You are on page 1of 21

Practical Methods of

Voltage Stability Analysis


Revisited

IEEE 2002 Summer Meeting, Chicago


Kip Morison
Powertech Labs Inc
kip.morison@powertechlabs.com.
1
IEEE 2002 Summer Meeting, 2002
Objectives

Review a procedure used in VS studies of actual systems

Clarify how analytical methods are used together

Stimulate ideas on improvement or development

2
IEEE 2002 Summer Meeting, 2002
Procedure
1) Define scope

2) Build models (steady-state & dynamic)

3) Conduct preliminary studies (select methods to use)

4) Perform full study and analysis

5) Analysis and Design remedial measures

3
IEEE 2002 Summer Meeting, 2002
Techniques Used
! Steady-state methods (powerflow : P-V Curves)

! Dynamic Analysis (Time-domain simulations)

! Modal analysis of Jacobian

! Optimization (Special OPF)

4
IEEE 2002 Summer Meeting, 2002
1. Define Scope
! Base System conditions (dispatch, topology, load)

! Contingencies

! Transactions

! Criteria

! Determine modeling requirements

5
IEEE 2002 Summer Meeting, 2002
2. Build Models
Dynamic and dynamic models where applicable
! ULTC
! Switchable devices (capacitors, reactors)
! Dynamic VAR sources/sinks (SVC, FACTS, HVdC, etc)
! Generator controls : (OEL / UEL / V/Hz)
! Protections :generator, line, load
! Special protections (i.e. load shedding)
! LOADS
Distribution detail as needed
Explicit representation of caps on LV (such as PF correction)
V and F dependency
Thermostatic effects
Extinction lights
motors

6
IEEE 2002 Summer Meeting, 2002
3. Preliminary Studies
! Use steady-state and dynamic approaches (PV & time-domain)

! Establish mechanism of instability

! Verify that the results, particularly limits, are similar

! Sensitivity studies on devices and load model

! Judge appropriateness / uncertainty in models and make adjustments

as indicated

! Select method & model most suitable for bulk of analysis

7
IEEE 2002 Summer Meeting, 2002
Monday, September 10, 2001, 10:30:19
Buf. Binary Result File Scenario Contingency
1 base-30.bin All ctg at base-30 MW transfer 4 -- G2
2 base.bin All ctg at base transfer 11 -- G2

Dynamic Analysis Bus # Bus Name Buf.


Bus voltage magnitude (pu) 13730 KTA33---33.0 1

(Time-domain) 1.200
13730 KTA33---33.0 2

0.950

0.700

0.450

V Pre-Contg. St a
ble
e
Uns
tabl

Case: 2001wp_wet_2hly
0.200 Contingency: G2
0 48 96 144 192 240


Time in seconds

TSAT
Powertech Labs Inc.


Nanjing Automation Research Institute

Post-Contg. Transient Security Assessment Tool Copyright 2001 All rights reserved

Post-Contg. Margin

Pre-Contg. Margin Steady-state analysis

Po Pcm Pm P
8
IEEE 2002 Summer Meeting, 2002
N I 2001 S U MME R C AS E
Area 11,12,21,22 LOAD - from S . Gen
OTA220--220. B us V oltage

1.02 P re-C ontingency


G4
1
H
0.98 I
J
0.96
p.u. V oltage

0.94

0.92

0.9

0.88

0.86

0.84

0.82
1300

1350

1400

1450

1500

1550

N.LOA D 1600

9
IEEE 2002 Summer Meeting, 2002
Typical Comparison of Static vs Dynamic Methods

0.0 Dynamic VS Margin of Multiple Contingencies


2.8 Dynamic VS Margin of Double Contingencies
5.5 Dynamic VS Margin of Single Contingencies
-3.3 Voltage Limit Margin
2.8 Static VS Margin of Multiple Contingencies
3.3 Static VS Margin of Double Contingencies
7.2
Static VS Margin of Single Contingencies

-5 0 5 10

% Margin

10
IEEE 2002 Summer Meeting, 2002
Load may introduce significant uncertainty
Statistical nature of load (amount and type)
Distributed collection of many devices and controls
Composition (residential, commercial, industrial)
% of motors + type of motors + driven load = equivalent motor ?
Load protection and control

44 kV ? 44 kV
M M
f

M M M M M M

M M M M
f f

M M M M M M M

M M M M

11
IEEE 2002 Summer Meeting, 2002
Sunday, April 14, 2002, 14:49:10
Buf. Binary Result File Scenario Contingency
2 A.bin Ctg A at base transfer 1 -- A

Bus # Bus Name ID Buf.


Induction motor speed (Hz) 13130 HEP33---33.0 A 2
52
15130 MNG33---33.0 A 2
18430 PEN33---33.0 A 2

42

31

21

Motors tripped
in the simulation
10

0
0 24 48 72 96 120

Time in seconds

Powertech Labs Inc.


TSAT
Transient Security Assessment Tool
Nanjing Automation Research Institute
Copyright 2002 All rights reserved

12
IEEE 2002 Summer Meeting, 2002
Sunday, April 14, 2002, 14:44:28
Buf. Binary Result File Scenario Contingency
1 A.bin Ctg A at base transfer 1 -- A

Bus # Bus Name Buf.


Bus voltage magnitude (pu) 13130 HEP33---33.0 1
1.400
15130 MNG33---33.0 1
18430 PEN33---33.0 1

1.120

0.840

0.560

0.280

0.000
0 24 48 72 96 120

Time in seconds

Powertech Labs Inc.


TSAT
Transient Security Assessment Tool
Nanjing Automation Research Institute
Copyright 2002 All rights reserved

13
IEEE 2002 Summer Meeting, 2002
Monday, September 10, 2001, 08:37:02
Buf. Binary Result File Scenario Contingency
1 base.bin All ctg at base transfer 15 -- I
2 base+50.bin All ctg at base+50 MW transfer 14 -- I

Bus # Bus Name Buf.


Bus voltage magnitude (pu) 17860 OTA220--220. 1
1.200
17860 OTA220--220. 2

1.000

0.800

0.600

Case: 2001wp_wet_2hly
0.400 Contingency: I
0 48 96 144 192 240

Time in seconds

Powertech Labs Inc.


TSAT
Transient Security Assessment Tool
Nanjing Automation Research Institute
Copyright 2001 All rights reserved

14
IEEE 2002 Summer Meeting, 2002
MW
Load Variation Stability
Limit

A Original case base

B
Changing motor load percentage (5% increase) -6%

C
Changing motor loading factor (0.7 to 0.6) +5%

D
Changing motor load characteristic ((1-s)2.0 to (1-s)1.8) -10%

E
Changing motor load characteristic ((1-s)2.0 to (1-s)1.6) -9%

F No limit
Changing motor load characteristic ((1-s)2.0 to (1-s)1.5)
found

15
IEEE 2002 Summer Meeting, 2002
Load Modeling

! Derive best approximation of load from composition data (or test)

! Build composite dynamic model

! Simulate step changes at each load bus

! Fit static model to dynamic responses

! Use these models in full analysis

16
IEEE 2002 Summer Meeting, 2002
4. Detailed Study and Analysis
For many large non-soft systems, steady-state analysis is often
sufficient and results are consistent with dynamics

For small or soft systems, motor load is more prone to stall or


near-stall and dynamic simulation may provide more realistic result

However, issues associated with load models must be recognized.


A more complicated but inaccurate model can produce poorer results
than a simple but reasonably accurate model.

17
IEEE 2002 Summer Meeting, 2002
British
Columbia

British
Columbia
Source

North generation
(Source)

Full analysis examines all,


East generation
(Source)
! Conditions
! Contingencies Tie
to
! Transactions Alberta

! Configurations
Lower Tie
Tie to
Mainland
Results, to
USA Load
USA

! Critical conditions
! Critical contingencies
! Stability Limits
! Assessment of cause and remediation

18
IEEE 2002 Summer Meeting, 2002
5. Analysis & Design of Enhancements
! Modal analysis useful for assessing areas prone to collapse
Only useful for critical point (zero eignevalue)
Participations used to site compensation and load shedding sites

! VS margin optimization (such as ROPES) produces similar results and


can also be used for device sizing and control settings

! Time domain used to understand mechanism by examination of control


operation (such as OEL or ULTCs)

Once enhancements are added, verification is done

with static and dynamic simulations

19
IEEE 2002 Summer Meeting, 2002
No. Bus No., Name Area Zone Part.Fac.
--------------------------------------------------
1 18430 PEN33---33.0 12 4 1.00000
2 18420 PEN22---22.0 12 4 0.96349
3 18460 PEN220--220. 12 4 0.56946
4 18450 PEN110--110. 12 4 0.32655
5 18490 PEN-*T10100. 12 4 0.31574
6 18411 PEN11T1011.0 12 4 0.30552
7 18030 PAK33---33.0 12 4 0.29870
8 18006 PAK-ET6-33.0 11 4 0.29738
9 18050 PAK110--110. 12 4 0.26572
10 22330 WIR33---33.0 12 4 0.26123
V Pre-Contg. 11 17852 OTA110-2110. 12 4 0.24721
12 22352 WIR110-2110. 12 4 0.23744
13 22351 WIR110-1110. 12 4 0.23544


14 17815 OTA11-G511.0 12 42 0.23033
15 17814 OTA11-G411.0
16 17816 OTA11-G611.0
12
12
42
42
0.23033
0.17399
Post-Contg.
17 10730 BOB33---33.0 12 5 0.16782
18 10750 BOB110--110. 12 5 0.15753
Modal Analysis 19 16250 MER110--110.
20 22530 WRU33---33.0
13
11
7
3
0.13434
0.11825
Point 21 16230 MER33---33.0 13 7 0.11377

Post-Contg. Margin

Pre-Contg. Margin

Po Pcm Pm P

20
IEEE 2002 Summer Meeting, 2002
Summary
! Different methods can produce unique insight
Steady-state
Dynamics
Modal analysis
Optimization

! Methods must be applied appropriately and used in a complementary


fashion

! Uncertainty in load models and dynamic device models must be


considered care must be taken when using more detail

21
IEEE 2002 Summer Meeting, 2002

You might also like