Smart Grid Resilience: Extreme Weather, Cyber-Physical Security, and System Interdependency

Preface
Contents
Part I Extreme Weather and Cascading Failure
1 Cascading Failures Under Extreme Temperatures
1.1 Introduction
1.2 Ambient Temperature in Blackout Modeling
1.2.1 Temperature Disturbance
1.2.2 Load Change Under Temperature Disturbances
1.2.3 Dynamic Line Rating
1.2.4 Probability of Line Tripping
1.2.5 Probability of Generator Tripping
1.3 Modeling Protection and Control Strategies
1.3.1 Undervoltage Load Shedding
1.3.2 Operator Re-dispatch
1.4 Timing of Events
1.5 Voltage Stability Margin Calculation
1.6 Blackout Model Considering Temperature
1.7 Simulation Results
1.7.1 Model Implementation and Parameter Settings
1.7.2 Typical Simulation Run Without Operator Re-dispatch
1.7.3 Typical Simulation Run with Operator Re-dispatch
1.7.4 Number of Simulations
1.7.5 Impact of Temperature Disturbances and Size of Selected Area
1.7.6 Identifying the Most Vulnerable Buses/Locations
1.7.7 Impact of Control Strategies
References
2 Cascading Failure Interaction Analysis
2.1 Introduction
2.2 Estimating Interactions Between Component Failures
2.3 EM Algorithm
2.3.1 A Coin-Flipping Example
2.3.2 Mathematical Foundation
2.4 Estimating Component Failure Interactions by EM Algorithm
2.5 Determining the Number of Cascades Needed
2.5.1 Lower Bound of M
2.5.2 Lower Bound of Mu
2.6 Results
2.6.1 Number of Cascades Needed
2.6.2 Interaction Matrix and Interaction Network
2.6.3 Identified Key Links and Key Components
2.6.4 Validation of Estimated Interactions
2.6.5 Cascading Failure Mitigation
2.6.6 Efficiency Improvement
References
3 Integrated Preventive and Emergency Responses
3.1 Introduction
3.2 Integrated Resilience Response
3.3 Mathematical Formulation
3.3.1 Preventive Response
3.3.2 Damage From Natural Disasters
3.3.3 Emergency Response
3.3.4 Integration of Preventive and Emergency Responses
3.4 Solution Methodology
3.4.1 NC&CG Decomposition-Based Algorithm
3.4.2 Computational Efficiency Improvement Techniques
3.5 Results on PJM Five-Bus System
3.6 Results on IEEE One-Area RTS-96 System
3.7 Results on IEEE Three-Area RTS-96 System
Appendix 1: PJM Five-Bus System
Appendix 2: IEEE One-Area RTS-96 System
References
Part II Cybersecurity of Smart Grid Monitoring
4 Risk Mitigation against Cyber Attacks Based on Dynamic State Estimation
4.1 Introduction
4.2 Power System Dynamic Model
4.2.1 10th-Order Nonlinear Power System Model
4.2.2 Linearized Power System Model
4.3 Unknown Inputs & Attack-Threat Model
4.3.1 Modeling Unknown Inputs
4.3.2 Modeling Cyber Attacks
4.4 DSE under UIs and CAs
4.4.1 Sliding-Mode Observer for Power Systems
4.4.2 SMO Dynamics & Design Algorithm
4.5 Asymptotic Reconstruction of UIs & CAs
4.5.1 Estimating Unknown Inputs
4.5.2 Estimating CAs
4.5.3 Attack Detection Filter
4.6 Risk Mitigation—A Dynamic Response Model
4.6.1 Weighted Deterministic Threat Level Formulation
4.6.2 Dynamic Risk Mitigation Optimization Problem
4.6.3 Dynamic Risk Mitigation Algorithm
4.7 Case Studies
4.7.1 Scenario I: Dynamic Reconstruction of UI & DSE
4.7.2 Scenario II: DSE Under UIs & CAs
References
5 Comparing Kalman Filters and Observers Against Cyber Attacks
5.1 Introduction
5.2 4th-Order Nonlinear Power System Model
5.3 Model Uncertainty and Cyber Attacks
5.3.1 Model Uncertainty
5.3.2 Cyber Attacks
5.4 DSE Algorithms
5.4.1 Kalman Filters for Power System DSE
5.4.1.1 EKF
5.4.1.2 UKF
5.4.1.3 CKF
5.4.2 Nonlinear Observers for Power System DSE
5.5 Numerical Results
5.5.1 Scenario 1: Data Integrity Attack
5.5.2 Scenario 2: DoS Attack and Scenario 3: Replay Attack
5.5.3 Discussion on Model Uncertainty Estimation
5.5.4 Discussion on Cyber Attack Detection
5.5.5 Non-Gaussian Measurement Noise
5.5.6 Computational Efficiency
5.6 Summary
References
6 Self-Healing PMU Network Against Cyber Attacks
6.1 Introduction
6.2 Motivation for Self-Healing PMU Network
6.3 System Model
6.3.1 Power System Observability
6.3.2 Rules in Network Switches
6.4 Optimization Formulation
6.4.1 PMU Connection Status Constraints (PCSCs)
6.4.2 Power System Observability Constraints (PSOCs)
6.4.3 PDC Connection Space Constraints (PSCs)
6.4.4 PMU Reconnection Constraints (PRCs)
6.4.5 Switch Rule Space Capacity Constraints (SRSCCs)
6.4.6 Routing Policy Constraints (RPCs)
6.4.7 Endpoint Policy Constraints (EPCs)
6.4.8 Optimization for Self-Healing Mechanism
6.4.8.1 Stage 1: Recover System Observability
6.4.8.2 Maximize System Observability
6.5 Greedy Heuristic Algorithm
6.6 Case Study on IEEE 30-Bus System
6.7 Performance Evaluation of Stage 1
6.7.1 Impact of Scale of Attacks
6.7.2 Impact of Hardware Resources
6.8 Performance Evaluation of Stage 2
References
Part III Cyber-Physical Security for Distributed Energy Resources
7 Cyber-Physical Security Research Framework for Distributed Energy Resources
7.1 Introduction
7.2 Cyber-Physical Power System with Large-Scale DER Deployments
7.2.1 Generic Architecture of Power Systems with DERs
7.2.2 Challenges of Maintaining DER Cybersecurity
7.3 Overview of DER Cyber-Physical Security Research Framework
7.4 Potential Cyber Attacks on Cyber-Physical Power System with DERs
7.4.1 Cyber-Physical-Threat Modeling
7.4.2 Threat Scenarios Targeting DER
7.4.3 Attack Threat Ranking
7.5 Attack Impact Analysis and Metrics
7.6 DER Cyber-Physical Security Design Principles
7.7 Attack Resilience at Cyber, Physical Device, and Utility Layers
7.7.1 Cyber Layer Attack Resilience
7.7.1.1 Cyber Layer Attack Prevention
7.7.1.2 Cyber Layer Attack Detection
7.7.1.3 Cyber Layer Attack Response
7.7.2 Physical Device Layer Attack Resilience
7.7.2.1 Physical Device Layer Attack Prevention
7.7.2.2 Physical Device Layer Attack Detection
7.7.2.3 Physical Device Layer Attack Response
7.7.3 Utility Layer Attack Resilience
7.7.3.1 Utility Layer Attack Prevention
7.7.3.2 Utility Layer Attack Detection
7.7.3.3 Utility Layer Attack Response
References
8 Distributed Load Sharing Under Cyber Attacks
8.1 Introduction
8.2 Inverter-Based Microgrid Structure
8.2.1 Physical Layer
8.2.2 Cyber Layer
8.3 System Dynamic Model
8.3.1 Small-Signal Model
8.3.2 Active Power Reference
8.4 System Performance Under Attack
8.4.1 FDI Attack Against Distributed Load Sharing Control
8.4.2 Effects of FDI Attack on Microgrid Performance
8.5 Case Studies
8.5.1 Stable Region
8.5.2 System Performance Under Attack Strategy 1
8.5.3 System Performance under Attack Strategy 2
References
9 Deep Learning Based Attack Detection for Microgrid Control
9.1 Introduction
9.2 Distributed Control and FDI Attack
9.2.1 Cyber-Physical Representation of AC Microgrids
9.2.2 Secondary Control Problem Formulation
9.2.3 Distributed Control Algorithm
9.2.4 FDI Attack Against Distributed Control
9.3 Deep Learning Based Multi-label Attack Detection
9.3.1 Multi-label Classification Problem Formation
9.3.2 Data Preparation and Preprocessing
9.3.3 Deep Learning Models
9.4 Performance Evaluation
9.4.1 Test System and Control Performance
9.4.2 Deep Learning Performance Metrics
9.4.3 FDI Attack Detection Results
References
Part IV Smart Grid Resilience Under System Interdependency
10 Interdependency Between Power System Outages by Branching Process
10.1 Introduction
10.2 Estimating Multi-type Branching Process Parameters
10.3 Estimating Joint Probability Distribution of Total Outages
10.3.1 n-Type Branching Process
10.3.2 Two-Type Branching Process
10.3.3 Validation of Estimated Joint Distribution
10.4 Number of Cascades Needed
10.4.1 Determining Lower Bound for M
10.4.2 Determining Lower Bound for Mu
10.5 Estimated Parameters of Branching Processes
10.6 Estimated Joint Distribution of Total Outages
10.7 Predicted Joint Distribution from One Type of Outage
10.8 Estimated Propagation of Three Types of Outages
References
11 Interdependency Between Power System Outages by Coupled Interaction Model
11.1 Introduction
11.2 Coupled Interaction Matrix
11.2.1 Definition of Coupled Interaction Matrix
11.2.2 Definition of an Auxiliary Matrix
11.3 Estimating Coupled Interaction Matrix by EM Algorithm
11.4 Coupled Interaction Model for Cascading Failure Simulation
11.5 Critical Link Identification
11.6 Coupled Interaction Network of IEEE 300-Bus System
11.7 Validation of Coupled Interaction Model
11.8 Choosing Critical Links for Mitigation
11.9 Cascading Failure Mitigation
Appendix: Discretization Unit for Each Load Bus
References
12 Interdependency Between Smart Grid and Transportation Network
12.1 Introduction
12.2 Mathematical Modeling
12.2.1 Renewable Energy Investor Modeling
12.2.2 Conventional Generators
12.2.3 ISO Modeling
12.2.4 Driver Modeling
12.2.5 Market Clearing Conditions
12.3 Computational Approach
12.4 Results on Three-Node Test System
12.4.1 Effects on Equilibrium Prices
12.4.2 Effects on Renewable Investment
12.4.3 Effects on System Costs
12.4.4 Effects on Flow Distribution
12.5 Results on Sioux Falls Road Network and IEEE 39-Bus Test System
Appendix: Proofs
References
Index