Post on 11-Jan-2016
description
Stability and Security of Power
NetworksG. T. Heydt
Arizona State University
ECEDHA 2004 Annual Meeting
March, 2004
Orlando, Florida
Outline
Stability and security: a general discussion
Weaknesses and strengths of the North American grid
Some theoretical considerations
Solutions: short range and long range
Propaganda: power engineering education
Conclusions
Stability
Power system stability basically refers to the
ability of operating an AC network with all
generators in synchronism, retaining
synchronism even after a large disturbance
Stability
Each synchronous generator has a ‘Newton’s
law’ second order nonlinear differential equation
that describes the machine angle – and control
systems (e.g., power system stabilizers) also
contribute a higher order nonlinear controller to
the dynamics
A large interconnection (WECC, e.g.) may have
about 200 generators + 150 PSSs = about 1000 to
10000 order nonlinear differential equations
s
tf
x
VETP
)sin(||||)(
Stability
The basic analysis technique is state space analysis /
eigenvalues for the linearized system, or simulation for the
nonlinear system. Typically, the dimension is very high –
in the 1000 – 10,000 range. The interconnection is
modeled as Ibus = Ybus Vbus which is reduced to
eliminate the non-dynamic nodes (i.e., remove the non-
generation nodes).
Power system stabilizers
A PSS is a controller that uses (usually local)
measurements to provide a signal to one
generator so that damping torque is produced by
the machine field winding. The basic concept is
that a linear controller is used with standard
feedback control technology to place the poles of
the linearized system solidly in the LHP. Virtually
all large generating units in North America are
fitted with PSSs.
Power system stabilizers
The main weaknesses of this approach are that
the nonlinear system may respond poorly, and
also dynamics external to the generator + PSS
are not modeled (nor included in the
measurements). Therefore modes that result
from inter area dynamics may not be damped.
xx
xx
x xx
xx
By injecting the appropriate signals from distant measurements in the system, transmitted through LEOS, the controller is able to obtain superior performance in terms of damping interarea oscillations compared to use of conventional local signals. The main concept is to use interarea signals for interarea controls
SPSSLOCAL
MEASUREMENTS
REGIONAL MEASUREMENTSREGIONAL MEASUREMENTS
Wide area robust power system stability control
Low Earth Orbit Low Earth Orbit Satellites LEOSSatellites LEOS
Hierarchical robust power system controller
Execution Level
Signal pre-processor
Actuator / Distributor
Operation Level
System modal identifier
SPSS damping loop
Management Level
Fuzzy logic based parameter tuner
Management Level
Operation Level
Execution Level
Power SystemPower System
Input DataInput Data ControlControl
Voltage Regulator With PSS and SPSS
GeneratorExcitation
System
+
-
Ref
Generator Field
Gen
Vt
+
PSS ,f, or Pa
VoltageRegulator
Remote SignalsSPSS
G1 G3
G2 G4
0.011+j0.11
Load 1 Load 2
Area 1 Area 2
Time (sec.)
Am
plitu
de
SYS impulse response--1st input to outputs
-1
0
1From: U(1)
To: Y
(1)
0 14 28 42 56 70-1
0
1
To: Y
(2)
Frequency (rad/sec)
Phase (
deg);
Magnitude (
dB
)
SYS bode graph--1st input to 1st-2nd outputs
-200
0
200From: U(1)
-400
-200
0
To: Y
(1)
-200
0
200
10-2 100 102-1000
-500
0
To: Y
(2)
Time (sec.)
Am
plitu
de
SYS+LMI1 impulse response--1st input to outputs
-0.2
0
0.2From: U(1)
To: Y
(1)
0 5 10 15 20 25 30-0.2
0
0.2
To: Y
(2)
Frequency (rad/sec)
Phase (
deg);
Magnitude (
dB
)SYS+LMI
1 open-loop transfer function Bode graph
-400
-200
0From: U(1)
-500
0
500To: Y
(1)
-400
-200
0
100 105-1000
-500
0
To: Y
(2)
Key issues
• Full scale nonlinear solution (transient stability study)
• Eigenvalues of the linearized system near the operating
point (small signal stability)
• Line and component ratings
• Voltage ratings (maximum and minimum)
• Coherency - groups of generators swinging together
• Synchronizing torque, PSSs
• Acceptable operating conditions (including operation
within about 50 mHz of 60 Hz)
Intentional Intentional human actshuman acts
NetworkNetwork MarketMarket
Information & Information & decisionsdecisions
CommunicationCommunicationsystemssystems
Natural calamitiesNatural calamities
InternalInternalSourcesSources
ExternalExternalSourcesSources
Security refers to the ability of the system to
respond only to intended operator
commands, blocking all unintended
operations
Security
Electric power system is vulnerable to failure due to
Natural disasters
Deliberate attack
Equipment failures
Operator error
Accidents
Tree-related events
High load periods
Software failures
PMU
Sensor Systems
Transformers
Substations
Monitoring of electric power networks
Advanced
Underground
Transmission Lines
Overhead Transmission
Lines
EMS
Energy management systems
Archiving
E M S
Operator interaction
State estimator
Generator controls
Sensory information
Command and control
Virtual Sensor Present
Virtual Data
EMS
Network vulnerability reduction through virtual sensor utilization
EMS
Network Data Lost
No Data!
EMS
Tradeoffs betweenvirtual and physical sensors
$$$$
$$$ $
$ $
Low Cost
Less Accurate
Physical Sensors
Virtual Sensors
High Cost
Greater Accuracy
V IZ = [H] X
What is needed to enhance both security and stability
•Ability to acquire and interpret extensive real-time information from diverse sources, ranging from sensors to satellites. Sensory data used in Hx = z state estimators to enhance system performance.•Ability to quickly evaluate system vulnerability with respect to catastrophic events in a market environment involving competing, self-serving agents•Ability to adapt protective device performance based on system-wide and external system assessment•Ability to reconfigure the power network to minimize system vulnerability•Ability to develop system restoration plans to minimize the impact of disruption
Failure
Analysis
Information
&
Sensing Vulnerability
Assessment
Self Healing
Strategies
Strategy
Deployment
GP
S
Sate
llit
e
LE
O
Sate
llit
eIn
tran
etIn
tern
et
Strategic Power Infrastructure Defense System
Communication system for strategic power infrastructure defense
Satellite dish
Protective device
Gateway
Strategic power infrastructure main system
Time synchronization (GPS) / Self healing / Info. Exchange (LEO)
IntranetEthernet or model based network is used in the Intranet. Each Intranet can have a “gateway” that handles IP addresses in the Intranet
Internet or any other communication channel for a number of Intranets
GPS or LEO satellite communication
Internet based communication channelInternet based or more direct and faster communication channel
The North American grid
NERC: policies, rules, reliability, plans, synchronous interconnections
North American Electric Reliability Council
• Sets standards for the reliable operation and planning
• Monitors, assesses and enforces compliance with standards
• Provides education and training
• Assesses, analyzes and reports on bulk electric system adequacy
• Coordinates with Regional Reliability Councils
• Coordinates the provision of applications, data and services
• Certifies reliability service organizations and personnel
• Coordinates critical infrastructure protection
• Enables the reliable operation by facilitating information
exchange and coordination among reliability service
organizations
• Administers procedures for appeals and conflict resolution
Weaknesses and strengths of the North American grid
• Basic transmission design is over 40 years old. Some
basic distribution circuits are over 60 years old.
• Never designed to handle high levels of bulk power
• Both transmission and generation constrained
• The impact of market driven exchange of power has
stressed the transmission grid
• The transition to market based infrastructure has
stressed the newly created control entities (e.g., ISOs)
– in an industry that is rapidly loosing corporate
memory
The Northeast blackout of 2003
Time 8/14/2003 4:09:57 PM EDT: The first significant events were initially recorded in Michigan and Ohio
The Northeast blackout of 2003
Time: 8/14/03 04:10:39 PM EDT: The disturbance was then recorded all over Michigan , Ohio , and the city of Buffalo, NY
The Northeast blackout of 2003
Time: 8/14/03 04:10:58 PM EDT: 19 seconds later, the disturbance had propagated to the eastern seaboard.
The Northeast blackout of 2003
Main causesFailure of state estimator in MISO to model ‘external’
system changes
Combination of heavy power exchanges, high reactive power flows, planned outages of transmission circuits and planned outage of a main generating facility (none of which are unusual)
Operator error / training of MISO operators / imprudent operation of an Ohio utility (generation outages)
Unplanned unit and line outages
The Northeast blackout of 2003
The Northeast blackout of 2003
Generation building boom of the past
0
20
40
60
80
100
120
140
160
180
200
1950 1955 1960 1965 1970 1975 1980 1985 1990 1995 2000
Coal Oil Gas Nuclear Other
GW
Ins
talle
d
in F
ive
Ye
ar P
erio
d
A hindsight view of the past building boom
0
20
40
60
80
100
120
140
160
180
200
1950 1955 1960 1965 1970 1975 1980 1985 1990 1995 2000
Coal Oil Gas Nuclear Other
GW
Inst
alle
d i
n F
ive
Yea
r P
erio
d
Generation Building Boom Follows the Baby Boom Labor Force Entry
17.93
29.41
11.93
19.23
11.69 Per
cen
t C
han
ge
in L
abo
r F
orc
e
0
5
10
15
20
25
30
35
Generation building boom of the future
0
200
400
600
800
1000
1200
1400
2000 2005 2010 2015 2020 2025 2030
By 2020, the U.S. will need 1300 new power plants at 300 MW each
By 2020, the U.S. will need 1300 new power plants at 300 MW each
Total System Generation Capacity
Cumulative Additions
GW
Employment at major IOU’s
TRANSMISSION
DISTRIBUTION
The N9s problem
Electric power quality Extreme bus voltage reliability, for example 'five
nines' (i.e., 0.99999 availability), or six nines or even higher
Utilization of new transmission and distribution technologies for improvement of reliability
Utilization of distributed energy sources (DERs) to improve reliability
Working with manufacturers of information technology equipment to reduce load vulnerability
24/7 UTILIZATION OF POWER SYSTEM ULTRA HIGH RELIABILITY
INFORMATION PROCESSING, FINANCIAL SERVICES, AIRLINES, POLICE, MILITARY
Reliability enhancement
Distributed rather than concentrated loads
Loop circuits for distribution systems
Information Technology and sensitive manufacturing loads
Independence of energy sourcesEnvironmental issues
AS A RESPONSE TO THE 1993 TERRORIST BOMBING OF THE WTC,
THE PRIMARY DISTRIBUTION SYSTEM IN THE BUILDING WAS
IMPROVED TO KEEP THE POWER ON IN THE CASE OF SEVERE
DISRUPTION OF THE SUPPLY / INTERRUPTION OF THE IN-BUILDING
PRIMARY DISTRIBUTION. THERE WERE TEN SUBSTATIONS IN EACH
WTC TOWER, ON FLOORS 7, 41, 75, AND 108, AND THE SOUTH TOWER
HAD AN ADDITIONAL TENANT OWNED DOUBLY FED SUBSTATION ON
FLOOR 43
THE USE OF MULTIPLE FEEDS, MULTIPLE SUBSTATIONS, AND ISOLATED POWER SUPPLIES KEPT THE POWER ON IN MOST OF THE WTC FOR 102 MINUTES AFTER THE INITIAL STRIKE. IT IS BELIEVED THAT THIS WAS THE MAIN FACTOR IN SAVING THE LIVES OF AS MANY AS 18,000 PEOPLE WHO ESCAPED FROM THE TOWERS BEFORE COLLAPSE
Independence of sources
The dependence of the sources will result in a much higher
outage rate than (1-P1)(1-P2)
TWO FEEDERS RELIABLE LOAD BUS
LOAD
1-P = (1-P1)(1-P2)
Modeling dependence of sources
The dependence effect of multiple sources can be
modeled using a difference equation of the form
qn+1 = Cqn+(1-C)(q1)1/n qn
where qn = 1-pn = outage rate of circuit upon addition
of nth feeder, C is a correlation coefficient
The (q1)1/n term is called a discounting term and it
accounts for increased potential for dependence for
cases of large n (large numbers of feeders)
Discounted model
C = 0 indicates no correlation between multiple
feeders
C = 1 indicates the feeder outages among several
feeders are ‘common mode’
Reliability of multiple feeds
100 % circuitcorrelation
1% circuitcorrelation
Zero circuitcorrelation
Rel
iabi
lity
expr
esse
d as
num
ber
of 9
s
Number of circuitfeeders
0 1 2 3 4 50
2
6
8
4
10
The addition of
feeders to improve
reliability has a
diminishing effect.
For practical cases,
use of more than
three ‘independent’
feeders of 100%
capacity is counter-
productive.
0.9 0.99 0.999 0.9999 0.99999 0.999999 0.9999999 0.99999999 0.9999999991 2 3 4 5 6 7 8 9
Onegenerator,FOR = 1%
Two generators,FOR = 1%
One generator, + 1feeder FOR = 1%
Two feeders FOR =1%, Dependence10%
1 day in 20years
1 day in 200 years
3 feeders FOR = 1%,Dependence 10%
Threegenerators, FOR= 1%
Probabilities of uncommon events
COMMON (?)
Event_______
Loosing at roulette
in Las Vegas – bet on 00
Loosing the
PowerBall
lottery
FAA design
criteria for
aircraft
POWER SYSTEM
RELIABILITY
Reliability N Outage time
99.9 3 8h 45 min / yr
99.998631 4.9 1 day / 200 yrs
99.999 5 5 min 15 s / yr
99.99999 7 3.2 s / yr
99.999999 8 18.9 cycles / yr
99.9999999 9 1.8 cycles / yr
LIFE
Probability, N
97.368, 1.6
99.99995, 6.3
0.999999999
0.999999999999,
9 to 12
Solutions: short range
Distributed generation
Added small generation units at all levels
Conservation / electronic control of loads
Investment in distribution systems
Sharp increase in research in both transmission
and distribution engineering
Recruiting of students to the power area at all
levels
Improvement of software tools
PROTON EXCHANGE
MEMBRANE FUEL CELL - 7.5 kVA
PHOSPHORIC ACID
250 kVA FUEL CELL
Low capacity, high speed units with electronic interface with 60 Hz bus
Alternative fuel sources (e.g., biogas, gasifier, pyrolysis, fuels that have less than 10% of heat content compared to fossil fuels)
Catalytic combustor to reduce nitrous oxide production
Heat recovery
Lower capacities -- e.g.,
5 - 300 kVA
High efficiency small units
New IEEE standard requires disconnection from the distribution system within a few cycles during low voltage or outage events
Microturbines
Solutions: long term
Added generation in larger units
Local solutions for high reliability requirements
Added capacity in distribution systems
Adaptive islanding of interconnected systems
Coordinate national energy policy with system
realities
The educational aspect of the
problem
U. S. Power engineering undergraduate enrollments
1960 1980 20000
500
1000
1500
2000
Source: G.T. Heydt and V. Vittal, “Feeding Our Profession,” IEEE Power & Energy Magazine, vol.1, issue 1, Jan/Feb 2003, pp 38-45
un
der
gra
du
ate
d
egre
e r
ecip
ien
ts
U. S. Power engineering graduate enrollments
1970 1975 1980 1985 1990 1995 20000
50
100
150
200
year
M.S.E.E.
Ph.D.
Source: G.T. Heydt and V. Vittal, “Feeding Our Profession,” IEEE Power & Energy Magazine, vol.1, issue 1, Jan/Feb 2003, pp 38-45
gra
du
ate
d
egre
e r
ecip
ien
ts
The general electrical engineering reality
There is a certain ebb and flow to the enrolments
in engineering nation-wide; since the all-time low
in undergraduate engineering in 1998, there has
been an uninterrupted growth in enrolments
In many electrical and computer engineering
programs, the growing tendency to select the
computer engineering option has resulted in the
majority of students seeing little or no subject
matter relating to energy and power
The general electrical engineering reality
Given the decreasing number of electrical
engineering undergraduates, there is good
progress in stopping the precipitous decline in
the undergraduate power engineering enrolments
to the point where many power programs are
experiencing record levels
Encouraging developments on the curriculum front
A determined movement away from the old
straight jacket curriculum to new enriched course
offerings with broadened choice New developments are evident in three principal
thrusts addition of microeconomic/finance elements introduction of energy, environment and
public policy aspects wider array of power systems, power
electronics and machines/drives courses
The impact of recent events
Restructuring of electricity and the California
crisis sharpened public interest in electricity
The September 11, 2001 tragedy brought to
prominence the issue of the security of the North
American interconnected power system
The 2003 mega-blackout produced keen interest
in the reliability of the interconnected grid
Conclusions
Stability of power systems is a well understood phenomenon, but complex numerical problem.
Stability enhancement controls are very complex to design, but the present research thrusts and engineering practice have yielded in-service designs (or designs nearly in-service) that are suitable to the task
The transition to a market based energy infrastructure may not have been well thought out, and system implications are just now being remedied
Conclusions
Distribution engineering, long a step-child of power engineering, is a focus of research – mainly related to high reliability, cost reduction, and distributed generation sources
System security is a point of focus in contemporary power engineering
Research on sensory systems is needed to enhance system security
Power engineering education and the production of power engineers at all levels seems to have a significant impact on the health of the national power system. It is unclear that the number of engineers needed will be attained by US educational institutions.