Faults: J ohn W. Shafter, May

1997

* IEEE Transactions in Power Systems Motor Load

Dynamics: Impact on Voltage Recovery PhenomenaGorge K. Stefopous, Georgia Tech, 2006.

References

AC Stalling Simulations (1)

Documents