The DoE / PSerc Initiative
Thrust Area 1
Electric Energy Challenges of the
Future
PSerc IAB Meeting
Golden, CO
December, 2012
Reported by
Kory Hedman, G. T. Heydt
Reporting on the research of D. Aliprantis, K. Hedman,
G. Heydt, J. McCalley, S. Venkatasubramanian
1
1.1 Integrating Transmission
and Distribution Engineering
Eventualities
1.2 A National
Transmission Overlay
The work of Dr. Heydt and
researchers at ASU
The work of Drs. McCalley
and Aliprantis and
researchers at ISU
1.3 Robust and Dynamic
Reserve Requirements
The work of Dr. Hedman
and researchers at ASU
1.4 Wide Area Control
Systems
The work of Dr.
Venkatasubramanian and
researchers at WSU
2
1.1 Integrating Transmission
and Distribution Engineering
Eventualities
PI: Prof. G. T. Heydt
Accomplishments
•Analysis of HVDC multiterminal systems and application in the
WECC system (making the Pacific DC Intertie a three terminal
system)
•Analysis of HVDC meshed networks and application in Southern
California – Mexico – Arizona (alleviating stress in the adjacent 500 kV
system, and effectuating a tie with CFE-Mexico)
•Analysis of the integration of HTLS and compact phase overhead
transmission in the contemporary AC system – with examples
•Analysis of six phase overhead transmission systems, and the
integration of those circuits into the existing three phase AC network
•What are the advantages and disadvantages of these technologies,
and cost effectiveness
3
Example: reconductoring Rinaldi-Tarzana 230 kV with HTLS
• A short line, ~18
miles
• Typical HTLS
reconductoring
• Long term thermal
limit improved
• Security limit not
impacted
• Voltage limit not
impacted
• Limitations imposed
by the loading of
adjacent circuits and
components at
Tarzana
Relieves a WECC critical path.
Losses increase under heavy loading.
4
Example: A HVDC meshed network in Southern California –
Arizona - Mexico
500 kV AC
IMPERIAL
VALLEY
AC/DC
Meshed HVDC illustration in the WECC system
* Maximum for the indicated conductor
NORTHconverter
GILA
Station
DC
SAN LUIS RIO
COLORADO
PUERTO PEÑASCO
mesh
WECC
CFE
To CFE
A comparison of simulation results for two
illustrative implementations of HVDC
transmission
North Gila, AZ
Imperial
Valley, CA
San Luís RC,
Sonora
Puerto Peñaso,
Sonora
*To retain all lines in thermal and security limits
Case considered
WECC system base
case with no added
HVDC
Same with added
meshed HVDC
Location of
increased
loading
Imperial Valley
and San Luis
Rio Colorado
Imperial Valley
and San Luis
Rio Colorado
Permissible
increase in
base case
loading MW *
992
3350
Length (km)
Conductor
Configuration
Conductor/ pole
Voltage rating
Power rating*
Converter stations
Interconnecti Technology
Voltage
on Region
Rating
WECC
WECC
WECC
CFE
VSC
±500 kV
Control
type
Current
Voltage
Current
Current
HVDC transmission
North Gila – Imperial
Imperial Valley – San Luis RC
Valley
114
96
San Luis RC – Puerto
Puerto Peñasco – North Gila
Peñasco
170
181
Falcon (1590 kcmil)
2 lines
2
± 500 kV
2.63 GW
Utilization of a current controlled HVDC system at the US-Mexico border to:
parallel 500 kV critical circuits; asynchronously connect CFE-Mexico to the
US and also CFE-Sonora to CFE-Baja California. Enhances security of the
southwest AC grid.
5
Innovative concepts for bulk power transmission
Objectives
• Maximize conductor and insulating systems in overhead and
underground transmission
• Maximize space utilization
• Maximize use of investment in transmission assets /
designing a favorable cost-benefit ratio
• Permit solutions to transmission problems that have few
viable alternatives
• Alleviate critical transmission paths
• Design new asynchronous connections for specialized
applications
6
Innovative concepts for bulk power transmission
High phase order cables
•Lower line-line voltage
•High power transmission
•Low electromagnetic impact
•In higher phase order systems,
some of these advantages are magnified
•Higher reliability (single pole out)
•Potentially narrower ROW, trenches and conduits
In a six phase system,
line-neutral and line-line
voltage magnitudes are
the same. Cable
systems can be
designed with less
dielectric stress
Challenges: protection, innovative interfaces to the three phase
system, loss of phase operation, innovative cable designs
Nonsinusoidal waveform applications
•Utilize dielectric / insulation systems maximally (not like a sine wave)
•Utilizes the conductor ampacity more fully
•Can be generated electronically – we already have the electronic interface in
many cases
Square wave Pav = |Vrms||Irms|ρ
•Low frequency outlet feeders from V
Sine wave Pav = ½|Vrms||Irms|ρ
wind farms and PV generation
Challenges: mitigate power electronic losses
and cost, lower RFI
Time
7
Innovative concepts for bulk power transmission
High phase order, compact phase designs for overhead and cable
transmission
Low phase to phase voltages which can be
combined with phase compaction to give
high power capacity for a given spatial
occupancy (cables or overhead)
Vln
Vll
Vll
360o/n
2
1.8
•Potential for operation under loss of a phase
•Low electromagnetic impact
•Maximal use of space
•“Positive sequence” impedance of the cable can be
simply calculated using Toeplitz matrices. AC
calculations simplified analogous to symmetrical
component calculations.
Vll if Vln is held at 1.0 V
1.6
1.4
1.2
1
0.8
0.6
0.4
0.2
0
0
10
20
30
Number of phases
40
50
60
Challenges: interface with the three phase system
Integrated – innovative technologies for bulk power applications
•Combine several innovative technologies to obtain enhanced transmission
characteristics
•AC and DC subsystems to obtain benefits of both
•Realize asynchronous interconnections at high power levels
•Permit islanding and connection of North American interconnections
8
1.4 Wide Area Control
Systems
PI: Prof. Mani
Venkatasubramanian
Task Objectives
• Wide-area control designs for future power systems
• High percentage of renewable energy sources
• Fast wide-area monitoring systems with PMUs everywhere
• What control designs are feasible?
• Voltage stability issues at substations from reactive power
demands of renewables and power electronic devices
• Small-signal stability problems from interactions of power
electronic devices with the grid
• Transient stability concerns from unforeseen operating
conditions and high order contingencies
9
Wide-area PMU
measurements
SCADA
data
State
Estimation
model
Wide-Area Voltage
Controller
• Hierarchical design
developed
• Substation Local
Voltage Controllers
• Central Coordinator
Operator
inputs
Supervisory
Central
Voltage
Coordinator
(SCVC)
Supervisory
Control
Signals
Local
PMUs
Supervisory
Control
Signals
Selected
SCADA data
Substation #1
Local Voltage
Controller
Switching
commands
. . .
SLVC #1
Substation #1
Controls
Selected
SCADA data
Substation #N
Local Voltage
Controller
SLVC #N
Local
PMUs
Switching
commands
Substation #N
Controls
Presented in IEEE
2012 PES General
Meeting paper
10
Substation Voltage Controller
• Control of substation VAR demands crucial in future
power systems
• Substation Local Voltage Controller developed
• Tested on a complex substation with heavy loads and
many power electronic converters
Local PMU
measurements
Local
Switching
commands
Selected
SCADA
data
ULTC
transformers
Capacitor banks
Reactor banks
Supervisory
control
signals
Substation
Local
Voltage
Controller
Other VAR
devices
Alerts
Local Operator
Alerts
Alarms to Grid
Control Center
11
Small-Signal Stability
• Challenging problem from lack of models and from
interaction of discrete and continuous dynamics
• Subsynchronous high frequency modes (5 Hz to 50Hz)
from power electronic devices may be problematic
• High sampling frequency data may be needed
• Controls on power electronic side may be culprits
• Detection and isolation crucial
• Which substation(s)? Which controls?
• New algorithms for monitoring and analysis of
subsynchronous modes developed
• Tested on 12.5 Hz SSR Mode seen in certain wind farms
in Oklahoma Gas and Electric from DFR data at 5760 Hz
sampling rate (NASPI February 2012 presentation)
12
Wide-Area Transient Stability Controller
• New model prediction based real-time angle stability
control developed
• Suited for uncertain operating conditions of the future and
for unplanned high order contingencies
• PMUs assumed everywhere
• Rotor machine angles assumed to be measured from
synchronous machines
• Fast communication and computation assumed
• Stable or unstable? What control actions? Algorithms
developed.
• Can stabilize highly complex contingencies
• Doctoral dissertation work of Greg Zweigle
• IEEE 2012 PES Transmission Expo paper
13
Future work
• This project focused on control designs for future power
systems
• How to get to the future?
• Transition road map needed
• Control designs with combination of SCADA and PMUs
• Control designs with increasing availability of fast
communication networks
• Road from here to the future needs to be planned out
• Implementations in phases
• Start working with utilities on actual designs and
implementations
14
1.2 A National
Transmission Overlay
PIs: Profs. J. McCalley
and D. Aliprantis
Objectives and
Accomplishments
Objectives:
• Develop explicit overlay design process (purely engineering study) to facilitate
growth of renewable (wind, solar, geothermal) to 2050;
• Design and evaluate several “good” US national transmission overlays for different
futures;
• Quantify benefits to building a national transmission overlay;
• Assessment of new HVAC-HVDC transmission overlay scenario of high renewable
energy penetration for multi-area speed dynamics;
• Account for the control capabilities of transmission technologies for overlay design.
Accomplishments:
• A well-formatted “co-optimization” planning framework;
• Transmission overlay designs for 4 scenarios respectively;
• Innovated transmission candidate selection process;
• Quantified benefits of a national transmission system;
• Innovated multi-state/circuit transmission optimization modeling method (to be
finished soon).
• Reachability analysis to incorporate dynamics in the designs.
15
Study Framework
GIS/climate data
Transportation
System
Estimated
Econ signal
Ventyx
Existing Tran. &
Gen. data
DOE AEO
NERC LTRA,
ES&D
2. Iterative re-weighting minimum
spanning tree algorithm
“Current transmission”
40-year
gen/load data
1. Generation planning
“Copper
Sheet”
Gen OM costs
(Various industry
sources)
No
Yes
500kV/765kV
800kV/600kV
Trans. data
Enough
Coordination?
4. Dynamic study and
robustness testing;
Plan evaluation
Candidate
circuits
3. Multi-stage network expansion
optimization using MILP
N-1 Security screening
Compare with
previous solution
Optimal transmission
design for given future
Robust Overlay
Design
16
Transmission Overlay Design
Reference
High Offshore
Transmission Candidate Set
High Geothermal
High Solar
17
Overlay and Speed Assessment
1) A given HVAC overlay is superimposed on a existing one (susceptances).
HVDC links are considered as speed/power sensitive to shift fluctuations.
Areas of low inertia are to be connected to areas of higher inertia via HVDC.
7
6
11
5
3
4
13
12
2
10
9
1
8
x 10
4
-2
-4
-5
x 10
0
-4
5
! 1 (p:u:) x 10
-4
-5
5
0
0
-2
0
-4
! 8 (p:u:)
! 5 (p:u:)
0
x 10
-4
-4
-5
5
! 1 (p:u:) x 10
0
-4
1
13 (p:u:)
-2
x 10
2) Modeling of overlay
technologies vs. slow
(‘governor’) dynamics.
-4
2
! 4 (p:u:)
0
5
4
2
! 3 (p:u:)
! 2 (p:u:)
2
x 10
-4
0
x 10
-3
! 3 (p:u:) x 10
5
-4
0.5
0
!
4
-4
-0.5
-5
-5
0
! 4 (p:u:)
5
x 10
-4
-5
-1
0
! 7 (p:u:)
1
x 10
-3
-1
-1
0
!
11 (p:u:) x 10
1
-3
3)
Reachability analysis is used to
generated bounds of speed deviations due to
fluctuation of renewables.
4) A ‘good’ overlay is the one that returns the
smallest volume in speed deviation (hypercube
volume to be tested with possible futures).
18
Conclusions
Developments supporting any transmission planning process:
• Multi-scenario study, value-based planning: Co-optimize generation, transmission
and transportation system.
• Transmission candidate circuit identification, multi-circuit/technology/stage
transmission expansion optimization, with locational differentiation among
generation and transmission costs.
• Application of reachability analysis for design verification in power system
planning.
• Transmission technologies HVAC and HVDC capabilities and limitations
(variability shifting).
Developments towards designing high-capacity interregional transmission :
• A process to performing such a design has been developed and illustrated.
• Several “good” designs have been identified, each of which increases overall
energy system resilience, reduces emissions, improves system planning flexibility,
facilitates renewable resources and reduces overall cost.
• An approach has been developed which accounts for slow power system
dynamics in developing the transmission design.
19
1.3 Robust and Dynamic
Reserve Requirements
•
•
•
•
•
•
PI: Prof. K. Hedman
Outline
Background and Motivation
Current Industry Practice
CAISO, ERCOT, and MISO
Achievements
Results
Future Work
20
Background and Motivation
• Reserve requirements: proxy method to ensure
reliability
• N-1 is not guaranteed due to congestion
• No guarantee of deliverability of reserves
• Project motivation
• Ensuring sufficient reserves (quantity + location) will
be increasingly more difficult with renewables
• Develop systematic tools to provide robust and
dynamic reserve requirements
• Improve reserve deliverability by developing dynamic
reserve zones
21
Current Industry Practice: Zones
• Today’s reserve zones are usually determined by:
• Utility ownership
• Identification of critical, congested paths
• Zones traditionally treated as static
• ERCOT Zones:
• Defined such that each generator or load within the
zone has a similar effect on commercially significant
constraints (CSC) (critical paths)
• ERCOT utilizes statistical clustering methods
ERCOT, “ERCOT Protocols, Section 7: Congestion Management” [Online].
Available at www.ercot.com/content/mktrules/protocols/current/07-070110.doc,
July, 2010.
22
CAISO, ERCOT, and MISO
* Zone 1 controlled by PJM
23
Achievements
• Developed systematic ways to determine
dynamic reserve requirements (zones and
levels)
• Improved reserve location/deliverability
• Transition from static rules to dynamic (operational
state dependent)
• Developed reserve rules for renewable resources
• Developed reserve rules flexible for network topology
changes
• Improvements in market efficiency and reliability
(reduction in out of market corrections)
24
Day Ahead Dynamic Reserves
Current Procedures
Dynamic Reserves
25
Dynamic Zone Numerical Results
Date
2-Jan
9-Jan
14-Jan
24-Jan
5-Feb
7-Feb
14-Feb
22-Feb
3-Mar
11-Mar
14-Mar
26-Mar
Ave. Load Shedding
Ave. Exp. Cost
Stochastic
Programming (Single
Zone)
5.285
4.51
3.51
6.3
6.475
6.3185
4.19
4.33
1.844
2.443
2.654
3.075
5.7355
3.043
4.89525
6.83925
3.25
3.2
3.2
2.68
3.757
7.525
2.45
4.184
Dynamic
Reserves (3
Zones)
2.23
2.53
1.6
2.4
2.72
2.675
2
2.33
1.65
1.5
2.43
2.07
4.25
4.23
2.17
0.7212
0.7356
0.7164
Traditional
Reserves (3 Zones)
Percent
Improvement
61.1%
16.9%
67.3%
64.9%
16.3%
16.4%
37.5%
13.1%
56.1%
80.1%
0.8%
50.5%
48.7%
2.6%
• IEEE118 test case
• Traditional reserves: max(largest contingency, NREL’s 3+5 rule) with
historical information to partition zones
• Dynamic reserves: partitions zones based on PTDF and probabilistic power
26
flow
Future Work
• Model refinement; post model + code online
• Investigate balance between deterministic,
proxy procedures (reserves) with stochastic
programming
27
Questions
Comments
28