Real-time Stability Management Under

Stressed Conditions: Voltage Stability

Damir Novosel, Quanta
Vahid Madani, PG&E
Mevludin Glavic, Quanta
Amsterdam, October 2012
PG&E and Quanta Technology LLC Proprietary and Confidential

Voltage Related Issues

Voltage Instability:
Short- and Long- Term

Fault Induced Delayed Voltage

Recovery (FIDVR)

PG&E and Quanta Technology LLC Proprietary and Confidential

Page 2

Wide Area Events: Voltage Instability

Inability to maintain voltage so that power and voltage
are controllable

High system loading,

followed by a fault,
line overload or
generators hitting
excitation limits

PV Curve

Voltage

Pre-contingency
case

Post-contingency
case

Grid gets overloaded, more

reactive power is consumed
and voltages drop

Margin
Margin
108% Peak
100% Peak

Followed by motors stalling

108% Peak
- in 3 years

PG&E and Quanta Technology LLC Proprietary and Confidential

Real
Power

Page 3

4
Disturbances: Understanding What Happened

Power - Voltage Curve

Voltage (kV)

245
240

235
230
225
220
170

175

180

185

190

195

200

205

210

215

220

Power (MW)

Alarm when measured PV-curve slope approaches a critical level

Source: Dmitry Kosterev, BPA

PG&E and Quanta Technology LLC Proprietary and Confidential

Page 4

Dynamic Reactive Power: Practical Measures

Additional generation and transmission lines

Switching on shunt capacitors and

switching off shunt reactors

Exhaust generation reactive

resources, e.g. AVR set points

Dynamic reactive power devices

On-line tap changer control

Fast unit start up

Demand Side Management

Wide Area Monitoring, Protection

and Control Control

Some voltage collapse examples:

Brazil-Paraguay, 11/11/09
Greece, 07/12/04
NE US and Canada, 8/14/03
WECC, 7/2/96
Japan 7/23/87 and 8/22/70
Sweden, 12/27/83
France, 12/19/78

PG&E and Quanta Technology LLC Proprietary and Confidential

Page 5

Voltage Stability Monitoring & Assessment

x

Undervoltage
signal
#1
r

Voltage
#2
instability
region
#1: inaccurate under-voltage detection
#2: under-voltage fails to detect

Simple voltage-only measurements

Dynamic, model-based simulation tools

Voltage Stability Assessment (VSA) based on State Estimation

contingency analysis

Tracking the relative distance from voltage instability continually in

real-time:

Distance to the nose of the PV curve

State Estimation based stability boundary

Important to validate model correctness

Measurement-based indicators

Source: ABB

Identify composite low voltage violations simultaneously over a broad

area

Monitor available reactive power levels (capacitor/reactor reserves,

tap-changers)

Distance of the loads apparent impedance to the Thevenin

impedance

PG&E and Quanta Technology LLC Proprietary and Confidential

Page 6

Measurement-Based Voltage Instability Detection

Simple, real time, voltage instability margin

detection, e.g. VIP*

Better than voltage-only methods, but simpler

than any other method
Much faster than EMS/SE contingency analysis
and not model dependent

Enabling tracking both slow changes and

system dynamics using PMUs (10-120
frames/s)

New Real-time Voltage Instability Indicator

(RVII)

Maximum power transfer

|Zapp | = |ZThev |
Point of collapse
E

ZThev
Thevenin

Zapp
Load

Major improvements in accuracy, numerical

stability, implementation variants, and ease of use
* K. Vu and D. Novosel, Voltage Instability Predictor (VIP) - Method
and System for Performing Adaptive Control to Improve Voltage
Stability in Power Systems, US Patent No. 6,219,591, April 2001.
PG&E and Quanta Technology LLC Proprietary and Confidential

Page 7

New RVII Advantages 1 (2)

Model-free, fast real-time voltage instability detection method*, independent of state

estimation

Implementation in several variants: bus, load center, transmission line, transmission

corridor
Calculates Q-margin & other indices
for proximity to voltage collapse
Stability boundary calculated with
real-time PMU data refresh rate

Easily combined with other

methods and indices
Reactive power monitoring
Could initiate model-based
contingency analysis,
e.g. by alarming the operator
New Q-margin
PG&E and Quanta Technology LLC Proprietary and Confidential

Loading margin
Page 8

New RVII Advantages 2 (2)

Ability to process data from different sources (PMUs, SCADA, simulation outputs (static
and dynamic))
Takes immediate advantage of available PMUs

Scales up well with increased number of PMUs

Excellent results from actual system tests on slowly changing system operating
conditions (load ramp) and in tracking system dynamics after large disturbances

Able to distinguish FIDVR from voltage

instability even if voltage is very low

Simple implementation in Control Center

tools local and/or IEDs for:
Operator tools to increase
situational awareness
Local automated actions
Addition to SIPS

Load center
PG&E and Quanta Technology LLC Proprietary and Confidential

Page 9

Real-Life System Test Results

Alarm when measuring

real-time Q margins

Realistic PQ-curve slope before

and after the disturbance

PG&E and Quanta Technology LLC Proprietary and Confidential

Page 10

Real-Time Simulation Results: PG&E

Line switching Two T-lines (Sequentially)
Lines switched are from outside of the single line

The P-Q plane - Use of 5 PMUs for A-B Corridor

@B
Reactive Margin Computation and Observability
- Time evolutions of margins for different PMUs
available

PG&E and Quanta Technology LLC Proprietary and Confidential

Page 11

Test Results
Comprehensive tests using real-life PMU and SCADA measurements and off-line
time-sequence simulation tools

Ability to detect instability even if voltage close to nominal

Accurate results for load centers, transmission lines, and corridors

Detection at highly-meshed high voltage systems (e.g. 500 kV) is more difficult

Results comparable to detailed, model-based off-line QV analysis; very

accurate closer to instability boundary

Discriminates between FIDVR and fast voltage instability

FIDVR cases (no voltage collapse) are accurately detected despite the fact the
voltage is low for some time

No false alarms
PG&E and Quanta Technology LLC Proprietary and Confidential

Page 12

Present and Future Activities

Deploy real-time application software based on RVII as part of

comprehensive WAMPAC system in control room

Design monitoring and control strategies based New

on RVII
Traditional

MODEL-BASED

MEASUREMENT-BASED
VOLTAGE STABILITY

VOLTAGE STABILITY

Identification

Analysis

RPM Methodology

(EMS)

Other EMS
Applications

SCADA
& Alarms

WAMS

State
Estimator

State
Measurement

Small Signal
Stability

Oscillation
Monitoring

Transient
& Voltage
Stability

Stability
Monitoring
& Control
Island Detection,

Island
Management

Resynchronization,

Other Indices

New Real-time
Voltage Instability
Indicator Solution

Alstom
e-terravision
approach
with new
RVII

New
Applications

& Blackstart

PG&E and Quanta Technology LLC Proprietary and Confidential

Page 13