Benchmark System for SMES

Lchen

Description

Preparation of Transient Simulation Data for PSCAD Relay Case Study of Manitoba Hydro D72V Transient Relay Testing Randy Wachal Ding Lin Manitoba HVDC Research Centre Manitoba Hydro Abstract: This paper presents the procedure for transient PSCAD Relay model. A series of transient fault efficiently preparing the data necessary to perform a cases to represent various fault conditions and current [1]PSCAD Relay transient simulation study for relay flows are described. In general, transient testing allows a [2]testing. An existing 1300+ bus ASPEN Oneliner phasor much more complete suite of cases to be used for testing, based system model was converted to an equivalent 4-bus including such items like applications of the fault at any PSCAD Relay model. A comparison of steady state 60 Hz phase angle, variation in telecommunication and breaker results for both 3 phase and single line to ground faults operating times. A set of fault cases was utilized for verify that the transient and phasor system models are transient testing of line protection for a new 230 kV equivalent. The transient model can be used with transmission line (D72V) recently commissioned in confidence to generate transient fault waveforms simply Manitoba Hydro system. not possible to develop with phasor based simulation method. Transient fault waveforms were used to investigate the operations of the forward and reverse 2. DEVELOPMENT OF SYSTEM MODEL ground ...

Informations

• Publié par: Zien
• Nombre de lectures: 43
• Langue: English

Extrait

Preparation of Transient Simulation Data for PSCAD Relay

Case Study of Manitoba Hydro D72V Transient Relay Testing

Randy Wachal

Manitoba HVDC Research Centre

Ding Lin

Manitoba Hydro

Abstract:

This paper presents the procedure for

efficiently preparing the data necessary to perform a

PSCAD Relay

[1]

transient simulation study for relay

testing. An existing 1300+ bus ASPEN Oneliner

[2]

phasor

based system model was converted to an equivalent 4-bus

PSCAD Relay model. A comparison of steady state 60 Hz

results for both 3 phase and single line to ground faults

verify that the transient and phasor system models are

equivalent. The transient model can be used with

confidence to generate transient fault waveforms simply

not possible to develop with phasor based simulation

method.

Transient

fault

waveforms

were

used

to

investigate the operations of the forward and reverse

ground directional overcurrent elements of the Nxtphase

L-PRO line protection relay.

Keywords

Relay Testing, Transient Simulation, PSCAD

Relay, System Equivalent, ASPEN Oneliner (ASPEN)

1.

INTRODUCTION

Transient testing of protection relays with waveforms of

the same quality and frequency response of the voltage and

current waveforms the protection uses from the system PT

and CT is becoming increasingly important

[3]

. This is true

especially as the speed and complexity of the digital

protection system increases. A form of dynamic testing has

been

developed based on the steady

state phasor solutions.

The pre-fault, fault and post fault steady state phasors

would be calculated using a phasor based simulation

program like ASPEN or PTI PSS/E. The prefault, fault

and post fault phasors for voltage and current would be

converted into time domain waveforms and then simply

concatenated together to create a type of dynamic

changing time domain waveforms. This dynamic STATE

testing ignores any transient effects when the fault was

applied or removed and works reasonable well there is a

high level of filtering applied by the protection relay.

A more accurate representation of the transient waveforms

is to simulate the power system using a time domain

simulation program to directly develop the transient fault

waveforms. These waveforms include all of the transient

effects.

One of the difficulties encountered is developing

the data necessary for time domain or PSCAD Relay

simulation. In many utilities there is a large database of

phasor based simulation models, which have been

developed over a period of many years.

This paper

illustrates the process of converting and validating the

existing phasor based ASPEN system model into a

transient PSCAD Relay model. A series of transient fault

cases to represent various fault conditions and current

flows are described. In general, transient testing allows a

much more complete suite of cases to be used for testing,

including such items like applications of the fault at any

phase angle, variation in telecommunication and breaker

operating times.

A set of fault cases was utilized for

transient testing of line protection for a new 230 kV

transmission line (D72V) recently commissioned in

Manitoba Hydro system.

2.

DEVELOPMENT OF SYSTEM MODEL

2.1 System Model

PSCAD/Relay case for the system under investigation was

developed from an ASPEN case. ASPEN is a fundamental

frequency fault program, used routinely by Manitoba

Hydro for protection studies. Manitoba Hydro maintains a

relatively large (1300+ bus) system model in ASPEN. The

conversion of a large system into a transient simulation can

be a significant effort. For the D72V test program, the

ASPEN system model was converted into a 4-bus PSCAD

Relay case using equivalent voltage sources at each bus to

represent the remaining system. A comparison between

results from the PSCAD Relay case for three phase and

single line to ground (SLG) faults at each bus and the

ASPEN simulation was performed with matching results.

This validation verified the system equivalence techniques

used to reduce the system size and the system model

conversion.

2.2 Procedure for ASPEN Equivalence Network and

Conversion

The Manitoba Hydro ASPEN system models consist of

approximately 1300 busses. This system was converted to

a 4-bus system including eight 3-phase transmission

sections and three 6-phase transmission sections.

A 6-

phase line section includes the mutual coupling effects

when two 3-phase circuits share the same tower.

The

PSCAD Relay Case developed for this testing is shown in

Figure 1.

A step-by-step illustration of the process of developing

equivalent sources at the 4 bus locations within the

ASPEN program is described in details in its on-line help

menu as well as its user manual (Reference 2: Appendix

- 1 -

G). This process is relative easy and would require less

than an hour of time for any user with some familiarity of

using the ASPEN program. Prior to proceeding to

conversion to PSCAD, it is important to ensure the faults

results generated in the full ASPEN system are the same as

the same fault case in the equivalence or reduced ASPEN

system

Once the equivalent electrical system is developed and

validated in ASPEN, this data is used to develop the

PSCAD Relay case. It is possible to develop a PSCAD

Relay case from a blank sheet but it is much quicker to

select a base PSCAD Relay from the prepared examples.

This example case is then modified into the study case.

Additional Transmission lines, breakers and voltage

sources can be added by copy and paste commands. The

following steps illustrate the process.

Step 1: Select the appropriate PSCAD Relay Example

case.

Step 2: Enter the Positive and zero sequence impedance for

each voltage source. Add additional voltages

sources as required.

Step 3: Enter the transmission line data parameters either

using the direct R, X, B values from the ASPEN

model or if available, transmission tower geometry

and conductor information in a PSCAD traveling

transmission line traveling wave model. If mutual

coupled transmission are utilized remember to input

transmissions as 6 or more conductor elements.

PSCAD supports mutual coupling of up to 20

conductors.

Add additional transmission lines as

required.

Step 4:Add Coupled Pi branch sections to accommodate

the fictitious branch data generated by the ASPEN

Equivalence procedure. This data will

have series R

and X but no shunt B data.

Step 5: Run the PSCAD solutions with no faults applied

and adjust the voltage source magnitude and angle

to give the desired prefault bus voltages and power

flow.

At this point the PSCAD Relay system model is ready for

comparing steady state faults results with results from

ASPEN case or to proceed with development of transient

test waveforms. Permanent single and three-phase faults

were applied and compared with steady state solutions

with ASPEN results for the same case.

2.3 Validation of Transient System Model

In order to compare PSCAD and ASPEN results it is

important to remember ASPEN simulation results can be

shown as phase or sequence quantities and that these

results are steady state in nature. PSCAD provides a time

domain voltage and current waveform similar to what can

be measured on the power system. In order to compare

ASPEN and PSCAD results, the time domain waveforms

must be converted to a phasor equivalent. Within PSCAD

there are “RMS” measurement blocks and 3 phase on-line

- 2 -

FT6

F6

F2

F7

FT7

F4

St. Vital 230 kV

Bus1

F5

FT5

V

Ph

230.0 [kV], 60.0 [Hz]

100.0 [MVA]

Z1 = 71.73 [ohm] /_ 47.39 [°]

V

Ph

230.0 [kV], 60.0 [Hz]

100.0 [MVA]

Z1 = 109.05 [ohm] /_ 62.52 [°]

Rosser 230 KV

Bus 3

Dorsey 230 kV

Bus 2

F1

FT1

V

Ph

230.0 [kV], 60.0 [Hz]

100.0 [MVA]

Z1 = 43.05 [ohm] /_ 63.9 [°]

V

Ph

230.0 [kV], 60.0 [Hz]

100.0 [MVA]

Z1 = 6.11 [ohm] /_ 84 [°]

V4

V5

F3

FT3

V1

FT4

3 Phase

RMS

V4rms

B3

27.16 [MVAR]

-94.01 [MW]

V6

Aspen1

B1

0.005922 [MVAR]

-0.0001996 [MW]

15 km

D36R1

B2

0.01282 [MVAR]

0.001017 [MW]

19.46 km

D5R

19.46 km

D13RD16R

B5

32.3 [MVAR]

-150.7 [MW]

B6

-31.4 [MVAR]

151.2 [MW]

B4

-32.35 [MVAR]

95.48 [MW]

V

3

r

m

s

3

P

h

a

s

e

R

M

S

V3

FT8

Aspen2

Aspen4

Ridgeway 230 kV

Bus 4

19.40 km

R32VD72V

Aspen4

16.41 km

R23R

20.07 km

D36RD72V

16.4 km

D72V_1

3

P

h

a

s

e

R

M

S

V5rms

3

P

h

a

s

e

R

M

S

V1rms

3

P

h

a

s

e

R

M

S

V6rms

3

P

h

a

s

e

R

M

S

V2rms

V2

FT2

F8

19.43 km

R33V

Figure 1: PSCAD Relay System

SLG Fault at Dorsey

Voltage

(V0)

at:

ASPEN

PSCAD*

*

Differenc

e

between

ASPEN

&

PSCAD**

% Error

between

ASPEN

&

PSCAD*

*

Dorsey

Bus 2

12.6

13.18

0.58

4.6%

Ridgeway

Bus 4

2.9

4.174

1.27

43.9%

Rosser

Bus 1

5.4

6.675

1.28

23.6%

St Vital

Bus 3

1.6

2.561

0.96

60.1%

Current

(3I0)

:

Bus 3 R33V

I1

28

17.54

-10

-37.4%

Bus 2 D36R

I2

297

412.9

116

39.0%

Bus 3 D72V

I3

248

298.6

51

20.4%

Bus 2 D72V

I4

248

296.4

48

19.5%

Bus 4 D36R

I7

297

414.5

118

39.6%

Bus 2 D13R

I8

387

424.9

38

9.8%

I fault

34114

36250

2136

6.3%

I1a

I1b

I1c

1

I1+

1

I1+ph

1

I1_0

1

I10ph

XA

XB

XC

Ph+

Ph-

Ph0

Mag+ Mag- Mag0

(7)

(7)

(7)

(7)

(7)

(7)

dcA

dcB

dcC

F F T

F = 60.0 [Hz]

1

Figure 2: PSCAD FFT Block with Sequence Outputs

FFT processing blocks that can provide positive, negative

and zero sequence information. Figure 2 shows the

PSCAD FFT block.

The results for comparison between the full (1300+ Bus)

ASPEN and the reduced (4-bus) PSCAD system illustrate

a close match for the positive and zero sequence voltages,

branch currents and fault currents. Samples of results for a

SLG and 3-phase fault are presented for a fault on Dorsey

bus are presented in Table 1 and 2. Results are presented

in both absolute value and % error. Care in interpreting

results is required. For example, in the SLG fault case the

zero sequence voltages at the non-faulted busses show a

large percentage error, while the absolute values are within

a very acceptable 1.3 volts.

Table 1:

Single Line to Ground Fault at

Dorsey Bus

3 Phase Fault at Dorsey

Voltage

(V+)

at:

ASPEN

PSCAD

Difference

between

ASPEN &

PSCAD

% Error

between

Aspen &

PSCAD

Dorsey

Bus 2

0

0

0

0.0%

Ridgeway

Bus 4

36.1

36.8

0.7

1.9%

Rosser

Bus 1

23.4

24.5

1.1

4.7%

St Vital

Bus 3

45.1

45.77

0.67

1.5%

Current

(I+):

Bus 3 R33V

I1

974

963

-11

-1.1%

Bus 2 D36R

I2

2177

2223

46

2.1%

Bus 3 D72V

I3

1759

1779

20

1.1%

Bus 2 D72V

I4

1759

1782

23

1.3%

Bus 1 R23R

I5

1600

Bus 3 R32V

I6

963

933

-30

-3.1%

Bus 4 D36R

I7

2177

2221

44

2.0%

Bus 2 D13R

I8

2491

2609

118

4.7%

I fault

37484

37150

-334

-0.9%

The minor differences can be attributed to the following:

1.

2.

3.

4.

Equivalence:

Results for

ASPEN system are 1300+

busses, while PSCAD are for the 4 bus system.

Note: When a 4-bus ASPEN system was solved the

results between ASPEN and PSCAD are within 1%.

Prefault load flow:

ASPEN has fault calculations

performed from a flat start position, while PSCAD

solves the system.

Even when the power flow is

reduced to zero, or near zero, the effects of the

transmission line charging are present.

Transmission

lines

are

not

identically

modeled.

ASPEN uses a coupled pi model with lumped R, X and

B values.

PSCAD calculates the traveling wave

parameters for the line based on geometrical conductor

configuration and conductor data. The 60 Hz lumped

parameters calculated by PSCAD are close but not

precisely the value used in ASPEN.

Table 2: Three-Phase Fault at Dorsey Bus

Mutual Coupling. The mutual coupling for some other

transmission lines on the same right of way as D72V

were not modeled in PSCAD but in ASPEN, because

the geometry data for these lines was not readily

available.

- 3 -

3. Development of Transient Test Cases

3.1 The Problem

D72V is a new transmission line with portion of it

constructed on the same towers of an existing line, and on

the same right of way (ROW) with some additional

existing lines. During state simulation testing of the relay,

the directional ground overcurrent elements of the relay

were giving some questionable results for some current

reversal conditions due to mutual coupling effect. It was

not clear whether the operation of these fast reacting

elements is affected by the unrealistic simulation of the

transition between states, or by different fault conditions

such as fault inception angle or prefault line loading. The

sensitivity of the forward and reverse ground overcurrent

elements 67F and 67R of the Nxtphase L-PRO relay on the

new D72V line was the focus of this transient testing

program.

3.2 The Test Plan

A number of PSCAD/Relay simulations were performed to

generate the required testing waveforms. An “A” phase to

ground fault was applied at the Ridgeway end of D36R, at

Fault Location F3 on Figure 1, in order to produce a

forward reverse current flow on D72V. The application of

fault angle was modified from 0 to 180 degrees in 30-

degree steps; and the power flow from Dorsey to St. Vital

on D72V was adjusted from 0, 100 and 200 MW.

In

addition, the telecommunications delay between line D36R

breaker opening at the Ridgeway, B1 shown in Figure 1,

and the breaker opening at the Dorsey end, B2 shown in

Figure 1, was selected at 30 or 100 msec. This set of tests

was performed using the multiple run feature of PSCAD,

generating a total of 42 test cases.

Each test case

generated the three voltage and three current signals

required for transient testing of the Dorsey and St Vital

D72V protection system. An example of the waveforms is

shown in Figure 3.

Initially 200 MW is flowing on D72V. A SLG fault is

applied at Ridgeway end of the D36R line. The voltage

and currents presented are recorded at the Dorsey end of

D72V. When the fault is applied, the D72V relay at

Dorsey end sees reverse current. The Ridgeway breaker

opens 50 msec (3 cycles) after the fault, changing the

direction of the current as seen at the Dorsey end of D72V.

The breaker on D36R remote from the fault opens 30 msec

after the local end (approximately 2 cycles) and removes

the fault current flow from D72V.

These faults waveforms were used for real time transient

testing of the D72V.

The overall development time for

PSCAD Relay Case development, validation with ASPEN

steady state and transient case study plan was a couple of

days, with the bulk of effort in the validation testing.

3.3 Results of the Testing Program

The transient waveforms were played into a Nxtphase L-

PRO relay configured with the appropriate setting D72V

- 4 -

DorseyD72V-2002-05-16_13.18.31.071 : 2002-05-16 08:18:31 .071 -- 51NAlarm

-10.0

6.0

Dorsey

2002/May/16 08:18:31.071181

Rosser

D72V Line Current A

-4.0

4.0

Dorsey

2002/May/16 08:18:31.071181

Rosser

D72V Line Current B

-4.0

4.0

Dorsey

2002/May/16 08:18:31.071181

Rosser

D729 Line Current C

-5.0

6.0

Dorsey

2002/May/16 08:18:31.071181

Rosser

3Io_Main

L

L

L

ProLogic 1

ProLogic 2

Comm. Scheme Send

Figure 4: Sample Relay Recording at D72V Dorsey

Figure 3: Sample Transient Test Waveforms

Voltage and Current at D72V Dorsey

files. Figure 3 illustrates the transient waveforms generated

by PSCAD Relay, which were played into the relay.

Figure 4 shows a set of sample waveforms recorded by the

L-PRO relay. The operation of the 67F and 67R elements

was verified over a large number of cases during a one-day

testing period. The transient testing program confirmed

that the relay operation was not dependent on the prefault

loading,

fault

inception

angle

or

the

protection

telecommunication delay on the faulty line, but the level of

positive sequence component of the fault current has an

effect on

the operation of the directional ground

overcurrent elements.

4. CONCLUSIONS

Transient simulation testing of protection offers many

advantages over the more traditional methods. Since the

transient waveforms produced represent realistic voltage

and current waveforms that the protection sees in service,

the overall confidence in the testing results is greatly

increased.

The process to develop a transient system

simulation model from a phasor-based system is not

difficult.

With PSCAD/Relay, it was possible to develop a study

system that produced the same results as a fundamental

frequency program. Once the positive and zero sequence

networks were confirmed, the development of particular

study

cases

of

interest

was

performed.

These

PSCAD/Relay generated waveforms were injected into the

protection system using a real time transient playback

system, allowing a thorough confirmation of the relay

performance. The development of a transient test plan can

be performed within PSCAD Relay with minimum effort.

These transient test waveforms can be used to verify the

relay performance with confidence for either single or GPS

based end-to-end testing.

The operation of the Nxtphase L-PRO relay was verified

over a large number of cases during a one-day laboratory

testing period.

5. REFERENCE

[1]

“PSCAD/Relay

Installation

and

Operations

Manual”, Manitoba HVDC Research Centre, Aug

2001.

[2]

“ASPEN Oneliner V2001 User’s Manual”

[3]

M.S. Sachdev, T.S. Sidhu, P.G. McLaren, Issues

and Opportunities for Testing Numerical Relays,

IEEE Power Engineering Society Summer Meeting,

Seattle, Washington, USA, 16 – 20 July 2000.

- 5 -