Power%20System%20Dynamics%20Analysis

1
Power System Dynamics Analysis

• Complexity in the WSCC
• August 10, 1996 System Disturbance
• David P. Chassin, October 16, 2002
• Contact david.chassin_at_pnl.gov

6Nov98
2
August 1996 West Coast Blackout Summary of Events

• WSCC 8-10-96 1548 PAST major disturbance
• 4 islands, 7.5M customers out up to 9 hours
• Initiating event 500 kV Keeler-Allston flashover
  to trees
• Trip cause poor right-of-way maintenance
• Numerous supporting events
• Follow-up events
• COI limit lowered to 3200 MW (Aug 11)
• Account for limits at McNary (exciter) and The
  Dalles (fish)
• Raised Aug 12-14 to 3600 MW to avoid blackouts in
  CA

3
PNNL and Power Systems

• DOE Transmission Reliability Program
• Real-Time Grid Reliability Management
• Reliability and Markets, Load as a Resource
• Distributed Energy Resources Integration
• Infrastructure Assurance Outreach Program
• Utility vulnerability assessments
• Interaction with NERC, others on security topics
• Interest in SCADA security
• Energy System Transformation Initiative
• Integrated econophysics modeling and simulation
• Next generation power technology development

4
Wide-Area Measurement SystemDynamic monitor
network supports advanced analysis
Better information supports better - and faster
- decisions.
5
Real-Time System DataCollected from various
monitors throughout the grid
Bonneville Power Administration Phasor Data
Concentrator
6
Power System Measurement ToolsTools for managing
data signal analysis
7
Ringdown Analysis ToolAdvanced dynamic analysis
Extracting a linear model from measured data

• Dynamic analysis
• Model validation
• Control design

8
Advanced Dynamic System Analysis and Model
Validation
August 10, 1996 post-disturbance analysis
Determine if model calibration is needed.
Currently working with western utilities to
improve power grid models
9
Power System Econophysics
PJM 1999 Load/Price Probability Distributions
Q Qmax (1?1/P2)4.54
Abstract Transactive Machine
10
Grid Friendly Appliances
Grid-friendly appliances rapid, automatic
response to grid crises platform for active
communication control pre-heat/pre-cool to
coast through peaks utilize value thermal
storage increase reliability
security unnoticeable by consumer mass
customization/marketing
100
106
Loads and Reserves on a Typical U.S. Peak Day
Residential (non-GFA) 10
Industrial 28
Residential (GFA) 20
GFA potential exceeds US operating reserve
requirements!
Operating reserves 13
Commercial 29
given enough ants, you can move a
mountain!... impromptu reaction from a utility
power engineer
11
Overview of WSCC August 10, 1996 1548 PAST
Disturbance

• Based on WSCC Disturbance Report
• Approved by WSCC Operations Committee on October
  18, 1996

12
High northwest transmission loading

• High imports to CA
• Heavy imports from Canada and Idaho
• COI at 4750 MW
• Similar to conditions prior to 7/2/96 disturbance
• Warning signs visible
• Previous high-load issues
• Small power/large voltage swings
• Suggested voltage support problems in NW

13
Equipment out of service

• Lines
• 500 kV Big Eddy-Ostrander, John Day-Marion, and
  Marion-Lane (reactive support around Portland)
  flashover to trees
• 115 kV Allston-Rainer degraded hdwr,
  Longview-Lexington cable installation.
• Breakers
• 500 kV Marion, Keeler (modifications)
• Transformers
• 500/230 kV Keeler (modification)
• Static VAR Comp (SVC)
• Keeler reduced support to500 kV (tied to 230 kV
  side)

x
x
x
14
Triggering Events

• 154237 Keller-Allston
• Sags into trees, flashes, trips
• Overload par-lines hold 5 min
• McNary react-power at max
• 154729 St. Johns-Merwin
• Lines trip on relay malfunc.
• KA par-line loads increase
• 154736 Ross-Lexington
• Tree flashover and trip
• 207 MW from Swift lost
• System voltage sags
• 154736 McNary
• Units trip, exciter problem
• System power/voltage osc. begins
• ID-UT-CO-AZ-NM-NV Surge
• COI power flows down east side
• Out-of-step trips
• CA-AZ remain tied together

!
!
!
!
!
15
Final Result

• 28,000 MW of under-frequency load shedding
• 20,000 MW of undesired generation loss
• Four major islands
• Northern California (North of LA to Oregon
  border)
• Losses 11,600 MW load 7,900 MW generation
• Frequency excursions 58.54 60.7 58.3,
  restored in 2.5 hrs.
• Southern California (SoCal, NV, AZ, NM, El Paso,
  Baja)
• Losses 15 820 MW load 13,500 MW generation
• Frequency excursions 61.3 58.5, restored in 70
  min.
• Northern (BC, OR, WA, MT, WY, ID)
• Losses 2,100 MW load 5,700 MW generation
• Frequency excursions 60.4, restored in 7 minutes
• Alberta
• Losses 970 MW load 146 MW generation
• Frequency excursions 60.4 59.0, restored in 6
  minutes

16
Open Questions re. Complexity

• Report only identifies the incontrovertible
  points and fails to address more controversial
  questions
• Why was line maintenance inadequate?
• Why was system operated in single-contingency
  mode?
• Why did AZ-CA separation scheme fail to operate?
• Why did models fail to predict oscillations?
• Why did system damping fail?
• Lesson for us
• Dont go just by the official reports. Much more
  is not reported or discussed because of
  sensitivities. The social dynamics of a tightly
  knit community factors into what is knowable. We
  have to look deeper.

17
Problems Persist
480 MW generation dropped
2900 MW generation tripped
1400 MW Chief Joe brake inserted
350 MW load loss
86 MW UFLS
18
Questions and Comments