DIAGNOSING VULNERABILITY, EMERGENT PHENOMENA, and VOLATILITY in MANMADE NETWORKS

1
MANMADE
MANMADE
DIAGNOSING VULNERABILITY, EMERGENT PHENOMENA, and
VOLATILITY in MANMADE NETWORKS
On the implementation of GIS and vulnerability
assessment of networks to natural hazards
F. Bono, E.Gutierrez, K.PoljansekJoint Research
Centre, Ispra - ITALY
September 25th 2009
2
Motivation
Critical Infrastructures
Electricity
Gas pipeline disruptions
Complex System Vulnerability
HV grid
Gas
Hazards
3
Gas disruptions
Failure to manage gas supply interruptions
properly and efficiently can result in widespread
disruptions in the supply of gas to industry and
gas-fired power generation plants.
Threats to gas supplies
Terrorism-related hazards Natural disasters
Materials failure Other hazards
4
Gas Failures
EGIG Report European Gas Pipeline Incident Data
Group
5
Electricity disruptions
UCTE Year 2006 82 events
6
Approach
Mathematical Methods
Fragility Curves
Probability of Failure
Adjacency Matrix
Graph Network
7
Overview

• Introduction
• Geographical Information Systems (GIS)
• GIS and networks
• Data processing issues
• The EU interconnected energy network

8
Introduction

• The recent improvements of GIS has led to the
  acquisition of massive data sets on territorial
  and infrastructural elements (transportation,
  utilities and social networks with a geographical
  component) that were previously unavailable.
• Georeferenced data are a precious commodity that
  allows us to compile and study man-made networks,
  provided that a convenient preprocessing of the
  data is performed.
• GIS data mining and processing methods to
  effectively combine heterogeneous data sources
  into meaningful real-world networks are
  presented, with an insight into the problems and
  solutions adopted within the ManMade project.

9
Geographical Information Systems (GIS)
Raster
Satellites
GIS
Gridded data
Maps
Database
Vectors
A geographic information system (GIS) integrates
hardware, software, and data for capturing,
managing, analyzing, and displaying all forms of
geographically referenced information.
10
GIS data
Vector data model. Discrete features, such as
infrastructural elements, are usually represented
using the vector model.
Raster data model. Continuous numeric values,
such as elevation, and continuous categories,
such as population distribution, are represented
using the raster model
LandScan Global Population Databases
GSHAP
Global Seismic Hazard Assessment Program
11
GIS and Network Analysis
12
GIS and networks

• GIS Network Analysis basic functionalities
• Topology issues
• Generation of interconnected networks from
  disjoint GIS data

13
Topology Issues
GIS
Networks
Network Analysis

• Nodes degree
• Strong Components
• Loops
• Closed Cycles

Map Images georefrencing
14
The Energy Network
Hydro Power
Power Plants (coal, nuclear, )
Electricity network
Wind Farms
Gas Power Plant
Other sources (solar, geothermic, )
LNG storage and extraction
Electricity Consumption
LNG terminal
GAS network
15
Relating network elements
GIS gas and electricity networks
16
Condensing the network

• Working with networks at continent level
• Focus on major networks
• Relating objects connected via the minor network

17
Inter-network connectivity
Plants and grids connections
Shortest path (red line) between a power plant
and the substation on the main network the
geographically closest substation is not the one
to be associated with the plant
18
Virtual Edges
Breadth first search of the shortest paths
between a power station and the substations on
the main network
19
Setting the hazard level on the networks elements
SOURCE The Global Seismic Hazard Assessment
Program (GSHAP)
Seismic hazard Map of Europe10 Probability of
exceedance in 50 Years, 475-Year Return Period
20
Gas Sources
LNG Terminals
Natural Gas Production Regions
Natural Gas Producing Countries
Open Sea gas nodes
21
Distribution substations
DEFINITION
Paris Urban Area

• One degree vertices or
• More degree vertices
• - inside urban area or
• connected to distribution
• network

- Transmission substations - Distribution
substations
22
Generation of adjacency matrices
4. GIS import of computed values
1. GIS data extraction
2. Network grid compilation(IDfrom, IDto, value)
3. Matlab and Pajek data import and processing
23
The interconnected Network
Gas pipelines 2212 nodes 2644 edges 75 Arcs 21
LNG teminals
Directed Network
Gas Sources
Electricity grid 5127 subs 6726 edges
Gas node
Power Plants
Power plants 998 Natural Gas 4383 Others
Subs
24
Approach and data requirements
NETWORK

• Topology
• Location
• Analysis of unweighted
• directed network

Seismic hazard
Interdependency effect

• Interoperability matrix

• Sesimic hazard maps
• Fragilty curves of the elements

NETWORKS response
Probabilistic reliability method MONTE CARLO
SIMULATION
Risk Hazard x Response
25
Fragility curves of elements
Vertices (PGA)
Edges

• Gas network
• Pipelines (material, joint type)

- Ground failure (PGD) - breaks - Seismic wave
propagation (PGV) -leaks
(un)anchored subcomponents

• Electricity network
• Transmission lines

FEMA. HAZUS-MH earthquake model (2003)
O'Rourke, Ayala (1993), Wald et al (1999)
26
Probabilistic reliability model
ELEMENT LEVEL
NETWORK LEVEL
ANALYSIS
Propagation of probabilities of failures
We know the probability of failure of each element
We would like to know the probability of failure
of the whole network
27
Performance measures
Albert et al.,Physical Review E 2004.
number of the source vertices connected to the
i-th sink vertex
CONNECTIVITY LOSS
the sum of the power plants power connected to
i-th distribution substation
POWER LOSS
Electricity network
IMPACT FACTOR ON THE POPULATION
the number of people assigned to i-th
distribution substation
network characteristic ranging 0-100 used
for the definition of network damage states
28
Modelling of network interdependency
Interdependency is a bidirectional relationship
between two infrastructure through witch the
state of each infrastructure influences or is
correlated on the state of the other (Rinaldi et
al. 2001,IEEE Control Systems).
(Dueñas-Osorio et al. ,2007, EESD)
INTEROPERABILITY MATRIX
GAS
ELECTRICITY
GAS
ELECTRICITY
29
Interoperability matrix
Physical interdependence
adjacent vertices
fuel
Direction of interaction
x
GAS
ELECTRICITY
Strength of coupling
30
Independent network analysis
20 CL
50 CL
80 CL
475 year
1.1g
0.9g
0.6g
0.4g
0.3g
Damage states
0.2g
0.1g

• Increasing hazard levels

GAS NETWORK (EU)
31
Gas source supply stream
Failure of the gas vertex
Gas Power Plant
32
Dependent network analysis
20 PL
19.5 electricity generation capacity from gas
power plants
EUROPE
50 PL
80 PL
Electricity network
Dependant network

Performance measure
POWER LOSS

33
Vulnerability of the Network

• Probability reliability model
• spatial distribution of network
• network response under seismic hazard
• increased vulnerability due to interdependency

475 year
0PL
0.8PGA
100PL
0.5PGA
Distribution substations
34
Vulnerability of the EU Network
Topological vulnerability of the EU electricity
network for events scaled from 0.8 to
2.5 Strength of coupling 100 Averaged value
of Power Loss for each distribution substation
35
Induced Vulnerability
Independent Networks
Gas Dependent Electricity Networks
Gas Dependent Electricity Network
Gas Independent Electricity Network
2.5 PGA
36
Population Affected
Density of Population is an important component
to evaluate the effects of possible events on the
grid
The amount of population affected by disruptions
is not directly proportional to the probability
of failure of the electricity distribution network
37
Conclusions

• GIS proves to be a valuable tool for critical
  infrastructures data processing
• Network Analysis must be involved from the outset
  when dealing with GIS data
• Complex Networks vulnerabilities must be
  considered along with their interactions

38
Thank you
www.manmadenet.eu