usa-argentina collaboration: smart grid applications in ......the 1996 blackout in the western north...
TRANSCRIPT
USA-Argentina Collaboration: Smart Grid Applications in Transmission Systems PRESENTER: MARCELO A. ELIZONDO, RESEARCH ENGINEER
September 18, 2013 1
Electricity Infrastructure Group, Energy and Environment Division Buenos Aires, Argentina
PNNL-SA-98382
USA-Argentina Collaboration
Smart Grids (SG): Advances in communications and computation to address power systems challenges U.S. DOE and PNNL expertise developing SG technologies in both transmission and distribution SG technology to address Argentina’s needs Summary of smart grid activities at PNNL on transmission systems. Several PNNL staff contributing. Marcelo’s background in USA and Argentina
UNSJ, San Juan, Argentina. Power system engineer. CONICET, Argentina. Scholarship
UNSJ, Argentina. PhD, power systems engineering. Carnegie Mellon University, USA. 2-year visitor.
Mercados (ME-C), Argentina. Consultant engineer. PNNL, USA, research engineer (Electricity Infrastructure Group)
September 18, 2013 2
Outline
PNNL and its Electricity Infrastructure Group Four examples of developments
Catalyst role in PMU technology development & mode meter Model validation Technology enabling finer selection of load shedding Control center tools for renewable integration and dynamic interchange
Open discussion: opportunities in Argentina
September 18, 2013 3
Who is PNNL?
4
The Pacific Northwest National Laboratory is a DOE Office of Science laboratory in Richland, WA. Operated since 1965 by Battelle, a global non-profit research and development organization committed to science and technology for the greater good.
Mission: Transform the world through courageous discovery and innovation. Vision: PNNL science and technology inspires and enables the world to live prosperously, safely, and securely. Values: Integrity, creativity, collaboration, impact and courage provide the foundation for all we do.
PNNL employs nearly 5,000 staff and has an annual operating budget of $1.1 billion. PNNL has over 100 people working in Electric Infrastructure – over 50 Power System Engineers.
Slide from: Paul Skare, PNNL September 18, 2013
PNNL draws upon core capabilities, facilities, and investments in Electric Infrastructure
September 18, 2013 5
Power system operation, planning and security
Power markets
Demand response
Renewable integration
Advanced analytic methods, HPC-based simulations, visualization
Staff Capabilities
Live PMU data from all three interconnections
PMU data archive
PowerNET lab
EMS/DMS displays
T&D-level data displays
Platform for tool evaluation, operator training
Physical Control Center (EIOC)
Live security data streams
Visual analytics
Co-located with classified assets that accelerate threat recognition and appropriate response
Emergency Response
Public / Private
Cyber Security / Resilience Center (EICC)
Networking and data management
Advanced analytic methods and HPC approaches for real-time modeling and simulation
Visualization and decision support
Next Generation EMS
Next Generation Simulation
Future Power Grid Initiative
Slide from: Paul Skare, PNNL
Outline
PNNL and its Electricity Infrastructure Group Four examples of developments
Catalyst role in PMU technology development & mode meter Model validation Technology enabling finer selection of load shedding Control center tools for renewable integration and dynamic interchange
Open discussion: opportunities in Argentina
September 18, 2013 6
7
Impact: Mode Meter
If used in ‘96, the mode meter would have provided operators the necessary lead time (4-6 min) to potentially avoid a $2B blackout .
Our approach: Improve power system performance and transmission reliability by extracting greater value from grid measurements and data. Key elements include:
U.S. Department of Energy (DOE) lead for the North American Synchrophasor Initiative (NASPI) – joint effort with the North American Electric Reliability Council (NERC) and industry to build out phasor measurement units (PMUs) across North America, enabling increased situational awareness and control Planning models validation through measurement-based analysis Decision support tools for operators
Mode meter – uses PMU data to improve detection of grid disturbances, enabling greater asset use and preventative measures; deployed in Western Interconnection Synchrophasor Project
Electricity Infrastructure Operations Center – providing utilities, vendors and researchers access to real-time grid data for testing in realistic operations environment
National Challenge
Ensure a reliable U.S. power system by
leveraging new data streams that provide
wide-area visualization,
monitoring and control.
Overview Transmission Reliability
Impact: Mode Meter
If used in ‘96, the mode meter would have provided operators the necessary lead time (4-6 min) to potentially avoid a $2B blackout .
Graphical Contingency Analysis
Real-time power flow visualization identifies/prioritizes issues, recommends corrective actions
NASPI
7 September 18, 2013 Slide from: Jeff Dagle, PNNL
North American SynchroPhasor Initiative
8
“Better information supports better - and faster - decisions.”
DOE and NERC are working together closely with industry to enable wide area time-synchronized measurements that will enhance the reliability of the electric power grid through improved situational awareness and other applications
April 2007 November 2012
http://www.naspi.org
Lessons Learned from August 10, 1996 Blackout in the Western Interconnection of North America
9
4000
4200
4400
4600
0 10 20 30 40 50 60 70 80 90
4000
4200
4400
4600
Time in Seconds
Simulated COI Power (initial WSCC base case)
Observed COI Power (Dittmer Control Center)
Data captured from high-speed time-synchronized Wide Area Measurement Systems were essential to support the blackout investigation
The need for better model validation was demonstrated
September 18, 2013 Slide from: Jeff Dagle, PNNL
Mode meter: new grid disturbance detection system using real-time phasor data
Challenge Today: limited real-time monitoring of grid oscillations that cause instability and result in constrained operations (renewables integration/demand response will increase instability) PNNL Approach: use real-time phasor data to improve oscillation detection, enabling increased grid asset utilization and preventative measures Status:
Mode meter to be used in Western Interconnection Synchrophasor Project to monitor oscillations Licensed to commercial energy management system providers for incorporation into utility software
The 1996 blackout in the western North American power system illustrated the critical need for an advanced view of grid disturbances before they cascade to outages. If used in 1996, the mode meter would have provided operators the necessary lead time (4-6 min) to potentially avoid the blackout.
Slide from: Jeff Dagle, PNNL
Situational awareness
Congestion management
Renewable integration
Increase in operating transfer capacity
System protection
REAL-TIME SYNCHROPHASOR APPLICATIONS AND THEIR PREREQUISITES
FUTURE Prerequisites
Applications
ANALYSIS
COMMUNICATIONS
Good data collection
Interconnection-wide baselining
System studies
High availability, high speed
TODAY
Functions
Appropriate physical & cyber-security
Redundant, fault-tolerant
Outage avoidance
Wide-area Monitoring Visualization Frequency and voltage monitoring Oscillation detection
Event detection Alarming
Operator decision support
Automated wide-area controls Reliability Action Schemes
USERS
Pattern detection
Model validation – system & elements
Familiarity Good visual interface Training
Interoperability standards
Synchrophasor Deployment Summary
DOE and PNNL has played a key catalyst role in the development and implementation of synchrophasor technology PNNL will continue to support DOE’s efforts to promote and enable widespread adoption of advanced monitoring technologies to ensure grid reliability DOE is actively supporting needed R&D to ensure that the full value of a North American phasor network will be realized • Hardware – measurement technologies • Network – data access and security • Software Applications – focus on reliability management objectives • Demonstrations – regional in scope
The American Recovery and Reinvestment Act of 2009 has provided
funding for investment grants and demonstration projects. This has enabled unprecedented advancement of synchrophasor technology deployment
September 18, 2013 12
Slide from: Jeff Dagle, PNNL
Outline
PNNL and its Electricity Infrastructure Group Four examples of developments
Catalyst role in PMU technology development & mode meter Model validation Technology enabling finer selection of load shedding Control center tools for renewable integration and dynamic interchange
Open discussion: opportunities in Argentina
September 18, 2013 13
Some Definitions
Model Validation – diagnosis of model problems Determine mismatch between model and measurement
Model Calibration – solution to model problems
Tune parameters to minimize the mismatch Model Verification – confirmation of good solutions
Compare model against measurement from multiple disturbance recordings to ensure reliable estimation
September 18, 2013 14 Slide from: Henry Huang and Ruisheng Diao, PNNL
Model Validation and Calibration Methodology
Model validation through event play-in using phasor data
15
Remainder of Power system Model
Subsystem for hybrid modeling
Remainder of Power system Model
(external)
Measurement boundary
Subsystem for hybrid modeling
Measured V,f(θ)
P,Q
Simulated P,Q
Measured P,Q = ?
Calibration
“Prediction” Dynamic
Simulation dx/dt = f(x, y, α)
Dynamic states x(k-1)
Parameters α(k-1)
Measurement z(k)
x(k), α(k)
x’(k)
“Correction” Measurement
Equations ∆ z = z – h(x, α)
k: time step
Event Playback
Model calibration using Kalman Filter
Slide from: Henry Huang and Ruisheng Diao, PNNL
Model Validation and Calibration
Conventional power plant model validation and calibration Device-level model validation and calibration Renewable generation model validation and calibration System-wide model validation and calibration Real-time model validation and calibration
September 18, 2013 16 Slide from: Henry Huang and Ruisheng Diao, PNNL
Power Plant Model Validation
September 18, 2013 17
Mismatch between Simulation and Measurement
-790
-785
-780
-775
-770
-765
-760
-755
-750
-745
-7405 10 15 20 25 30 35 40
t (s)
P (M
W)
P3_ref P3_sim
P4_ref P4_sim-50
-30
-10
10
30
50
70
90
5 10 15 20 25 30 35 40
t (s)Q
(MW
)
Q3_ref Q3_sim
Q4_ref Q4_sim
(Data from August 18 2002, 1500 MW Navajo units #2 & #3 drop)
Slide from: Henry Huang and Ruisheng Diao, PNNL
Calibration of Inertia Constant and Exciter Gain
18
0 10 20 30 40-7.8
-7.6
-7.4
t in seconds
real
pow
er fl
ow p
er u
nit
0 10 20 30 403.5
4
4.5
5
5.5
t in seconds
Iner
tia C
onst
ant
0 10 20 30 40300
320
340
360
380
t in seconds
Exc
iter G
ain
0 10 20 30 40-0.6
-0.4
-0.2
0
0.2
t in seconds
reac
tive
pow
er p
er u
nit Measurement
Calibration
4.98
311
(Data from August 18 2002, 1500 MW Navajo units #2 & #3 drop)
Slide from: Henry Huang and Ruisheng Diao, PNNL
Methodology
Automated model validation wrapping around PSLF BPA EPCL codes + PNNL MATLAB codes
Standalone MATLAB codes for model calibration A GUI is available User-selectable for what parameters to be calibrated.
Goal: Integrate calibration with PSLF (PSSE, …)
19
MATLAB
PSLF Play-in .sav .dyd .csv
PLOT
• Plots: model validation
• Calibrated parameters • Plots: model validation,
model calibration, model verification
Play-in Calibration
Data Handling
Call PSLF Call PLOT
Model Validation Model Calibration
Outline
PNNL and its Electricity Infrastructure Group Four examples of developments
Catalyst role in PMU technology development & mode meter Model validation Technology enabling finer selection of load shedding Control center tools for renewable integration and dynamic interchange
Open discussion: opportunities in Argentina
September 18, 2013 20
Grid Friendly™ Appliances Provide Fast, Autonomous Reliability Resource
September 18, 2013 21
Autonomously detects under-frequency events and sheds load 150 new Whirlpool clothes dryers, 50 retrofitted water heaters No one noticed in hundreds of curtailment events! Can displace spinning reserves and increase reliability Reacts within 1/2 second Delays & randomizes service restoration to avoid grid shock Low cost: no communications required
Slide from: Robert Pratt, PNNL
“When the inevitable occurs … people get stuck in elevators and high-value uses of power are shut off along with all the lowest priority uses of energy. It's the meat-ax approach to interrupting power flows.” Dr. Vernon Smith, 2002 Nobel Prize Winner, Economics
Outline
PNNL and its Electricity Infrastructure Group Four examples of developments
Catalyst role in PMU technology development & mode meter Model validation Technology enabling finer selection of load shedding Control center tools for renewable integration and dynamic interchange
Open discussion: opportunities in Argentina
September 18, 2013 22
23
December 2008
Simulation, Modeling, and Control
Integration of forecasting and renewable energy production tools into grid resource planning and operation tools
Forecasting generation capacity and ramp ranges needed to balance the system (orange bands)
Incorporating all sources of uncertainty/variability: wind and solar generation and demand
Example Outcome: predicting generation deficiency above the available range (gray band)
Tool is installed at the CAISO Control Center to help real-time dispatchers anticipate and address ramping needs.; Planned deployment to other ISOs
* Patricia Hoffman (Assistant Secretary, DOE ), “Maximizing Renewable Energy in the US Electric Grid,” Presentation at The Road to a 100% Renewable Energy System Workshop, August 1, 2011
Ramp and Uncertainty Prediction Tool
Slide from: Yuri Makarov and Pavel Etingov, PNNL
Focus Areas
September 18, 2013 24
Sources of uncertainty (discrete and continuous): • Traditional:
Load forecast Unexpected forced
outages Generation units failure
to startup • New challenges:
Wind generation forecast Solar generation
forecast Wind ramps
Balancing process: • Day-ahead schedules
Unit commitment Economic dispatch
• Real-time dispatch Unit commitment Economic dispatch
• Spinning reserve • Regulation • Contingency reserve
• Control performance • Reliability
Slide from: Yuri Makarov and Pavel Etingov, PNNL
Models for Balancing Requirements Uncertainty: Multidimensional Uncertainty Analysis
25
Existing approaches are frequently limited to one dimension of the uncertainty problem – capacity. But capacity is not a single sufficient descriptor of the problem. Operational performance of a power system can be demonstrated through four basic metrics, forming the “first performance envelope”.
Capacity (π). Ramp rate (ρ). Ramp duration (δ). Energy (Є).
Ramp Duration, min
Energy, MWh
Capacity,MW
Time
MW
Ramp Rate, MW/min
Net Load ORLoad Following ORegulation Curve
* Y.V. Makarov, C. Loutan, J. Ma, and P. De Mello. “Operational impacts of wind generation on California power systems,” IEEE Transactions on Power Systems, vol. 24, pp.1039–1050, May, 2009.
Probabilistic Tool Design
September 18, 2013 26
Forecast error, %
-20 -10 0 10 20
Statistical analysis Cumulative distribution function
1. Data acquisition 2. Uncertainty assessment through
statistical analysis 3. Prediction of future grid
balancing requirements for specified time horizons and confidence levels
1
2
3
Slide from: Yuri Makarov and Pavel Etingov, PNNL
Real-Time Requirements Screen
27 Slide from: Yuri Makarov and Pavel Etingov, PNNL
Dynamic Interchange Adjustment (DINA) Tool
September 18, 2013 28
Objectives of PNNL dynamic interchange adjustment (DINA) tool: Provide an online estimation of the secure range for possible ISO-NE intra-hour net interchange adjustments within the next dispatch interval. Consider all contributing factors, such as:
expected changes to system load and interchange, spinning reserve requirements, system ramping capability, relevant uncertainties significantly impacting this estimation, including:
load forecast errors, random interchange variations, uninstructed deviations on generators, uncertainties in generation dispatch and reserve requirements.
Use ISO-NE market and other data in an agreed form Slide from: Yuri Makarov and Pavel Etingov, PNNL
Example of Actual Interchange Adjustment Range Estimation
29
Summary
PNNL experience Catalyst role in PMU technology development & Mode meter Model validation Technology enabling finer selection of load shedding Control center tools for renewable integration and dynamic interchange
Advancements to improve system reliability and resilience Opportunities in Argentina?
September 18, 2013 30
Thank you
Marcelo A. Elizondo, PhD Research Engineer
www.pnnl.gov
September 18, 2013 31
PNNL-SA-98382