Accidents in the Energy Sector and Energy Infrastructure Attacks in the context of Energy Security

Wir schaffen Wissen – heute für morgen

CP Expo, 29 – 31 October 2013, Genova, Italy

Paul Scherrer InstitutPeter Burgherr (PSI) & Jennifer Giroux (ETHZ)Accidents in the energy sector and energy infrastructure attacksin the context of energy security

CP Expo 201329 – 31 October 2013

Burgherr & Giroux: Accidents and Intentional Attacks in the Energy SectorCP Expo, 29 – 31 October 2013, Genova, Italy

• Comparative Risk Assessment in the Energy Sector Energy-related Severe Accident Database (ENSAD) [PSI]

• NaTech Accidents affecting Energy Infrastructures ENSAD [PSI]

• Intentional Attacks on Energy Infrastructures Energy Infrastructure Attack Database (EIAD) [ETHZ-CSS / PSI]

• Terrorism Risk for Energy Installations Probabilistic tool for site-specific assessment [PSI]

• Conclusions

Content


Comparative Risk Assessment in the Energy Sector


• Comparative assessment of accident risks is a central aspect in a comprehensive evaluation of the performance of energy technologies(Fritzsche 1989; Inhaber 2004; Rasmussen 1981)

• In the past 40 years catastrophic accidents affected the entire energy-related business and industry (Sutton, 2012)

• Society is often risk averse towards low-probability high-consequence events, but at the same time a lack of urgency can be observed among the public and decision makers (Garrick, 2008)

• No adequate treatment of energy accidents in terms of completeness and data quality (Fritsche, 1992)

To close this gap, the PSI initiated in the early 1990s a long-term research activity, at the core of which is the

ENergy-related Severe Accident Database (ENSAD)

Background


Overview of Accidents in the Energy Sector

Kocaeli earthquake (Tur),fire at refinery

Lightning struck oil storage tank

Fire/explosion at LNG facility (Algeria)

Prestige, Galicia (Spain)

Deepwater Horizon (USA)

Explosion of tapped gasolinepipeline, Nigeria

Silane explosionIn PV plant

Wind turbine collapse

Coal mine explosion

Refinery fire/explosion

Gas pipeline explosion

LPG explosion

Biodiesel plant explosion

Dam failure

Induced seismicityat geothermal well


Approach for Comparative Risk Assessment

Burg

herr

et al.

2013

• Full energy chains because accidents can occur at all stages

• Evaluation period: ENSAD contains accident data for more than four decades 1970–2008 (2012/13)

• Data normalization to ensure comparison across different energy chains GWeyr

• Regional aggregation at different spatial scales individual countries, country groups (OECD, EU, non-OECD, globally)


Severe Accident Definition and Consequence Indicators

Risk description

Impact Category ENSAD severity threshold

Consequence indicator

Human health FatalitiesInjuries

≥ 5≥ 10

Fatalities per GWeyrInjured per GWeyr

Societal Evacuees Food consumption ban

≥ 200yes

Evacuees per GWeyrNominal scale

Environmental Release of hydrocarbonsLand/water contamination

≥ 10’000 t≥ 25 km2

Tonne per GWeyrkm2 per GWeyr

Economic Economic loss ≥ 5 Mio USD (2000) USD per GWeyr


Historical Development of ENSAD

Burgherr et al. 2013


Geographic Distribution of Severe Accidents (1970-2008)

• Top 10 countries accounted for 83% of all fatalities; countries with 100 to 1000 cumulated fatalities contributed another 15.1%, and remaining countries summed up to only 1.9%.

• China: 53343 fatalities; 25772 in coal mine accidents (mostly Shanxi, Henan, Guizhou, Heilongjiang and Hunan); 26000 in Banqiao/Shimantan dam failure (Henan).

• Nigeria: oil chain accounts for over 98% of fatalities; Delta, Lagos, Rivers, Osun and Abia states.• USA: most fatalities in the Gulf of Mexico area (e.g. LA, TX), where O&G activities are concentrated.

Burgherr et al., 2010

Nigeria


Fatality Rates & Maximum Consequences (1970-2008)



NaTech Accidents affecting Energy Infrastructures


NaTech Accidents BackgroundWhat are NaTech accidents? Accidents triggered by natural hazard

(Showalter & Myers, 1994; Steinberg & Cruz, 2004; Krausmann et al., 2011)

Why look at NaTech accidents? Risk of such accidents is expected to increase in the future due to

- growing industrialization- change of natural hazard occurrence patterns by climate change- increasing vulnerability of the society

(Girgin & Krausmann, 2013)

How can NaTech accidents be analyzed? Qualitative risk screening (Cruz & Okada, 2008) Index method ranking natech hazards (Sabatini et al., 2008) Qualitative screening using Multi-Criteria Decision Model (Busini et al., 2011)

Quantitative assessment (Antonioni et al., 2007) ENSAD


NaTech Accidents in ENSAD

• 398 NaTech accidents in ENSAD (1970-2008), mostly fossil energy chains.• In the coal chain out of 95 accidents 83 occurred in China, and of these 69 were severe(≥5

fatalities)• In the oil and natural gas chains 246 and 37 Natech accidents were recorded, with most of them

located in the USA (115 and 24).• Flooding, windstorm and lightning were the most common triggering events, followed by

earthquake and landslide.

• Highest frequencies in Coal China.• OECD countries exhibit lower

frequencies and smaller maximum consequences than non-OECD.

• However, currently available ENSAD data are rather limited.

• Key research focus in currently ongoing database update.



Intentional Attacks on Energy Infrastructures


Energy Infrastructure Attack Database (EIAD)• Focus on violent non-state actors (VNSA)• Covers all forms of non-state violence aimed at energy infrastructure

- all human (energy sector personnel)- physical (energy sector physical assets) - information (energy sector cyber systems supporting operations)

• EIAD does not code for motivation but rather for - attack type (e.g., assassination, assault, bombing, etc.) - instrument used (e.g., firearms, explosive-dynamite, arson/firebombing, etc.).

• All information in EIAD is from open-source information sources• EIAD incident record structure:

- Unique ID- Date (including extended incidents such as hijacking and kidnappings)- Location (desriptive and geo-coded information)- Information (summary, event type, and whether event was part of a multiple attack)- Attack Information (attack type, instruments used, combination attack, second attack type)- Target Information (specific target, energy sector, energy infrastructure, second target)- Perpetrator Information (individual/group, group type)- Incident Consequences (casualties and fatalities, reported downtime, infrastructure impact, hostage information)- Additional Information - Source Information (media reports, social media, cross-reference to other databases, etc.)


Summary of EIAD Content• Currently EIAD contains 8602 records for the

years 1980-2011.

• Significant increase over time.

• Vast majority of attacks classified as successful (8211)

• Most attacks directed to “linear” targets that are difficult to protect (50% electricity transmission lines and substations, oil pipelines (15%), oil transports by road tanker and natural gas pipelines (each 7%).

• Almost 40% (3413) of EI attacks were considered multiple attacks. Multiplicity within a specific country points to the power of ‘tactical contagion’ and the ‘criminal micro-cycles’ within certain contexts, contributing to the development of spatio-temporal “waves” or hotspots.

Burgherr & Giroux 2013


EI attacks by attack type • 80.4% of attacks were carried out with some

type of bombing device.

• Several other attack types with more than 100 events cumulatively amount to 15.0%.

• Notable that also 4.1% could not be assigned due to incomplete information.



Spatial distribution of EI Attacks by Country


• Top 3 countries are Colombia (1381 EI attacks), Iraq (1085) and Pakistan (1009), totaling 49.7%.• More than 200 attacks: El Salvador, Peru, Afghanistan, Nigeria, Chile (25.7%).• More than 100 attacks: India, Angola, Philippines, Thailand, Russia (11.8%). • More than 50 attacks: Spain, Turkey, Yemen, Guatemala (3.8).• Remaining 69 countries 8.9%; roughly two thirds of them 0.1% or less.


Temporal Patterns in Top 3 Countries



Comparison EI Attacks and Accidents

Burg

herr

& Gi

roux

2013

• Country-specific patterns for severe (≥5 fatalities) and smaller (1-4 fatalities) accidents and EI attacks.

• Maximum consequences of accidental events are generally higher than for EI attacks. • Caution: in the case of Nigeria and Pakistan EIAD data might not be complete for EI attacks ongoing updates.


Terrorism Risk for Energy Installations


Terrorist Threat Analysis (EU Project SECURE)Risk of fatalities through terrorism total

1.E-061.E-051.E-041.E-031.E-021.E-01

1.E+001.E+011.E+02

EPR(Gen III)

LMR(Gen IV)

HTR(Gen IV)

Dam Refinery LNG

Nuclear Hydro Oil Gas

Ris

k pe

r ins

talla

tion

per y

ear

China

USA

Europe

Eckle

et al.

, 201

1

• First-of-its-kind analyses of terrorist threat bymeans of scenario quantification for selectedenergy infrastructure.

• In spite of large uncertainties the analysis indicatesthat the frequency of a successful terrorist attackwith very large consequences is of the same orderof magnitude as can be expected for a disastrousaccident in the respective energy chain.

• This is primarily due to the fact that centralizedlarge energy installations are hard targets andrelatively easy to protect, requiring sophisticatedattack scenarios to cause significant damage andlasting impacts.


Conclusions

• Evaluations are based on quantitative data from the databases ENSAD and EIAD.

• Accidents are typically rare and independent events.

• Intentional attacks are often multiple events and concentrated both in time andspace, resulting in distinct hotspots.

• The severity distribution for accidents generally exhibits a fat-tail, with low-probability high-consequence events being an important factor of energy chainperformance.

• For intentional attacks severe consequences are less an issue because energyinfrastructures are often targeted in remote areas and difficult to protect (e.g.pipelines and transmission lines), but when frequently attacked can result insubstantial business and supply disruptions.

• The joint analysis of accidents and intentional attacks provides a comprehensiveand complementary approach on two types of risks that have rather differentproperties, but are essential in a holistic energy security perspective.

CP Expo, 29 – 31 October 2013, Genova, Italy Burgherr & Giroux: Accidents and Intentional Attacks in the Energy Sector

Thank you for your attention!

Laboratory for Energy Systems Analysishttp://www.psi.ch/[email protected]

Accidents in the Energy Sector and Energy Infrastructure Attacks in the context of Energy Security

Technology

energy sector coal

energy infrastructure

energy sectorcp expo

energy infrastructurescp

overview of accidents

natech accidents background

italyburgherr giroux

years catastrophic accidents