The Role of Sensing, Estimation and Communications

The Role of Sensing, Estimation and Communications
in the Emerging Electric Energy Systems
Marija Ilić [email protected]
Electric Energy Systems Group (EESG) http://www.eesg.ece.cmu.edu/, Director
Invited keynote talk IEEESmartGridComm2014
Venice, Italy
November 2014
Outline
Basic vision for the emerging energy systems; new challenge
Today’s use of sensors, estimation and communications
Lessons learned from recent pilot experiments
Remaining major open problems/challenges:
--Need for new operating paradigm
--Bridge physics-based modeling and automation design to the
architecture design of communication networks
 Toward new paradigms
--Standards/protocols for operating the emerging power grids;
and/or
--Business models in support of innovative technologies at value
 Some open problems to support performance of new
paradigms
CMU on-going work: DyMonDS and simulations test beds
Basic Vision for Smart Grids:
Massive Systems Integration Opportunity and Challenge [1,2]
 Multi-national scale integration of coordinated complex systems:
- energy system (power grid, power electronics, energy resources)
- communication system (hardware and protocols)
- control system (algorithms and massive computation)
- economic and policy system (regulation and stimulus)
 Why? On-line IT enables:
- 20% increase in economic efficiency (FERC estimate)
- cost-effective integration of renewables and reduction of emissions
- differentiated Quality of Service without sacrificing basic reliability
- seamless prevention of blackouts
- expanding the infrastructure (generation, T&D, demand side) for
maximum benefit and minimum intrusion
 Who?
- Huge intellectual challenge – must be university led
- industrial partners include leading technologists in all four systems
- government partners
Key R&D Challenge— CPS Design of Future Electric Energy Systems
New technical problem:At present the physical
energy system, including its communications and
control, does not readily enable choice and multiparticipant information exchange and processing
for aligning often conflicting goals.
 It is essential to design intelligence for T&D
operations to align these goals and consequently
to make the most out of available resources
while simultaneously offering robust and
affordable quality of service.
New flexible energy processing equipment will
also be needed to handle increased variety and
bandwidth of many participant requests.
Key R&D Challenge— CPS Design of Future Electric Energy Systems
(2)
 New social science problem: The culture of managing vital services in
a robust way by assigning responsibilities and expectations, in
addition to technical and institutional design.
 New institutional problem: Integration can be achieved through
different institutional design which will have qualitatively different
impacts.
 Options include:
-Fully regulated monopolies and centralized planning and operations.
-Complete, carefully designed markets.
-Common set of interface standards and protocols.
-Common regulator level and/or lose cooperation of distributedregulators.
 The biggest challenge is to capture the interdependence of the
technical, social and institutional problems so as to reconcile the
need for choice, and affordable and robust electric energy services.
5
Basic cyber system today –backbone SCADA
The Role of State Estimation
Measurements
Static:
State
Standard
Deviation
Visualize System Operating
Point on Monitor.
Power
System
Measurements
AC State
Estimation
Estimated
Measurements
On-line DC OPF
Input System State to
Security Analysis Programs.
7
8
Today’s state estimation (SE)
Assumption:
Power System
 Changes slowly
Solution:
Previous State Estimate
Newton’s method
Current Global Optimum
8
Physical and Information Network Graphs Today
Network graph of the physical system
Information graph of today’s SCADA
Local serving entities (LSEs)
Load serving entities (LSEs)
Local Distribution Network (Radio Network)
PQ
Dies
el
PQ
Predicted  ,
Backbone Power Grid
and its
Local Networks (LSEs)
PQ
Wind
PQ
Information flow: MISO
State information exchange
Backbone
LS
E
LS
E
LS
E
LS
E
LS
E
LS
E
LS
E
LS
E
LS
E
LS
E
LS
E
LS
E
LS
E
LS
E
LS
E
LS
E
LS
E
LS
E
LS
E
LS
E
LS
E
Redundant
measurement sent to
Control center (hub)
LS
E
Potential of Measurements, Communications and Control
DLR
Constrained Line
Line-to-Ground Clearance
Transfer Capacity in Real
Time
PMU
Control
10
From preventive to corrective management of
future electric energy systems
System Load Curve
Every 10 min Real Time Load of NYISO in Jan 23, 2010
220
Forecasted Load
Actual Load with Disturbance
180
160
Real Power Load P(p.u.)
Real Power Load P (p.u.)
200
10 min Real Time Load of NYISO in Jan 23, 2010
140
120
0
5
Upper Bound
200
180
10
Lower Bound
0
10
Time (min)
15
Time (hours)
5
20
11
25
12
Predictable load and the disturbance
NYISO August 2006 Load Reactive Power Data for 3 Hours, Measurement Frequency = 0.50 Hz,Power Factor = 0.8
24
Pre-planed Load Value
Real Load Evolution, with 0.5Hz Sampling
23
NYISO August 2006 Load Data for 3 Hours, Power Factor = 0.8
21.1
Load Reactive Power Evolution (p.u.)
Load Reactive Power Evolution (p.u.)
21
20.9
20.8
20.7
20.6
20.5
20.4
21
20
19
18
17
16
20.3
0
22
20
40
60
80
100
Time (mins)
120
140
160
180
15
0
20
40
60
80
100
Time (mins)
120
140
160
180
12
Issues with SCADA today
Quite inaccurate transmission/sub-transmission SE
(mix of algorithmic, sensing and communications
problems); bad data, topology status and
computing intertwined.
Time synchronization problems (inconsistency of
planning and EMS data; mapping of models with
breakers into models w/o breakers)
Disconnect between T&D data management
Fast relays and primary controllers local; only
some experiments with SPS/coordinated RAS
13
Implications on system performance
Hard to implement corrective actions for
enhanced efficiency and reliability
Poor voltage/reactive power data prevent
optimization of voltage support
Disconnect of transmission/sub-transmission
and distribution SEs disables use of demand
response in operations
The trade-off between local enhanced
control/relays and faster communications
14
Utilities’ major concern: Reliability
Dynamical problems
Types of
Component
Small
signal
instab.
Transienti SS
nstab.
R
SSCI
Freq.
Volt.
Power flow
instab. Instab. imbalance
Synchronous
generators
?
?
?
?
?
?
?
Wind
generators
?
?
?
?
?
?
?
Solar plants
?
?
?
?
?
?
?
FACTS
?
?
?
?
?
?
?
Storage
?
?
?
?
?
?
?
Table 1.
15
Recent pilot experiments
Industry-government(-academia) collaborations on
hardware for smart grids
University campuses (``micro-grids”) –UCSD, IIT
Chicago
Utilities deploying AMIs, synchrophasors (PMUs)
Lessons learned—Familiarity with new smart
hardware
The remaining challenge (protocols for systematic
integration of scalable technologies at value)
16
Lessons learned from pilot experiments—familiarity with
new hardware (AMIs, PMUs, PHEVs, EVs, microgrids)
Energy Sources
Transmission Network
Electromechanical
Devices
(Generators)
Photo-voltaic
Device
Energy Sources
Load
(Converts
Electricity into
different forms of
work)
Electromechanical
Device
PHEVs
Demand
Respons
e
Future Smart Grid (Physical system)
Contextual complexity
ISO – Market Makers
FERC
Scheduling
Power Traders
Generator
Generator
Generator
XC
PUC
Transmission
Operator
Distribution
Operator
Some Utilities
Are all Three
Supply
Aggregators
Customer
Customer
Customer
Demand
Aggregators
Critical: Transform SCADA
 From single top-down coordinating management to
the multi-directional multi-layered interactive IT
exchange.
 At CMU we call such transformed SCADA Dynamic
Monitoring and Decision Systems (DYMONDS) and
have formed a Center to work with industry and
government on: (1) new models to define what is the
type and rate of key IT exchange; (2) new decision
tools for self-commitment and clearing such
commitments. \http:www.eesg.ece.cmu.edu.
New SCADA
DYMONDS-enabled Physical Grid
Aligning technical and social interactionsMaking the most out of the naturally available resources? [4]
“Smart Grid”  electric power grid
and ICT for sustainable energy systems
Core Energy
Variables
• Resource
system (RS)
• Generation
(RUs)
• Electric Energy
Users (Us)
Man-made Grid
• Physical network
connecting
energy
generation and
consumers
• Needed to
implement
interactions
Man-made ICT
•
•
•
•
Sensors
Communications
Operations
Decisions and
control
• Protection
• Needed to align
interactions
Typical challenge: From old to new paradigm—
Flores Island Power System, Portugal [11]
26
Controllable components—today’s operations
(very little dynamic control, sensing)
H – Hydro
D – Diesel
W – Wind
*Sketch by Milos Cvetkovic
Two Bus Equivalent of the Flores Island Power System
Generator
Diesel
8.15
8.15
Transmission
line
From Diesel
to Load bus
[]
0.3071
[]
0.1695
Base values
 = 10
 = 15
0.5917
State
Equilibriu
m
0.5917
0.9797
2.35
AVR
Diesel
Diesel
 []
400
 []
40
 []
0.02
 []
0.6
1.3
 
1/0.03
1
 []
0.2
0.0173
2.35
 
1
2.26
 []
0.8527
0.005
0.7482
Governor
 []
0.1667
 []
0.03
 []
1
0
0.01
0
Base values  = 10,  = 0.4
Information exchange in the case of Flores---new
(lots of dynamic control and sensing)
Transmission grid
PMU
DSO
DSO
DSO Module
PMU
SVC
Power-electronics
Module
Phasor
Measurement Units
Dynamic Purpose
Communication
Market and
Equipment Status
Communication
Smart Grid: More Local Optima Problems—Open
problems
Smart Grid Components:
Distributed
Generator Plants
Vehicle
Charging/Discharging
Various Demand
Management Software
Feature: More Uncertainties State Changes Fast
Previous Estimate
Current Global Optimum
 More Local
Optimums
Need for new operating paradigm
Different cyber support for different evolving power grids
(islands, micro-grids, distribution grids, transmission
grids)
 Remaining major open problems/challenges:
--Need for new operating paradigm
--Bridge physics-based modeling and automation
design to the architecture design of communication
networks
 Toward new paradigms
--Standards/protocols for operating the emerging power
grids
--Business models in support of innovative technologies at
31
value
Our proposal: TCP/IP like standards
 Given specified disturbances and range of operating
conditions within a known system:
- specified with e.g voltage, power
- similar to LVRT curves for wind turbines
- with specified duration
 All components (synchronous gens, wind gens) should
guarantee that they would not create any of the
problems in Table 1. (Clear objectives goals for
components, assigned responsibility for system
reliability)
 Two key questions: Q1-- Why does it matter?
Q2)--- Can this be technically done?
Not one way to achieve these!
32
A1: Examples of iBAs—it matters for ensuring both
reliable and efficient operations
iBA 1
iBA 3
Storage
iBA 2
33
Possible to create iBAs for meeting transient
stability distributed standard [13]
18
21
22
17
16
Given disturbance
Tripping of generator 1
20
19
15
14
13
11
24
23
12
iBA
3
9
10
6
4
5
8
1
2
7
34
Q2: Can we have a unifying theoretically sound
approach to TCP/IP like standards for smart grids?
Basic functionalities
Simple transparent TCP/IP like functionalities
Transparency based on a unifying modular
modeling of network system dynamics
Provable performance-difficult
Proposal—use interaction variables to specify
family of standards sufficient to avoid operating
problems
 Measure of how well modules balance themselves in steady
state
 Measure of rate of exchange of stored energy between a
module and the rest of the system over different time horizons
36
Value of dynamic system-wide coordination
 Minimal coordination by using an aggregation-based notion
of ``dynamic interactions variable”
Zoom-in
Zoom-out
…
38
Interaction variable-based two-layer model [12]
39
Physics/Model Based Spatial Scaling Up
CONFLICTING OBJECTIVES—COMPLEXITY
AND COST OF COMMUNICATIONS VS.
COMPLEXITY AND COST OF SENSORS,CONTROL
-SBA: Smart Balancing
Authorities (Generalization
of Control Area)
-IR: Inter-Region
-R: Region
-T: Tertiary
-D: Distribution
-S: Smart Component
-The actual number of
layers depends on
needs/technologies
available/electrical
characteristics of the grid
``SMART BALANCING AUTHORITY” CREATED IN A BOTTOM-UP WAY
(AGREGATION)--DyMonDS;
--COMPARE WITH CONVENTIONAL TOP-DOWN DECOMPOSITION
40
Market solutions
Fundamentally require alignment of technical
and financial/economic/policy interaction
variables
Primal/dual decompositions define
interaction variables in the generalized statespace
IT essential to support
Examples --- market-based integration of
renewables and demand response at value
41
Information flow within- and among modules
Based on the temporal decomposition of exogenous
inputs and control
Dynamic Monitoring and Decision Systems
(DyMonDS)-basis for algorithms
Working on computer platform
Video demonstrating SGRS; send us request for more
info.
NIST funded activity. Will be Webex demos for
industry, academia and government
42
Must simplify!
Utilities are having hard time adding all these
new components and their smarts for
simulating system-wide dynamics
Is there a ``smarter” way to model and define
modular functionalities so that the
interconnected system meets system-level
performance (Table 1)?
80% of each solution is modeling (PetarKokotovic,
Chalenges in Control Theory, Santa Clara, circa 1982)
43
Must proceed carefully…
The very real danger of new complexity.
Technical problems at various time scales lend
themselves to the fundamentally different
specifications for on-line data
No longer possible to separate measurements,
communications and control specifications
 Major open question: WHAT CAN BE DONE IN A
DISTRIBUTED WAY AND WHAT MUST HAVE FAST
COMMUNICATIONS
Thank you! Questions?
45
Smart Grid in a Room Simulator (SGRS) []
First demos of end-to-end SGRS
Physics-based modeling of diverse power
system components and their interactions
over time
 Software embedded into modules
Information exchange
(learned/communicated); the rest
private/secure
*Huge* R&D—what information to exchange
for what
Ilic,purpose?
M. DyMonDS Computer Platform for Smart Grids, PSCC 2014.
46
General Module Definitions
47
Information Exchange Between Modules
48
General Module Structure
49
Module Structure – Generator Example
50
51
Open questions
Interactive on-line practical software to
support system operator decisions and his
interactions with the grid users
Measurable software benefits (in terms of
reliability, efficiency and environmental
impacts)
***Academic challenge: Multi-physics
verifiable emulator of real-world power
systems***
Conclusions
 Our proposal: Interaction variable-based
 Standards/protocols for interactive iBAs can set the
basis for plug-and-play in smart grids—bounds on stored
energy change and on rate of change of stored energy
for T of interest
 Standards need to define transparent protocols for all
dynamic components
 Complexity of smart grids can be managed this way
 At the same time system performance is guaranteed
 With current NERC standards system performance
cannot be mapped into responsibilities of different
components
52
References
[1] Ilic, M., Toward a Multi-Layered Architecture of Large-Scale Complex Systems for
Reliable and Efficient Man-Made Infrastructures, Proc. of the MIT/ESD Symposium,
Cambridge, MA, March 29-31, 2004.
[2] Marija Ilic, IT-Enabled Electricity Services: The Missing Piece of the Environmental
Puzzle, Issues in Technology and Policy, 2011 IAP Seminar Series, MIT, Jan.,19,2011
[3] Yang Weng, “ Statistical and Inter-temporal Methods Using Embeddings for Nonlinear
AC Power System State Estimation, PhD thesis, CMU, ECE, August 2014
[4] ] ElinorOstrom, et al, A General Framework for Analyzing Sustainability of SocioEcological Systems, Science 325, 419 (2009).
[5] Ilic, M., E. Allen, J. Chapman, C. King, J. Lang, and E. Litvinov. Preventing Future
Blackouts by Means of Enhanced Electric Power Systems Control: From Complexity
to Order., Proc. of the IEEE, vol. 93, no. 11, pp. 1920-1941, November 2005.
[6] Ilic, Marija and Liu, Zhijian, “A New Method for Selecting Best Locations of PMUs
for Robust Automatic Voltage Control (AVC) and Automatic Flow Control (AFC),
IEEE PES 2010, Minneapolis, MN, July 25-29, 2010.
[7]MilosCvetkovic, “Power-Electronics-Enabled Transient Stabilization of Power Systems.
PhD thesis, CMU. ECE Department, May 2014
[8] Kevin D. Bachovchin, “Electromechanical Design and Applications in Power Grids of
53
Flywheel Energy Storage Systems, PhD thesis, CMU, ECE Dept. May 2015.
[9] M. Prica and M. Ilic, ” Maximizing Reliable Service by Coordinated Islanding and
Load Management in Distribution Systems under Stress” Proc. of the 16th Power
Systems Computation Conference, July 14-18, 2008, Glasgow, Scotland.
[10] SiriphaJunlakarn, “Differentiated Reliability Options Using Distributed Generation
and System Reconfiguration, PhD thesis, CMU. EPP Department, May 2015.
[11] Ilic, M. D, Le, X., Liu, Q. (co-Eds) Engineering IT-Enabled Sustainable Electricity
Services: The Tale of Two Low-Cost Green Azores Islands, Springer August 2013
[12] Ilic, et al, CDC 2014 tutorial.
[13]S.Baros, M.Ilic intelligent Balancing Authorities (iBAs) for Transient Stabilization of
Large Power Systems IEEE PES General Meeting 2014
[14] Smart Grids and Future Electric Energy Systems, 18-618, CMU, ECE, Fall 2014.
[15] Marija Ilic, Integration of Heterogeneous Small Test Beds for Emulating Large Scale Smart
Power Grids: The Emphasis on Cyber Architectures, MSCPES2014Workshop, Invited keynote
Berlin, CPS week, April 14,2014
[16] Ilic, Marija; Xie, Le; Joo, Jhi-Young, “ Efficient Coordination of Wind Power and
Price-Responsive Demand Part I: Theoretical Foundations, IEEE Trans. on Power Systems.
[17] Marija Ilic, Jhi-Young Jho, Le Xie, Marija Prica, NiklasRotering, A Decision-Making
 Framework and Simulator for Sustainable Energy Services, IEEE Transactions on
 Sustainable Energy, vol. 2, No. 1, January 2011, pp. 37-49.
54