1

Copyright © SEL 2012

Modern Solutions for the

Control and Protection of Electric

Power Systems

Edmund O. Schweitzer, III Schweitzer Engineering Laboratories, Inc.

June 8, 2012

They provide us the most versatile form of energy,

at the flick of a switch... at the speed of light...

at a fair price.

Our Electric Power Systems

Are Modern Marvels

Reliable

Economical

Efficient

Safe

2

The Future of Power Systems

Solutions must be: efficient, low-cost, robust, reliable

No Blackouts

Faster Tripping

Islanding

State Measurement

New Sources

Distributed Generation

Intermittent Sources

Innovations

Better Control

Fast Remedial Action

Wide-Area Control

Secure

We Must Operate, Control, Protect Our

Wide-Area Systems Better

Time-coordination is for BACKUP.

Quicker loop times

Cycles or seconds…vs…minutes

Automation / operators

State-based control

Measure the state

Drive the system to desired state

Distributed control/successful islanding

3

First Principles

Local decision making is most dependable.

Security is critical.

Logic-based: “If I lose these lines when

load is big, then I must shed load.”

Thermal-based: “This device will fail if

overload continues, so load must be

reduced or device must be tripped.” “I

should tell others when I am overloaded.”

Voltage and Angle Stability

Power-voltage characteristics are

becoming more “brittle.”

Low voltage can be worse than NO voltage.

=> Time-undervoltage characteristics.

Angle stability includes swings, out-of-step,

frequency.

Angles can indicate voltage stability

problems too.

4

Load Trends Over Time

Fewer resistive loads (P = V2/R)

More switchers (P = const.)

More “brittle” systems

Increasing risk of voltage collapse

Conservation voltage reduction is less

effective today…and may even be counter-

productive!

50” Flat-Panel TV Test

I

V

1.5

2.5

P

Old TV (est.)

2.0

250

200

150

60 80 100 120

P

I

5

What Is the “State”

of the Power System?

A vector of the complex voltages at every

node in the system, known at the same

instant of time.

You can estimate it from traditional SCADA

data, using a “state estimator,” …

…or you can directly measure it:

SYNCHROPHASORS.

Directly Measure the State

Detect bad data

Average

measurements

Determine topology

Calculate V at

adjacent stations:

V’ = V + ZI

Network

Sub 1

RTAC . . .

Sub n

SVP

Synchronized

Data

V1

V2

. . .

Vn

V1

Vn

6

Sharing the Complete State Vector

Every Location Gets the State Vector

N = j + k + l + m + n

V1

.

.

.

Vj

. . . j k l m n

Vj

Vk

Vl

Vm

Vn

= S

Lateral Communications:

ICON, Radio

Vj = Vk Vl Vm Vn

Time Sub 1 Sub 2 Sub 3 Sub 4

0 V0 V0 V0 V0

1 V1 V1 V1 V1

2 V2 V2 V2 V2

3 V3 V3 V3 V3

4 V4 V4 V4 V4

5 V5 V5 V5 V5

6 V6 V6 V6 V6

7 V7 V7 V7 V7

8 V8 V8 V8 V8

9 V9 V9 V9 V9

… … … … …

Newest state is 1

Two-Second Scan, One-Second Synchrophasors

= State Every Second, Two Seconds Latency

Newest state is 3

Newest state is 5

Direct state measurement, second by second, with one scan of latency.

7

Generator Relays Will Directly Measure

a

'a

N

S

E

IZ

V

E

Z

V, I

Generator

+

-

Rotor Test System

Load

Box

Synchronous

Generator

Prime

Mover

Optical

PickupVs and Is

8

Western

Interconnection

>40,000 units

Texas

Interconnection

>18,000 units

Eastern

Interconnection

>125,000 units

SEL Synchrophasors Are Everywhere,

and Growing Every Day!

Reliable, Redundant Precise Time

Satellites: GPS is great when available

ICON networks +/– 0.5 µs of all nodes with

or without GPS

Rubidium and cesium standards

WWVB 60 kHz time broadcast

You can build systems today that enjoy the

precision of GPS and provide wide-area

synchrony even when GPS is lost.

9

GPS Time Is Not Guaranteed!

Jamming or interference

(NAVWAR)

Equipment failure

DoD control

Solar flares

“On December 6, 2006, a solar flare created an unprecedented intense solar radio burst

causing large numbers of receivers to stop tracking the GPS signal.”

-- NOAA Press Release

Close the Loop for Real-Time Control

Locally and over wide areas

Relay

SVP or RTAC Relay

Relay

10

RTAC & SVP:

Relay-Speed Processing, Anywhere

State Equations: Stability, Thermal

Phasor Math: Self-Checks, Interpolation

Synchrophasor Processing

Synchrophasor Vector Processor

One-cycle loops

Local data qualification

Regional decision making

Relay speeds, special-protection schemes

Real-Time Automation Controller

Sub-second loops

Ideal SCADA interface

Directly in the Relays

11

What Do I Do With the State?

Measure it

State equations predict where state is

heading.

If it’s going somewhere bad, then act, to

keep it safe.

Repeat.

Making It Easier

Break down the system, model locally

Generation Plants

Transmission Networks

Distribution Systems

Build equivalents, for simplicity

The farther away, the simpler

12

Communications for Critical Infrastructure

1 2

3

1

2

3

7 ms

12 ms

7 ms

Secure, 5 ms healing, reliable, redundant, 0.5 µs absolute time

Relay – SVP – ICON – SVP – Relay Three to four cycles, across wide areas

t E t E t E

SVP

PMU

13

Copyright © SEL 2012

Georgia: Tiblisi Relies on Enguri Dam Line Loss Requires Load Shedding

Enguri HPP

115 kV Network

400 kV

900 MW

AngosturaTapachula

City

400 kV

Southern Region Load

SabinoChicoasen

National

System

400 kV

Synchrophasors Maintain Stability

at CFE

Synchrophasors Trip

Generator

14

Florida Power & Light Anti-Islanding

T

DG

Relay

Relay

Angle

Frequency

Slip

Acceleration

SVP

State Interpolation or “Linear Estimation”

V1 = ¼ VL1 + VL2 + VL3 + VL4 if all breakers closed and

if all voltages about the same

V2 = V1 + Z12 I1

Line 4 Line 1

Line 2 Line 3

Sub 1

Sub 2

V2

VL2

VL1 VL4

VL3

Z12

I1

15

Determining Wide-Area Voltages

Use Local V and I to Determine Remote V

1 2 3

4 5 6

7 8 9

V2’ = V1 + Z12 • I12

Utilities Are Operating Closer to the Edge

MarginMargin

1.0 1.3 1.5 1.7PU Nominal Load

1.0

0.5

0.0

Operating

Point

Bifurcation

Point

V

16

Monitor Angles Between Transmission

System and Critical Distribution Buses

SEL-421

SEL PDC

Modem

Modem

Server

SEL-421

Detect Voltage Collapse by Angle

138 kV

138 kV

69 kV

23 kV

17

Inverse-Time Undervoltage Elements

Shed Low-Voltage Loads First No Communications Needed!

0.2 0.4 0.6 0.8

Voltage (pu)

Seconds

0

5

10

15

Governor Mode Control

Utility Power

System

Relay

Plant Distribution

System

Relay

Generator Governor

Abbott

Generator

Frequency and Angle

Measurements Across

Relay-to-Relay Logic

Communications

Regulate

Power

Regulate

Frequency

18

Governors Today and Tomorrow

Power setpoint

Most sources

Matches frequency of “system”

Frequency setpoint

One plant per system

Others follow by “droop”

Phase setpoint??

Why Drive Phase Instead of Frequency?

δ = ω * t ω = δ’

P = [ V1*V2 / X ] sin δ

Historically, operators adjusted frequency to

get 24 x 60 x 60 x 60 cycles per day

Going forward, command the ANGLES to drive

the desired power flows…and keep up instant

by instant.

Angle Is a State Variable

19

Traditional Control to Isolate Line

Bus 1 Bus 2 224 kV 224 kV

Master Control

Traditional Control to Isolate Line

Bus 1 Bus 2

Master Control

230 kV 206 kV

20

Traditional Control to Isolate Line

Bus 1 Bus 2

Master Control

230 kV 206 kV

Traditional Control to Isolate Line

Bus 1 Bus 2 224 kV

Master Control

Close Cap

230 kV

21

Traditional Control to Isolate Line

Bus 1 Bus 2 224 kV 224 kV

Master Control

Tap Down Close Cap

Synchronous Control to Isolate Line

Bus 1 Bus 2 224 kV 224 kV

Master Control

Local Control Local Control

22

Bus 1 Bus 2 224 kV 224 kV

Master Control

Local Control Local Control

Recipes Sent to Local Controllers

Bus 1 Bus 2

Execute

Recipe

Simultaneously

224 kV 224 kV

Master Control

Local Control Local Control

Recipes Execute at Precise Instant

Tap Down Close Cap

23

Performance of Future Systems

Decentralized processing for robustness,

speed, security, successful islanding.

Clearing faults with no intentional delay.

Nearly instantaneous restoration.

Add distributed sources, anywhere.

Prevent blackouts: measure state, predict

trajectory, decide before “crashing.”

Lower capital and operating costs.