Critical Infrastructure

Energy.pdf

Home >Government homework help >Critical Infrastructure

213

EnErgy

Energy—derived from fossil fuels, wind, solar, and nuclear— propels everything. It runs power plants, heats our homes, powers our cars, and air-conditions our offices. Without energy, the telephone and Internet wouldn’t work, and the modern conveniences we take for granted vanish. The United States consumes roughly 100 quads of energy/year. This is equivalent to 17.5 billion barrels or 735 billion gallons of crude oil. Approximately, 20% of this energy comes from coal, 23% from natural gas (NG), and 35% from petroleum. Figure 12.1 further indicates that 38% is converted into electricity to run computers, homes, offices, and factories. Transportation in the form of cars, trucks, busses, airplanes, ships, and trains consumes 27%. Incredibly, 55% is lost due to conversion, transmission, and nonconservation inefficiencies. For example, an incandescent light bulb based on Edison’s invention wastes 95% of the energy needed to light up a room.

The energy sector is transitioning to renewable fuels such as wind, solar, tidal, and geothermal. This transition will take 40–50 years but radically alter another infrastructure—trans- portation. Automobiles, trucks, trains, and airplanes are likely to be running on electrons, hydrogen, or some fuel other than gasoline by 2030. Conversion of the energy sector from fossil fuels to some alternative such as electricity will most likely bring major sociopolitical and economic shifts with it.

This chapter examines critical nodes of the fossil fuel supply chains1 that deliver much of the energy consumed by the United States. We illustrate how this supply chain works

through case studies: the Powder River Basin coal supply chain, the Gulf of Mexico to Northeastern U.S. supply chain, and the Northeast storage facility located in New Jersey. The following summarizes the result of these analyses:

• Energy versus power: Energy is the ability to do work; power is the rate of doing work. Therefore, power is mea- sured in units like kilowatts (kW) and horsepower (ft-lb/s), and energy is measured in units like kilowatt-hour (kWh). The mathematical relationship is energy = power × time. The energy sector is concerned with extraction and delivery of fuels to power plants, and the power sector is more concerned with producing power from power plants and delivering it to consumers.

• Coal supply chain: This critical supply chain is highly dependent on transportation in the form of rail delivery of coal to power plants. The United States leads the world in coal reserves. The Powder River Basin is the largest source of coal in the United States, and its delivery depends on rail service to power plants. While the United States is highly dependent on coal as a fuel, environ- mental regulations continue to reduce the availability of coal for generation of electrical power. Coal is a declining source of energy for the United States, but not China.

• Gas and oil supply chains: Most U.S. energy comes from vast oil and NG supply chain networks that are highly self-organized—they have high betweenness centrality and low levels of robustness. They are vul- nerable to disruptions of their large and unique trans- mission pipelines that form supply chains from Canada, the Gulf of Mexico, and foreign sources.

1According to Dr. Warren H. Hausman of Stanford University, “The term supply chain refers to the entire network of companies that work together to design, produce, deliver, and service products.” http://www.supplychainon line.com/cgi-local/preview/SCM101/1.html

Lewis, T. G., & Lewis, T. G. (2014). Critical infrastructure protection in homeland security : Defending a networked nation. ProQuest Ebook Central <a onclick=window.open('http://ebookcentral.proquest.com','_blank') href='http://ebookcentral.proquest.com' target='_blank' style='cursor: pointer;'>http://ebookcentral.proquest.com</a> Created from apus on 2020-11-12 09:53:33.

C o p yr

ig h t ©

2 0 1 4 . Jo

h n W

ile y

& S

o n s,

I n co

rp o ra

te d . A

ll ri g h ts

r e se

rv e d .

214 ENERGy

• Critical assets: Major components of oil and NG supply chains are wellheads, refineries, transmission pipelines, storage, distribution pipelines, and SCADA. Supplies are vulnerable to disruption because of highly clustered refineries, pipeline ruptures, and storage facility damage.

• Competitive exclusion: Ownership of gas and oil refin- eries and transmission pipelines is highly concentrated. Dense clusters of refineries, limited transmission links, and concentration of storage terminals characterize energy supply chain networks. Robustness is limited because few assets are redundant.

• Vulnerabilities: Coal is vulnerable to transportation dis- ruptions, and oil and NG supply chains are vulnerable to disruption because of asset clustering and low robust- ness of transmission. They are also heavily dependent on power (for running pumps and SCADA).

12.1 EnErgy FundamEntals

Most people are ignorant of perhaps the most important CIKR sector to modern civilization—the energy sector. Energy has become a commodity like water and the air we breathe, but like water and air, modern civilization would come to a halt in less than a month without a continuous supply of energy. A sudden cessation of gasoline, coal, or electrical power, for example, would throw the world back 500 years to medieval times. And yet, few consumers understand how this sector works. Even fewer understand how energy is measured, produced, delivered, and consumed. Taking this CIKR for granted may be a major mistake, however.

As a simple illustration, consider the architecture of the U.S. energy sector as shown in Figure 12.1. The United States consumes approximately 18 million barrels of oil/day (2012),

Hydro 2.7%

Nuclear 8.4%

Electric generation

38.2% Coal 19.8%

Other 1.5%

Nat. Gas 23.4%

Residential 11.3%

Commercial 8.5%

Industrial 21.8%

Transportation 27.0%

Petroleum 35.3%

Other 8.9%

25.3%

7.8%

7.6%

3.2% 1.7% Lost

54.6%

4.4%

20.2%

2.3%4.9% 4.7%

4.5%

3.0%

.03%

7.0%

1.5%

18.3%

8.4% 26.1%

2.7%

1.4%

FigurE 12.1 Most energy consumed in the United States comes from coal, NG, and petroleum—fossil fuels—and over half is lost due to inefficiency.

C o p yr

ig h t ©

2 0 1 4 . Jo

h n W

ile y

& S

o n s,

I n co

rp o ra

te d . A

ll ri g h ts

r e se

rv e d .

ENERGy FUNDAMENTAlS 215

representing about 27% of all energy consumed. To gauge how large this number is, consider the energy content of one barrel of oil, which contains 42 gallons of crude oil. (This crude is converted into 44 gallons of gasoline, which is roughly equivalent to 1 week of fuel for a typical consumer.) The typical U.S. consumer uses approximately 54 barrels of energy/year.

Energy isn’t created or destroyed, except by nuclear reac- tion. Instead, it is converted. A coal-burning power plant releases energy by burning coal and converting it into heat and motion by heating water to make steam, which drives a rotating turbine. The kinetic energy of the turbine is further converted into moving electrons by a generator. Thus, the energy stored in coal is converted into other forms of energy— heat, kinetic, and electrical—so it can be transmitted to where it is converted back into movement of an electric car, lighting of a house or office, or computation inside of a computer or cell phone.

Energy isn’t created or destroyed, but some of it is wasted in conversion. For example, burning coal is roughly 30% effi- cient, meaning that 70% of coal’s pent-up energy is lost during conversion to heat, light, and chemical reaction. Sadly, most energy is lost due to conversion along the energy supply chain that extends from the coal mine to the home or office.

Energy is measured in terms of work performed over some time period, so energy and work are equivalent. On the other hand, power is work performed per unit of time. It is the rate of doing work. The difference is subtle so here is an example. Mostly everyone in the West is familiar with horsepower. James Watt defined horsepower as the amount of work a horse could do over a certain amount of time. He concluded that one horse could lift 33,000 lb of coal out of a mine in a minute. For example, if Watt’s horse lifted 330 lb of coal 100 ft in a min, or 1,000 lb 33 ft in 1 min, his horse exerted 1 hp, regardless, because 330(100) = 33(1,000) = 33,000 ft-lb/min.

The relationship between power and energy (work) is

Energy power time= ×

Therefore, the amount of work done by Watt’s horse depends on time: if his horse exerts 33,000 ft-lb/min for an hour, 33,0 00(60 min/h) = 1,980,000 ft-lb of work is done. If the horse works for 2 h, 3,960,000 ft-lb of energy is transferred from the horse to the coal (the coal has higher potential energy as a result of being lifted out of the mine).

There are many different measures of energy and power. In the energy sector, most technologists prefer to measure energy in kWh in place of foot-pounds (ft-lb) and power in kW instead of hp. 1 kW equals 1.34 hp, for example. A 100 W light bulb equals 0.134 hp, an electric car with 100 kW motor equals 134 hp, etc.

The amount of work made possible by a fuel depends on its energy density. Table 12.1 lists a few popular fuels and their energy densities for comparison. Electrical energy is

typically measured in kWh—a 100 W incandescent light bulb burning for 10 h converts 1 kWh of energy into heat and light. Modern U.S. urban dwellers convert 500–1000 kWh of electricity into heat, cooling, computation, and communica- tion, for example, per household, per month.

Mechanical energy is typically measured in British thermal units (BTU) instead of kWh. 1 kWh is approxi- mately 3412 BTU. An electric car with a 100 kWh battery stores 341,200 BTU of energy when fully charged. If its efficiency is 3 kWh/min, the electric car can travel 3(100) = 300 miles on a charge. How long would it take? It depends on the power of its electric motor. How big is the battery? It depends on the energy density of the chemicals used to store electrons in the battery.

The United States converts approximately 100 quads/year of energy from all forms of fuel (see Fig. 12.1). A quad is an extremely large number equal to 1015 BTU, 8,007,000,000 gallons of gasoline, or 293,083,000,000 kWh. The United States uses nearly 100 times this amount of energy. The entire planet converted fuel of one kind or another into 446 quads in 2004; therefore, the United States used (95/446) = 21% of all energy captured by the planet in 2004.

Over half of all energy produced in the United States is wasted—largely due to inefficiency. Electrical power trans- mission is the most wasteful followed by gasoline consump- tion. Together, they account for 46% of all energy—losses that escape in the form of heat, friction, etc. This fact alone may drive future policy decisions regarding the CIKR sector. For example, conversion of personal transportation from the internal combustion engine (ICE) to an alternative such as fuel cells or electric motors can reduce the 20% loss to per- haps 5%, because electric motors are three to four times more efficient than ICE. As another example, consider the possi- bility of distributed generation, whereby electric power gen- eration is brought closer to its point of consumption, thereby reducing transmission losses. Solar panels on residential homes and shopping malls, for example, reduce the need for long-haul transmission lines—the major source of electrical energy loss.

These and other transitions brought on by new technologies and new energy policies will radically alter the architecture of

tablE 12.1 Energy density of familiar fuels: uranium-235 is off the scale, while lithium-ion battery storage barely registers. Fossil fuels are comparatively dense sources of energy but also produce large volumes of greenhouse gases such as CO

Fuel Density (kWh/Gal) CO

2 emission

(lb/million BTU)

li-ion battery 3 0 Natural gas 27 117 Gasoline 38 157 Coal 76 216 U-235 1,500,000,000 0

C o p yr

ig h t ©

2 0 1 4 . Jo

h n W

ile y

& S

o n s,

I n co

rp o ra

te d . A

ll ri g h ts

r e se

rv e d .

216 ENERGy

the energy sector. But consumption is unlikely to diminish because energy is the engine of economic prosperity and secu- rity. The confluence of increasing demand and diminishing inefficiencies levels the slope of future energy needs. Figure 12.2 summarizes where we have been and where we are going with respect to this CIKR. Projected energy demand is most likely to track population and economic trends, which are typically 2.5–3.5% per year.

For the past 100 years, civilization has not only increased its demand for energy but increased the rate of increase. Energy demand has accelerated because of greater mobility (transportation), adoption of technology (computers and cell phones), and increased population (from 1.5 to 7 billion in the past century). The supply of energy will hit an inflection point sometime in the twenty-first century, for a number of reasons. First and foremost, fossil fuels are becoming scarce and more expensive to exploit. Second, technology advances are likely to provide economically and environmentally more desirable alternatives such as solar and fuels derived from plants and new processes.

The following historical and contemporary analysis is unlikely to persist beyond the decade because population, technology, and economics will force a radical revamping of policy and strategy in the United States and abroad. Security and safety considerations are also likely to undergo radical modification as the energy sector transforms along one of several possible development directions outlined in the following.

12.2 rEgulatOry struCturE OF thE EnErgy sECtOr

Congress delegates oversight of gas, oil, and electric power to the Federal Energy Regulatory Commission (FERC). The FERC must work with the EPA, the Department of Energy (DOE), and the Office of Pipeline Safety (OPS) positioned within the Office of Pipeline and Hazardous Materials Safety Administration (PHMSA), which in turn is positioned within the Department of Transportation (DOT). The Atomic Energy Commission (AEC) is responsible for oversight of nuclear power plants.

12.2.1 Evolution of Energy regulation

The FERC evolved out of the Federal Power Commission (FPC) established by Congress in 1920 to regulate hydro- electric projects. But the FPC quickly evolved as legislation increased federal control. A brief history as it pertains to the gas and oil industries is summarized below3:

• Regulation of sale and transportation: The Federal Power Act of 1935 and the Natural Gas Act (NGA) of 1938.

• Regulation of NG facilities: 1940 amendments to the NGA and 1954 Phillips Petroleum Co. v. Wisconsin.

• Interstate commerce: In 1967, intrastate utilities became jurisdictional if they connected their supply lines to others outside of the state.

• The FPC becomes the FERC: Congress reorganizes and expands the FPC in 1977 as the FERC.

120

100

Q ua

dr il

li on

B T

0 1980 2008 2020

Year

Coal

Natural gas

Nuclear

Liquids

Biofuels

Energy use by fuel: 1980–2035 ProjectionsHistory

Renewab les

FigurE 12.2 Past, present, and future energy consumption in the United States: fossil fuels will remain the most consumed energy for the foreseeable future2.

2U.S. Energy Information Administration, Annual Energy Review 2008, DOE/EIA-0384(2008) (Washington, DC, June 2009). Projections: AEO2010 National Energy Modeling System, run AEO2010R.D111809A.

3http://www.ferc.gov/students/ferc/history.asp

C o p yr

ig h t ©

2 0 1 4 . Jo

h n W

ile y

& S

o n s,

I n co

rp o ra

te d . A

ll ri g h ts

r e se

rv e d .

REGUlATORy STRUCTURE OF THE ENERGy SECTOR 217

• Unified interstate commerce: The National Energy Act (NEA) of 1978 unifies intra- and interstate gas markets.

• Deregulation: The 1985 FERC Order 436 and 1992 FERC Order 636 opens pipelines to competitors and introduces price controls.

The energy sector has been shaped by a century of regulation and deregulation. It forms an industrial commons balanced between the forces of Gause’s law and competitive exclus- ion and the tragedy of the commons. left to free- market forces, this commons would evolve into a monoculture and monopoly. left to an overregulated command economy, this commons would die from the ravages of the tragedy of the commons. In its early years, the federal government treated this sector like a natural monopoly. Since Congress changed its stance on regulation through the Public Utility Regulatory Policies Act (PURPA) of 1978 and Energy Policy Act (EPACT) of 1992, the energy commons evolved somewhere in between. Pipelines are open to competitors, but there is a relatively high concentration of pipeline own- ership and operation. Energy costs are low for consumers, but government regulates (and limits) profitability of the industry.

12.2.2 Other regulation

The PHMSA regulates the safe, reliable, and environmen- tally sound operation of the nation’s 2.6 million miles of gas and oil pipeline and the nearly 1 million daily shipments of hazardous materials by land, sea, and air. It consists of two offices: the OPS, and the Office of Hazardous Materials Safety (OHMS).

Gas and oil pipeline companies are considered common carriers, which means they must operate as all interstate commerce companies: they must provide nondiscrimina- tory access to their networks. But the energy supply chain overlaps other domains because of its impact on the envi- ronment and dependence on transportation. Operational dependencies lead to complex regulatory controls. The NG and oil supply chains cross many regulatory boundaries as well as geographical boundaries.

The Natural Gas Pipeline Safety Act (NGPSA) of 1968 authorized the DOT to regulate pipeline transportation of various gases, including NG and liquid natural gas (lNG). As a consequence, the DOT created the OPS, but the emphasis was on safety, not regulation. The National Environmental Policy Act (NEPA) of 1969 made the FPC responsible for reporting environmental impacts associated with the construction of interstate NG facilities. And then the task of coordinating federal efforts to cope with electricity shortages was taken from the FPC and given to the Office of Emergency Preparedness in 1970. To complicate matters even more, the FPC was converted into the FERC in 1977, and the NEA of 1978 that includes the PURPA required

gradual deregulation of NG. In 1979, Congress passed the Hazardous liquid Pipeline Safety Act (HlPSA), which authorized the DOT to regulate pipeline transportation of hazardous liquids.

The EPA of 1992 required the FERC to foster competition in the wholesale energy markets through open access to trans- mission facilities. By 1996, the FERC issued a series of orders forcing common carrier companies to carry electricity, NG, or petroleum products from a variety of competing suppliers. This reregulation of the energy sector was euphemistically called “deregulation.”

By 2003, the regulatory structure was cloudy at best. The FERC handled regulation, the DOT/OPS handled pipeline safety, and the EPA handled hazardous materials. The DOE provided information on the energy sector through its Energy Information Administration (EIA at www.eia.doe.gov). HSPD-7, issued by President Bush in December 2002, dis- tributes responsibility for the energy sector across the DOE, DHS, DOT, and EPA. The PHMSA was created under the Mineta Research and Special Programs Improvement Act of 2004, combining the OPS and responsibility for hazardous material spills.

The FERC sets prices and practices of interstate pipeline companies and guarantees equal access to pipes by ship- pers. It also establishes “reasonable rates” for transporta- tion via pipelines. These charges typically add a penny or two to the price of a gallon of gasoline, for example. But the FERC is not responsible for safety, oil spills, or the construction of the energy supply chain. Safety is regulated by the PHMSA, except for nuclear energy, which falls on the DOE and the AEC.

For example, when a pipeline ruptured and spilled 250,000 gallons of gasoline into a creek in Bellingham, Washington, in June 1999, the OPS stepped in. This accident killed three people and injured eight. Several buildings were damaged, and the banks of the creek were destroyed along a 1.5 mile section.

Is there any difference between NG and oil when it comes to regulation? The FERC is empowered to grant permission for anyone to construct and operate interstate pipelines, interstate storage facilities for NG, or liquefied NG plants. It also handles requests by anyone who wants to abandon facil- ities when, for example, pipelines get old and need to be upgraded or replaced with a new pipeline. But the FERC does not regulate local distribution pipeline companies— state public utility commissioners regulate them.

12.2.3 the Electric sector isaC

The mission of the Electric Energy Sector ISAC (ES-ISAC) is to provide threat and warning information to member companies in the electric energy sector.

The list of ISACs is constantly changing. For the most current list, visit http://www.isaccouncil.org/memberisacs.

C o p yr

ig h t ©

2 0 1 4 . Jo

h n W

ile y

& S

o n s,

I n co

rp o ra

te d . A

ll ri g h ts

r e se

rv e d .

218 ENERGy

html. The electric sector ISAC is typical of ISACs. It pro- vides its members with:

• Information on threats and vulnerabilities (physical, cybersecurity, and interdependencies)

• How to respond to threats (both physical and information security tips)

• A forum for members to communicate best practices in a secure environment

12.3 intErdEpEndEnt COal

Coal has been a cheap and plentiful source of energy for thousands of years and will continue to be a major source throughout the world. Its use in the modern age accelerated in the 1880s in concert with the rise of the industrial revolu- tion. Coal powered the rapid rise of steel, rail, transportation, skyscraper, petrochemical, and electric power industries. Its relatively high energy density makes it an attractive source of energy (see Table 12.1). Without coal, the modern age would not be very modern.

Global reserves of oil will last about 50 years at current consumption rates. Coal, on the other hand, will last about 112 years at current consumption rates.4 The United States has the largest reserves of coal (22%), and China has 12%. The United States produced 14% of the global supply in 2011 and consumed 12%. But China produced and con- sumed 50% of all coal produced in 2011. China is currently the largest market for coal.

Coal produces 20% of all power consumed in the United States, but this ratio is changing because of the high CO

emissions caused by burning coal (see Table 12.1). Coal- powered power plants produced 30% of output from all power plants. An EPA proposal in 2013 cuts CO

2 emissions

from power plants by nearly 40% (from 1800 to 1100 lb/ MWh). This regulation poses the largest threat to coal as a source of energy.

12.3.1 interdependency with transportation

Surprisingly, the largest source of coal in the United States is Wyoming, not West Virginia. Two mines—the Black Thunder and North Antelope Rochelle mines—in Wyoming produce almost as much coal as West Virginia. Over half of the coal produced in the United States is produced in the Western coal region, and Wyoming is the largest producer. Nine of the top 10 U.S. mines are located in Wyoming.5

One region in Wyoming is particularly critical to the coal supply chain. Approximately 68% of U.S. coal comes from

the Powder River Basin region. But markets for this coal are spread all over the globe. How does Powder River Basin coal reach these markets? In general, 72% of mined coal reaches the power plants that consume it by rail, 11% by barge, 10% by truck, and 9% by pipeline. Powder River Basin coal depends on rail, because only rail can move such large vol- umes of coal, economically.

The Powder River Basin coal supply depends on Union Pacific (UP) and Burlington Northern Santa Fe (BNSF) to move over 300 million tons of coal to power plants in the Midwest and westward to markets in Asia (see Fig. 12.3).6 These trains are dedicated to the task—they haul 100–125 cars of Powder River Basin coal to electric utilities and then return back to the mines empty. BNSF, for example, supplies 10% of all electrical power by transporting coal to power plants.

A 103 mile section of railway called the Joint Line is the artery through which most Powder River Basin coal reaches the rest of the United States. It is the busiest stretch of rail- road in the world, handling 60 mile-long coal trains a day.7 Moving the same volume of coal by truck—currently the only alternative to rail—is both prohibitively expensive and restricted by available trucks and drivers. Rail is the only practical way to transport this critical resource.

The destruction of an important bridge, like the High Triple Bridge over Antelope Creek, would stop coal trans- port on one of the two primary lines feeding rail hubs for distribution to multiple states. Destruction of one to three similar targets immediately before peak periods of seasonal electricity demand could disable much of the country’s gen- eration capacity for periods of weeks to months.

Powder River Basin coal illustrates one of the most vital interdependencies among CIKR sectors. A normal accident to the railway network could magnify and spread to fuel shortages in the Midwest. These shortages could reduce or halt the generation of electric power to much of the United States. Blackouts or brownouts might then propagate to more populous states like the 2003 Blackout did, causing millions of people to go without power. If the coal shortage lasted long enough, the cascade failure could more broadly damage the economy of the entire country.

12.4 thE risE OF Oil and thE autOmObilE

When Edwin Drake discovered petroleum in Titusville, Pennsylvania, in 1859, the market for oil was to make kero- sene for illuminating homes and offices.8 Pennsylvania soon dominated the kerosene lamp oil business, holding 80%

4http://www.worldcoal.org/coal/where-is-coal-found/ 5http://www.eia.gov/

6https://en.wikipedia.org/wiki/BNSF_Railway 7http://www.sourcewatch.org/index.php?title=Powder_River_Basin#Rail_ history 8http://www.pipeline101.com/History/index.html

C o p yr

ig h t ©

2 0 1 4 . Jo

h n W

ile y

& S

o n s,

I n co

rp o ra

te d . A

ll ri g h ts

r e se

rv e d .

THE RISE OF OIl AND THE AUTOMOBIlE 219

market share. But then two important things happened: (1) Edison perfected the electric light bulb, and (2) Henry Ford revolutionized transportation with the mass-produced gasoline-powered automobile. The Edison electric light bulb was better and cheaper than a kerosene lamp, and the automo- bile was a better form of transportation than the horse and buggy. Both transformed the fledgling oil industry from a niche business to a mainstream consumer products economy. What was not so evident in the 1860s was the extent to which oil would become dependent on transportation and vice versa.

12.4.1 Oil

The oil business quickly became intertwined with transpor- tation, because the well that produced kerosene was in Pennsylvania and the kerosene and gasoline consumers lived miles away in New york. Getting the product from wellhead to market became the tail that wagged the dog. In the early days, crude oil was poured into wooden whiskey barrels and carried by wagon from the wellhead to the train station where it was transported by rail to Northeast refineries. After refining, it was once again distributed by horse-drawn wagons to consumers in New york City and other Northeaster

cities. Moving oil became as profitable as the oil itself—a problem that John D. Rockefeller quickly solved by making an exclusive deal with the railroad company.

The teamsters controlled the supply lines from wellhead to consumer. And like any monopoly, they began charging high prices for moving crude to refinery and refined kero- sene to consumer. The cost to move a whiskey barrel of oil 5 miles to a rail station was greater than the railroads charged to move the same barrel from Pennsylvania to New york City. An alternative to this chokehold had to be found. Thus was born the first pipeline in 1863. Pipelines were soon, and still are today, the most economical way to move petroleum from wellhead to refinery and from refinery to consumer.

The famous Tidewater pipeline, opened in 1879 to bypass even more costly middlemen, was the first trunk line—major energy transmission pipeline. It was half owned by John D. Rockefeller who secured control of the oil supply chain in order to control the oil market. Thus, monopoly power passed from the teamsters to Rockefeller. He rapidly expanded the pipeline network to Buffalo, Cleveland, and New york.

The business model for petroleum and NG was well established when Texas oil was discovered. Extract raw

Powder River Basin

FigurE 12.3 BNSF railway network connects the largest source of coal in the United States to power plants throughout the Midwest and markets abroad (Asia) via the Columbia River Gorge between Oregon and Washington. This is a file from the Wikimedia Commons: https:// en.wikipedia.org/wiki/File:BNSF_Railway_system_map.svg.

C o p yr

ig h t ©

2 0 1 4 . Jo

h n W

ile y

& S

o n s,

I n co

rp o ra

te d . A

ll ri g h ts

r e se

rv e d .

220 ENERGy

crude from the ground, send it to the Northeastern refineries by transmission pipeline, and then distribute the refined product to large metropolitan centers in the East. Thus, a profitable energy business was linked to control of a very long transportation network that often exceeded 10,000 miles in length. The barrier to entry was significant—how many competitors could afford to build such a pipeline? It soon became a monopoly—Standard Oil—run by Rockefeller.

Standard Oil was not the only monopoly held by a single powerful industrialist in the late 1800s. But it was one of the most persistent. The Sherman Antitrust Act was passed in 1890, making certain practices of a monopoly illegal. The Hepburn Act of 1905 declared transborder transmission lines a form of interstate commerce and hence subject to regula- tion by the federal government. But it was not until 1912 that Standard Oil was forced to break up after a lengthy and col- orful struggle involving President Theodore Roosevelt and powerful industrialists of the Gilded Age. In the end, seven regional companies replaced Rockefeller’s Standard Oil Company, as the petroleum industry continued to grow at a rapid pace. After World War I, pipeline networks were serv- ing much of the nation, rising to 115,000 miles in the 1920s. Today, they exceed 200,000 miles.

12.4.2 ng

NG, and its more compact form, lNG, is also transported largely by pipeline. Pipeline middlemen controlled the market and hence the price of getting NG to consumers. So in 1938, Congress passed the NGA, which allowed the FPC (forerunner of the FERC) to set the prices charged by inter- state pipelines—but not the prices charged by producers.9 In 1940, the NGA was amended to add regulation of the NG facilities themselves to FPC’s responsibilities.

In the famous Phillips Decision of 1954, the U.S. Supreme Court determined that the NGA covered wellhead prices as well. The FPC was now responsible for regulating the prices along much of the supply chain. The Court sought to protect consumers from “exploitation at the hands of natural gas companies.”10 The government sought to control the source, refining, and distribution of gas and oil for the benefit of tax- payers and voters.

Oil production in the United States was outstripped by demand during the 1950s and 1960s, which led to an increasing dependence on imported oil. The Colonial Pipeline constructed in 1968 to deliver oil products from the Gulf of Mexico states to the Northeastern United States was the largest privately financed project in history up to that time. The 800 mile trans-Alaskan pipeline delivered 2 million barrels/day when it opened in 1977. It delivers 1 million barrels today.

Increasing demand and decreasing domestic supply pushed importation of oil ever upward. By 2003, the United States was consuming 20 million barrels of oil per day, 68% of it delivered by pipeline networks and 27% by boat. The remainder was moved by truck and rail. Most NG and oil are moved from wellhead to refinery, to distributor, and then to consumer by pipe. This is the most economical way to pro- vide an enormous quantity of product over such long distances to so many consumers. But it is also the source of vulnerabilities, as we shall see later in this chapter.

The following network analysis of one major pipeline system connecting Canadian sources with U.S. refineries and consumers indicates a low spectral radius but a high betweenness centrality similar to pipeline networks in other CIKR. like most flow networks, energy supply chain net- works are characterized by low spectral radius and high betweenness centrality, which make them vulnerable to denial-of-service failures.

12.5 EnErgy supply Chains

Supply chain management (SCM) is the management of all steps in the delivery of a product or service to consumers. The gas and oil industries are principally an SCM commons consisting of exploration, drilling, operation of crude pipe- lines, and operation of refineries for the production of fuels, plastics, etc. Trunk line or “transmission pipes” are the arte- rials that deliver refined products such as gasoline and aviation fuel to terminals located around the country. Distribution refers to the sale and delivery of these products to consumers from storage terminals.

A greatly simplified supply chain model is shown in Figure 12.4. Unrefined product from a wellhead is trans-

Wellhead

Re�nery

Storage

Consumer SCADA

Crude

Transmission

Distribution

FigurE 12.4 Simplified supply chain model of NG and petroleum energy infrastructure.

9There seems to be a pattern here: federal government regulators attempted to put limits on the prices that an electric utility may charge consumers but allowed the wholesale prices of power to float. This backfired in states such as California because of shortages in the capacity of the power grid to distribute power to consumers. In the case of NG, regulation had to be expanded to cover the entire supply chain. Is this the eventual fate of electric power grid operators? 10Phillips Petroleum Co. v. Wisconsin, 1954. This Supreme Court deci- sion resulted in an expansion of the FPC’s jurisdiction, and in the after- math of the decision, NG applications under the NGA exploded, far exceeding the volume of electrical and hydroelectric regulation handled by the FPC.

C o p yr

ig h t ©

2 0 1 4 . Jo

h n W

ile y

& S

o n s,

I n co

rp o ra

te d . A

ll ri g h ts

r e se

rv e d .

ENERGy SUPPly CHAINS 221

ported to a refinery by a pipeline system or boat. The refinery converts the crude into refined products, which in turn are transmitted over long distance to terminals, where they are stored—typically in large tanks. Then a distribution net- work of trains, pipelines, and trucks delivers the product to consumers. Transmission generally refers to the long-haul pipes and distribution to the local short-haul delivery to consumers.

The vast miles of pipeline are monitored by various SCADA systems that report anomalies such as leaks and broken components. An example of a pipeline SCADA network is given in the chapter on SCADA—the crude pipeline that delivers oil from the fields around Bakersfield, California, to the refineries in the los Angeles area. SCADA is an increasingly important part of the supply chain because of environmental and security concerns.

12.5.1 petroleum administration for defense districts

For purposes of tracking supply and demand reporting, the supply chain is divided into five regions called Petroleum Administration for Defense Districts (PADDs) (see Fig. 12.5). They were created during World War II when gasoline was rationed. The division is somewhat artificial today but still used to record in- and outflows of petroleum and petroleum products. Broadly speaking, the five PADDs cover the East Coast (PADD1), the Midwest (PADD2), the

Gulf Coast (PADD3), the Rocky Mountain Region (PADD4), and the West Coast (PADD5). Supply chain data can be obtained from the U.S. DOE’s EIA, which collects and pub- lishes oil supply data by PADD.11

For example, in 2001, the percentage of oil produced and imported by each PADD was as follows.

PADD # Production Importation

1. East Coast ~0% ~100% 2. Midwest 10% 90% 3. Gulf Coast 90% 10% 4. Rocky Mountain ~100% ~0% 5. West Coast 45% 55%

Today, oil flows mainly from Canada or the Gulf Coast to refineries in PADD3, and refined product flows mostly from PADD3 to PADD1 and PADD2 (East Coast and Midwest). The West Coast supply chain is almost a stand-alone network that depends on Canadian, Alaskan, and foreign sources.

12.5.2 refineries

The United States has more refining capacity (20%) than any nation in the world and refines 96% of all petroleum prod- ucts it uses. Refineries are distillation factories. That is, they

AK HI

Los Angeles

San Francisco NV

Seattle MT

Deniver

WI MI

IL IN

AL GA

Miami

SC PADD1C:

Lower Atlantic

PADD1B: Central Atlantic

PADD1A: New England

PADD 1: East

Coast

PADD 3: Gulf Coast

PADD 2: Midwest

PADD 4: Rocky

Mountain

VA WVTN

MD DE

NJ PA

MA Bostan Cleveland

Chicago CT RI

New York

TX LA

Houston

PADD 5: West Coast,

AK, HI

FigurE 12.5 Petroleum Administration for Defense Districts (PADDs) are still used today to track oil production and consumption. http://www.eia.gov/.

11www.eia.doe.gov

C o p yr

ig h t ©

2 0 1 4 . Jo

h n W

ile y

& S

o n s,

I n co

rp o ra

te d . A

ll ri g h ts

r e se

rv e d .

222 ENERGy

take in crude oil and distill it into various petroleum products according to their specific gravity (density). Crude oil is sep- arated into a variety of petroleum products because lighter molecules percolate to the top of the distillery and heavier particles stay near the bottom.

A 42 gallon barrel of crude oil produces 44 gallons of refined product, because of gains in the distillation process. less than one-half is turned into gasoline. Assuming your automobile tank holds 20 gallons, each of the 140 million registered automobiles and trucks in the United States con- sumes 140 million barrels of oil just to “fill ‘er up”! (In 2001, U.S. consumers used 840 million gallons/day. Filling up all registered vehicles consumed at least one-sixth of total consumption for 1 day.)

The products obtained from a barrel of oil are summa- rized below along with the amount of each. Note that one barrel of oil is roughly equal to one tank of gasoline for a typical passenger automobile.

Gallons Product

19.5 Gasoline 9.2 Heating oil 4.1 Jet fuel

11.2 Asphalt, feedstock, lubricants, kerosene, etc. 44 Total

In 2003, there were 152 refineries in 32 states. The top 10 refineries produced almost 20% of the total, and the top 2 (Baytown, TX, and Baton Rouge, lA) produced over 5% of the national supply. Refineries are highly concentrated along the Gulf Coast. In fact, most of the high-volume refineries are clustered along the coastline between Galveston, TX, and Baton Rouge, lA. This high concentration poses a major vul- nerability in the refinery component of the energy supply chain.

12.5.3 transmission

large pumps and compressors move billions of gallons of product along pipelines each year. It is not uncommon for a pumping station to be powered by a 4000 hp diesel or electric motor, for example. While these pumps and compressors are backed up with auxiliary power, it shows how dependent the energy sector is on the power sector. Ironically, without power, energy cannot be moved along its supply chain. And without energy, power cannot be generated.

In addition, operators depend on SCADA to monitor the pipeline and its contents—looking for leaks and other anom- alies. SCADA is less critical, but without it, operators are blind. Writing for the Allegro Energy Group, Cheryl Trench describes how pipeline SCADA works:

Pipeline employees using computers remotely control the pumps and other aspects of pipeline operations. Pipeline con- trol rooms utilize Supervisory Control And Data Acquisition

(SCADA) systems that return real-time information about the rate of flow, the pressure, the speed and other characteristics. Both computers and trained operators evaluate the information continuously. Most pipelines are operated and monitored 365 days a year, 24 hours per day. In addition, instruments return real-time information about certain specifications of the product being shipped—the specific gravity, the flash point and the density, for example—information that are important to product quality maintenance. Oil moves through pipelines at speeds of 3 to 8 miles per hour. Pipeline transport speed is dependent upon the diameter of the pipe, the pressure under which the oil is being transported, and other factors such as the topography of the terrain and the viscosity of the oil being transported. At 3–8 mph it takes 14 to 22 days to move oil from Houston, Texas to New york City. [1]

Figure 12.6 illustrates how petroleum products are “sequenced” in a pipeline. like packets of data on the Internet, multiple segments travel on the same pipe. For example, a segment of kerosene may be transported along with a segment of gasoline. This leads to transmixing—the unintentional mixing of products, which often requires some reprocessing at a terminal. But this is an extremely efficient way to move different products over the same (expensive) network. Allegro Energy Group President Cheryl Trench describes how sequencing works (Colonial Pipeline is the largest oil pipeline in the United States):

Pipeline operators establish the batch schedules well in advance. A shipper desiring to move product from the Gulf Coast to New york Harbor knows months ahead the dates on which Colonial will be injecting heating oil, for instance, into the line from a given location. On a trunk line, a shipper must normally “nominate” volumes—ask for space on the line—on a monthly schedule… As common carriers, oil pipelines cannot refuse space to any shipper that meets their published conditions of service. If shippers nominate more volumes than the line can carry, the pipeline operator allocates space in a non-discriminatory manner, usually on a pro rata basis. This is often referred to in the industry as “apportionment.” (Pages 15–16 in Ref. [1])

12.5.4 transport4

In August 1999, several major pipeline companies formed a joint venture to create Transport4 (T4)—an SCM website dedicated to scheduling product sequences through the

Transmix

Premium gasoline

Regular gasoline

Diesel fuel

FigurE 12.6 Pipelines are “multiplexed” by combining different products on the same pipeline according to their density.

C o p yr

ig h t ©

2 0 1 4 . Jo

h n W

ile y

& S

o n s,

I n co

rp o ra

te d . A

ll ri g h ts

r e se

rv e d .

THE CRITICAl GUlF OF MEXICO ClUSTER 223

major pipelines. Buckeye Pipe line Company, Colonial Pipeline, Explorer Pipeline, and TEPPCO Pipeline opened up T4 to connect any pipeline carrier to any customer. Today, 80% of all petroleum products are scheduled via T4.

Advanced technologies such as horizontal drilling and hydraulic fracturing—fracking—have opened up new oil fields, which in turn have outstripped pipeline reach and capacity. Fracking uses water to break apart rocks to release oil deposits. Horizontal drilling has made the Bakken oil play—an area spanning Alberta, Canada, North Dakota, and parts of Wyoming—the second largest oil field in North America and North Dakota second only to Texas in terms of crude oil production.

The opening of new oil fields has had a dramatic impact on freight rail transportation, in regions where pipelines do not exist to transport the crude to refineries. The slack has been taken up by railroads that transport barrels of oil from the North and West to refineries in the East and South. Newfound sources of oil in Canada and the United States have outpaced building of pipelines. The KeystoneXL pro- posal to enhance pipeline reach and capacity is studied later in this chapter.

12.5.5 storage

Products transmitted through major pipeline systems like the 5,500 mile Transcontinental Pipeline (Transco) that delivers 95 million gallons/day to the East Coast (PADD1) end up in storage farms where millions of gallons of heating oil, aviation fuel, and gasoline are stored prior to being distributed to consumers. These tanks are large, conspicuous, and concen- trated in a few critical places. As a consequence, they con- tribute enormously to supply chain risk. This topic is addressed in more detail later.

12.5.6 ng supply Chains

The NG network architecture is much like the petroleum network architecture. It is characteristically subject to low spectral radius, high betweenness centrality, and geographical concentration of critical assets. In some ways, it is even more critical because NG currently provides 23% of U.S. energy, and the percentage is rising. Thus, its criticality is growing.

The NG supply chain is even more expansive than the petroleum supply chain. It consists of 280,000 miles of transmission pipeline and 1.4 million miles of distribution pipeline. Furthermore, it is expanding at a rate of 14%/ decade. Over the next 20 years, the United States will need 255,000 more miles of pipeline.12 The National Petroleum Council estimates that utilities will spend $5 billion/year to build out this capacity.

12.5.7 sCada

SCADA systems are designed to keep pipeline systems safe. They do this by monitoring pumps, valves, pressure, density, and temperature of the contents of the pipeline. An alarm sounds when one or more of the measurements go out of bounds, so operators can shut down pumps and compressors. Operators must consider the local terrain, the product that is inside a pipe, and numerous physical characteristics of the pipeline. If safety limits are exceeded, an SCADA system can automatically shut down a pipeline within minutes.

As we shall see, SCADA plays a relatively minor role in vulnerability analysis because an SCADA shutdown may cause loss of revenue, but not loss of life. Also, SCADA vulnerabilities can be mitigated for a relatively modest investment compared with the investment required to replace a refinery, storage tank, or section of pipeline.

12.6 thE CritiCal gulF OF mExiCO ClustEr

Refineries, major transmission pipelines, and major storage facilities are the most critical assets in the energy sector, because of their large capacities and geographical concen- tration. In addition, many of these critical components are wide open—they are easily accessed and therefore at risk due to symmetric and asymmetric attacks, weather-related damage, and industrial accidents.

Gas and oil supply chains are rich targets for many haz- ards from nature and humans. The most obvious hazards are weather related, especially along the hurricane-battered Gulf Coast states. This analysis is limited to the most frequent threat–asset pairs:

Major refineries along the Gulf Coast Refinery–equipment failure (ref–equipment) Refinery–human error (ref–human) Refinery–miscellaneous accidents (ref–misc)

Major transmission pipelines serving the Northeast market Transmission–equipment failure (trans–equipment) Transmission–corrosion (trans–corrosion) Transmission–aging (trans–age)

Major storage facilities serving the Northeast market Storage–lightning/static electricity (store–lightning) Storage–operational accidents (store–operation) Storage–leaks and ruptures (store–leak)

Figure 12.7 shows the two major networks containing critical assets. The Gulf of Mexico oil fields where explora- tion and drilling occur and the refineries that turn crude oil into refined petroleum products feed petroleum products into the massive Colonial Pipeline network running north

12National Petroleum Council: Meeting the Challenge of the Nation’s Growing Natural Gas Demand, December 1999.

C o p yr

ig h t ©

2 0 1 4 . Jo

h n W

ile y

& S

o n s,

I n co

rp o ra

te d . A

ll ri g h ts

r e se

rv e d .

224 ENERGy

from Houma, louisiana, to linden Station, New Jersey. The (also massive) storage complex at linden Station stockpiles gasoline and commercial airliner fuel until it is needed. Disruption of any link in this supply chain can have severe consequences, because it supplies most of the energy consumed by the major metropolitan areas in the Northeast.

12.6.1 refineries

The Gulf Coast area south of Houston, TX, stretching east to lake Charles, lA, and then to Baton Rouge, lA, con- tains 5 of the top 10 refineries in the nation. These 10 produce nearly 20% of all refined petroleum products for the nation:

1. Baytown, TX

2. Baton Rouge, lA

3. Texas City, TX

4. Whiting, IN

5. Beaumont, TX

6. Deer Park, TX

7. Philadelphia, PA

8. Pascagoula, MS

9. lake Charles, lA

10. Wood River, Il

The five largest refineries located along the critical Gulf Coast region produce 11% of the total U.S. supply. Table 12.2 lists the top five and how many barrels they produce each day. The largest refinery in the United States belongs to ExxonMobil—located in Baytown, TX (in the Galveston Bay south of Houston), and surrounded by smaller refin- eries, as well. It produces aviation fuels, lubricants, base stocks, chemicals, and marine fuels and lubricants.

In 1900, Galveston was the site of the deadliest hurricane in U.S. history. Over 8000 people lost their life. Then in 1947, the largest port disaster in U.S. history wiped out Texas City. A ship containing fertilizer caught on fire and spread through- out the port and much of the town. Over the years, the worst refinery disasters have occurred in large clusters like Texas City: Whiting, Indiana; Texas City, Pasadena, and Amarillo, Texas; Baton Rouge, louisiana; Romeoville, Illinois; and Avon and Torrance, California. The bottom line is that con- centrations such as the Galveston Bay cluster are both vulner- able to damage and highly critical to the oil supply chain.

Refineries can be shut down because of fires, lack of power, or weather-related damage. Power outages may last for only a few hours. Destruction of crude oil pipelines can deny service to a refinery for perhaps days. An explosion or fire can cause longer-term damages—perhaps for months. Hurricanes can flood pipelines and refineries, rendering them unproductive. The cost and likelihood of each incident vary and may be dif- ficult to estimate when performing a risk analysis.

FigurE 12.7 Refined petroleum products flow from the Gulf of Mexico oil field and lOOP terminal to refineries along the Gulf Coast and then to consumers in the Northeast by way of the Colonial Pipeline and storage network. (a) The Gulf of Mexico network of oil fields and refineries.

(a)

C o p yr

ig h t ©

2 0 1 4 . Jo

h n W

ile y

& S

o n s,

I n co

rp o ra

te d . A

ll ri g h ts

r e se

rv e d .

THE CRITICAl GUlF OF MEXICO ClUSTER 225

(b)

FigurE 12.7 (Continued) (b) Colonial Pipeline connecting refineries to markets in the Northeast13.

tablE 12.2 rank, name, and location of the most productive refineries in the gulf of mexico region that produce 11% of national refined product (2.2 of 20 mmbl/day)

Rank# 1 2 3 5 9

Corporation EXXON MOBIl EXXON MOBIl BPPlC EXXON MOBIl PDV AMERICA refiner ExxonMobil ExxonMobil BP Products ExxonMobil CITGO Petroleum location Baytown, TX Baton Rouge, lA Texas City, TX Beaumont, TX lake Charles, lA # of barrels/day 523,000 491,500 437,000 348,500 324,300 market (%) 11.0

13http://subseaworldnews.com/wp-content/uploads/2012/01/USA- Discovery-to-Expand-Pipeline-System-in-Deepwater-Gulf-of-Mexico.jpg

C o p yr

ig h t ©

2 0 1 4 . Jo

h n W

ile y

& S

o n s,

I n co

rp o ra

te d . A

ll ri g h ts

r e se

rv e d .

226 ENERGy

Figure 12.8 lists the most frequent causes of refinery accidents and shutdowns. The top three are related to equip- ment failures (25%), human error (24%), and miscellaneous (22%) incidents such as fire, falls, hazardous material spills, and electrical malfunctions.14 For example, 17 refineries in louisiana reported 301 accidents in 2011. Refinery acci- dents can cost hundreds of millions of dollars.

The cost of closing a refinery depends on the volume of output (lost revenues) and the size of the refinery. Refinery replacement can cost more than $1 billion, and the loss of production (500,000 barrels/day) can have severe implica- tions on revenues as well as shortages that lead to price increases at the gasoline station. The economic impact of faults in this supply chain is inestimable.

12.6.2 transmission pipelines

According to the PHMSA, there is over 2.3 million miles of transmission and distribution pipelines in the United States carrying NG, petroleum and refined petroleum products, as well as chemicals and hydrogen. Crude oil pipes transport unrefined oil from wellhead or ship to refinery. Product pipe- lines move refined products from refinery to storage facilities at the head end of distribution networks. Distribution networks complete the delivery of refined products to consumers.

The largest gas and oil transmission pipelines are owned and operated by conglomerates such as Kinder Morgan, Colonial Pipeline, Transco Pipeline, and Keystone. For example, the 5500 mile-long Colonial Pipeline transmits 95 MMbl/day of petroleum products to customers in PADD1, eventually ending up in Perth Amboy, New Jersey, where it is stored before being distributed by Buckeye Pipeline. It delivers gasoline, kerosene, home heating oil, diesel fuels, and national defense fuels to shipper terminals in 12 states and the District of Columbia. It transports 20% of the entire national supply of refined petroleum products.

Colonial is owned by a joint venture among several major energy companies.

Koch Capital Investments Co. 25.27% HUTTS llC 23.44% Shell Pipeline Co. lP 16.12% CITGO Pipeline Investment Co. 15.8% Phillips Petroleum International Investment Co. 8.02% Conoco Pipe line Co. 8.53% Marathon Oil Co. 2.82%

Figure 12.9 shows the top hazards affecting both gas and oil pipelines. Equipment failures, corrosion, operational accidents, aging, and miscellaneous accidents are the top causes of pipeline spillage. The largest spills can cost upward of $85 million, but most are much smaller. The fractal dimension of consequences for gas pipeline accidents is approximately 1.0, which means risk is at its tipping point between low and high risks. The fractal dimension of conse- quences for oil pipelines is approximately 0.50, which means oil pipelines are high-risk networks. Fortunately, fatalities and injuries from pipeline accidents have been declining since 1970.

12.6.3 storage

Storage tanks hold lNG and refined petroleum products while they are waiting to be distributed through a network of jobbers and resellers. For example, Cushing, OK, is the site of the largest crude oil storage facility in the world, with a capacity of 61 million barrels. The U.S. DOE Strategic Petroleum Reserve (SPR) can hold up to 727 million barrels in underground caverns spread around the Southeastern United States. The nation’s largest NG storage site is an underground cavern near Cedar Creek Field, Montana, with a capacity of 287 billion cubic feet. linden Station, located at Perth Amboy, NJ, is the largest refined product storage facility in the Northeastern United States and the terminal point for the Gulf of Mexico supply chain.

Re�nery hazards: 2005–2011

H um

an er

ro r

Eq ui

pm en

M isc

el la

ne ou

Pi pi

W ea

th er

Co rro

sio n

Po w

er o

ut ag

Hazard

30%

25%

20%

15%

10%

0% F

re qu

en cy

FigurE 12.8 Equipment failure and human error are the top refinery hazards.

14www.SouthernStudies.org data for 17 refineries in louisiana during 2005–2011.

C o p yr

ig h t ©

2 0 1 4 . Jo

h n W

ile y

& S

o n s,

I n co

rp o ra

te d . A

ll ri g h ts

r e se

rv e d .

THE CRITICAl GUlF OF MEXICO ClUSTER 227

In 1996, Colonial Pipeline Co. delivered 820.1 million gallons of jet fuel directly to airports in PADD1: Dulles International Airport, Baltimore–Washington International Airport, Nashville Metropolitan Airport, Charlotte Douglas International Airport, Raleigh–Durham International Airport, Greensboro Triad International Airport, and Hartsfield– Jackson Atlanta International Airport. Thus, the transporta- tion sector depends on this CIKR.

Colonial and other common carriers terminate at Perth Amboy, New Jersey, in the New york Harbor. This storage terminal contains a cluster of storage tanks that are obvious supply chain vulnerabilities. While this analysis is focused on the linden Station, note the numerous other storage facil- ities in Figure 12.10. Hundreds of storage tanks are located in close proximity to one another providing a high-value target to terrorists and criminals.

On January 2, 1990, a ruptured pipeline at linden Station spilled an estimated 567,000 gallons of fuel oil into the

Arthur Kill waterway between New Jersey and Staten Island. The spill caused extensive environmental damage and economic consequences. Pumping continued for 9 h after the rupture, because the owner/operator (Exxon) was slow to detect the spill—operators had disabled the leak detection system. A Coast Guard team in a small boat saw oil bubbling to the surface and determined that the pipeline was the source after operators conducted a pressure test by pumping more oil into the water.

In 1996, storage tank faults were the second most significant cause of spillage after pipelines.[2] Generally, the most frequent hazards affecting storage tanks are lightning strikes and static electricity (38%), operational/ reaction accidents (27%), and leaks/ruptures (13%) [2] (see Fig. 12.11). A ranked exceedence probability distri- bution study of the largest storage faults between 1963 and 2002 produced a fractal dimension of 1.14. This short-tailed exceedence probability suggests that storage

FigurE 12.9 Equipment failure, corrosion, operational accidents, and aging are the top gas and oil pipeline hazards. (a) The top oil pipeline hazards are equipment failure, corrosion, and miscellaneous accidents. (b) The top gas pipeline hazards are corrosion and operational accidents.

30%

35%

25%

20%

15% F

re qu

en cy

10%

Oil pipeline hazards: 2002–2012 (a)

H um

an er

ro r

Co rro

sio n

M isc

el la

ne ou

H um

an er

ro r

Ex ca

va tio

O ut

sid e f

or ce

N at

ur e

Hazard

25%

(b)

20%

15%

F re

qu en

cy %

10%

Co rro

sio n

A ge

O pe

ra tio

Hazard

Gas pipeline hazards: 2002–2009

N at

ur e

Ex ca

va tio

M isc

el la

ne ou

H um

an er

ro r

C o p yr

ig h t ©

2 0 1 4 . Jo

h n W

ile y

& S

o n s,

I n co

rp o ra

te d . A

ll ri g h ts

r e se

rv e d .

228 ENERGy

FigurE 12.11 lightning/static electricity, operational/reaction accidents, and leaks/ruptures are the top storage hazards.

30%

40%

20%

F re

qu en

cy %

10%

Storage hazards: 1963–2002

Li gh

tn in

g/ sta

tic

O pe

ra tio

n/ re

ac tio

Le ak

s/r up

tu re

Eq ui

pm en

Sa bo

ta ge

O pe

n �a

m e

N at

ur e

Hazard

FigurE 12.10 linden Station is the focal point of the massive storage facility at the end of the Colonial Pipeline supply chain. The major storage facilities and pipelines are shown here as a network straddling the New Jersey Turnpike.

C o p yr

ig h t ©

2 0 1 4 . Jo

h n W

ile y

& S

o n s,

I n co

rp o ra

te d . A

ll ri g h ts

r e se

rv e d .

THREAT ANAlySIS OF THE GUlF OF MEXICO SUPPly CHAIN 229

failures are low risk. However, the consequences of the largest failures can exceed $250 million. Fortunately, they are very rare.

12.7 thrEat analysis OF thE gulF OF mExiCO supply Chain

Figure 12.12a contains a general fault tree for the energy sector and its major components at risk—refineries, trans- mission, and storage. This fault tree is applied to the Gulf of Mexico supply chain shown in Figure 12.7. Hypothetical values are used to protect the supply chain’s security, but it is obvious from the foregoing analysis that it will fail if one or

more refinery, transmission pipeline, or storage facility fails. There is little to no redundancy in this CIKR network.

The three major asset types are represented in Figure 12.12a by three components. The refinery component faces three threats—equipment failure (ref–equipment), human error (ref–human), and miscellaneous accidents (ref–misc). The transmission component faces three threats—equipment failure (trans–equipment), corrosion (trans–corrosion), and aging (trans–age). Finally, the storage component’s three most likely hazards are lightning/static electricity (store– lightning), operational accidents (store–operation), and leaks/ ruptures (store–leak). These threat–asset pairs are represented in Figure 12.12a along with probabilities obtained from data presented in Figures 12.9, 12.10, and 12.11.

FigurE 12.12 General fault tree model of probable threat–asset pairs in the energy supply chain for three major components: refineries, pipelines, and storage facilities. (i) General fault tree for the energy sector contains threat–asset pairs for critical assets in the supply chain.

Energy

(a)

Transmission

Re�nery

Storage OR

OR OR

Miscellaneou...

Human error

Equipment

Leaks

Operation

Lightning

Age

Corrosion

C o p yr

ig h t ©

2 0 1 4 . Jo

h n W

ile y

& S

o n s,

I n co

rp o ra

te d . A

ll ri g h ts

r e se

rv e d .

230 ENERGy

For simplicity, consequences from all hazards are assumed to be $2000 million, and all threats are set to 100%. Thus, initial vulnerability and elimination costs are the only vari- ables in Table 12.3. Initial risk is $4484 million, and initial vulnerability (probability of a supply chain failure) is 93%.

Figure 12.12b suggests that fault tree vulnerability is more difficult to reduce than risk. Once again, this is due to the OR gate logic in the model. If any one or combina tion of threat– asset pairs fails, the entire supply chain fails. Risk falls below 50% after approximately $300 million is invested, so this number will be used in the following analysis.

If the risk ranking investment strategy is used to allocate $300 million, the threat–asset pairs would be funded in order: lightning–store (R = $760 million), equipment–trans (R = $638 million), and operation–store (R = $540 million). However, this is nonoptimal, because of the differences in initial vulnerability and elimination costs. An optimal allocation strategy reduces the vulnerability of threat–asset pairs in different order: equipment–trans (allocation = $61 million), lightning–store (allocation = $53 million), and operation–store (allocation = $50 million). This illustrates an important difference in strategies. Is the objective to

reduce the worst-case risk or overall risk? Risk minimiza- tion reduces overall risk.

Optimal risk reduction can be a harsh strategy. For example, two threat–asset pairs are denied any investment at all: leak– store and human–refinery pairs receive zero funding. Why? These two have the highest elimination cost to consequence ratios: leak–store = 0.50, and human–ref = 0.20. All other threat–asset pairs range from 0.13 down to 0.03. Risk minimi- zation ranks threat–asset pairs according to the return on investment. lower ratios mean risk reduction is less expensive and therefore preferred over more expensive risk reductions.

12.8 nEtwOrk analysis OF thE gulF OF mExiCO supply Chain

PADD3 (Gulf of Mexico) oil fields form a major network of refineries, pipelines, and a major import port called louisiana Offshore Oil Port (lOOP). In the foregoing introduction, refineries were shown to be critical, because the nation’s largest are nodes in this network. In addition, lOOP accounts for approximately 13% of the nation’s total import of crude.

tablE 12.3 allocation of $300 million reduces energy fault tree risk from $4484 million (100%) to $2073 million (46%) for hypothetical input data

Name Threat (%) Vulnerability (%) Elimination cost

$(millions) Consequence $(millions) Risk initial

Allocation $(millions)

Vulnerability reduced (%) Risk reduced

lightning 100.00 38.00 100.00 2000.00 760.00 53.23 5.48 109.63 Operation 100.00 27.00 150.00 2000.00 540.00 50.46 8.91 178.18 leaks 100.00 13.00 1000.00 2000.00 260.00 0.00 13.00 260.00 Equipment 100.00 31.90 200.00 2000.00 638.00 61.04 11.09 221.74 Corrosion 100.00 25.00 250.00 2000.00 500.00 40.78 14.79 295.76 Equipment 100.00 25.00 300.00 2000.00 500.00 28.19 18.47 369.49 Human error 100.00 24.00 50.00 2000.00 480.00 32.20 3.10 62.02 Miscellaneous 100.00 22.00 400.00 2000.00 440.00 0.00 22.00 440.00 Age 100.00 18.30 100.00 2000.00 366.00 34.10 6.79 135.84

100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0%

$– $200$100 $400$300 $600$500 $700 $800 $900 $1,000

$Investment

Energy general fault tree: Risk, Vulnerability vs. Investment (b)

%Risk

%Vulnerability

% R

is k,

% V

ul ne

ra bi

li ty

FigurE 12.12 (Continued) (ii) Risk and vulnerability versus investment for hypothetical values plugged into the general fault tree show that vulnerability is more difficult to reduce than risk.

C o p yr

ig h t ©

2 0 1 4 . Jo

h n W

ile y

& S

o n s,

I n co

rp o ra

te d . A

ll ri g h ts

r e se

rv e d .

NETWORK ANAlySIS OF THE GUlF OF MEXICO SUPPly CHAIN 231

Taken all together, the PADD3 network in Figure 12.7 is vital to the energy sector.

Network analysis of the Gulf oil field network shows that it is self-organized around both hubs and high- betweenness nodes and links. Its fundamental resilience equation from Chapter 4 indicates high resilience against cascade failures but low robustness of pipeline links. The fundamental resilience equation supports the claim of high cascade resilience:

log . .

. . %

q( ) = − ∴ = =

0 70 0 38

0 448 44 80

γρ γ

Deeper analysis of the Gulf oil field network shown in Figure 12.7 identifies its critical nodes and links, assuming betweenness and degree centrality are the most important properties. This network has 106 nodes, 121 links, and a mean degree of 2.28 links/node. Its spectral radius is 4.11, or 1.8 times mean degree. This modest amount of self- organization means cascade failures are less important than flow failures. When degree and betweenness centrality are combined, self-organization becomes apparent, because of relatively high betweenness centrality.

link robustness is very low at (121–106)/121 = 12%, but node robustness is much better at (1–1/4.11) = 75.6%. Node robustness is high because many terminal (wellhead) nodes lie under the ocean in the Gulf of Mexico. There are 42

blocking nodes (40%) that hold this portion of the Gulf of Mexico oil supply chain together. They lie on a path established by nodes and links connecting the Gulf oil fields and refineries to linden Station.

This critical path contains critical links and nodes (SS, ST, MC, and JE nodes are under the water):

Critical nodes: Houma, Gibson, SS-28, ST-300, and Clovelly

Critical links: SS-28 → Gibson, MC-311 → Nairn, Erath → Gibson, West

Columbia → East Houston, and JE → Gibson

A subset of these critical nodes and links form a critical path leading to Capline and the Colonial Pipeline (see Fig. 12.13). A fault in any one of these nodes or links leads to a denial of oil to the Colonial Pipeline and there- fore to the markets in the Northeastern United States. Summarizing, the critical nodes and links along this critical path are:

Critical nodes along critical path: Houma, Gibson, Clovelly, St. James, and Capline

Critical links along critical path: Houma → Gibson, SS-28 → Gibson, SS-332 → Houma,

JE → Houma, and Erath → Gibson

FigurE 12.13 The core of the Gulf of Mexico oil field network is centered on Houma and Gibson in louisiana.

C o p yr

ig h t ©

2 0 1 4 . Jo

h n W

ile y

& S

o n s,

I n co

rp o ra

te d . A

ll ri g h ts

r e se

rv e d .

232 ENERGy

Simulation of flow through this core network produces a high-risk flow exceedence probability distribution with a fractal dimension of approximately 0.37. Therefore, this net- work is extremely fragile. Failure in one critical node or link can reduce output as much as 60%, assuming the hypothet- ical values used by the simulation. This, of course, is due to the very high betweenness centrality and low link robust- ness, which leads to a very critical path from wellhead to refinery to storage tank.

To harden the entire Gulf oil field network, 40% of all nodes would have to be 100% hardened. This may be prohib- itively expensive and practically impossible. Without a redun- dant network of pipelines and refineries, the Gulf of Mexico supply chain will remain vulnerable to natural or human-made disruptions.

12.9 thE kEystOnExl pipElinE COntrOvErsy

The Cushing Oil Trading Hub (COTH) in Cushing, Oklahoma, is a dramatic example of Gause’s law. Oil entre- preneurs rushed to the Glenpool Basin area 15 miles south of Tulsa following discovery in 1905 of a major oil gusher. Oklahoma quickly became the nation’s largest oil producer as other fields were discovered and exploited. By 1914, the Cushing field was producing 50,000 barrels/day, or one- quarter of the entire state’s production. Cushing became a center for exploration and production of nearby oil fields. In 1928, the Oklahoma City Field was discovered and soon became the nation’s largest oil producing basin. Major oil companies followed, including Sinclair Oil Corporation, Marland Oil Company (merged with Conoco in 1929), Cities Service Oil Company (CITGO), Phillips Petroleum Co., American Association of Petroleum Geologists, Halliburton Company, Noble Corporation, Anderson & Kerr Drilling Company (Kerr-McGee, purchased by Anadarko Petroleum in 2005), and others.

As oil fields began to run dry in the 1940s, production became less important and brokering became more impor- tant. The COTH became the official price settlement point for the West Texas Intermediate (WTI) crude—the pricing benchmark for North American crude. It evolved into a storage depot and an oil trading center handling crude oil from Gulf of Mexico imports and domestic sources to Midwest refining markets. As supplies from domestic sources declined, the crude oil pipeline system was reversed to transmit crude oil produced in Alberta, Canada, to the COTH where it could be distributed to both Midwest and Gulf Coast refineries. More recently, other crude pipelines reversed flow direction due to increasing domestic supplies from shale reserves in North Dakota and Texas, bringing even more crude oil through the COTH. The COTH has become the nation’s hub for crude oil.

The COTH receives imported Canadian crude via the Keystone Pipeline (590,000 barrels/day) for distribution to refineries. A proposed KeystoneXl pipeline from Canada to the COTH and then to Port Arthur, Texas, will increase the Keystone Pipeline System capacity to 1.3 million barrels/ day. The Obama administration denied construction of the pipeline on environmental concerns, but a study done by larrañaga et al. concluded that the addition of KeystoneXl increases resilience of this supply chain by 55% [3]. The KeystoneXl analysis is reported in Chapter 4.

The KeystoneXl case illustrates the dramatic impact that policy and regulation have on CIKR. In fact, regulation is the major factor shaping most interstate infrastructure. It is responsible, along with economic concerns, for creating hubs, betweeners, and dangerous concentrations of critical assets. Self-organized criticality of the energy sector is mainly due to economic and political factors.

12.10 thE ng supply Chain

The NG (methane) supply chain is very similar to the oil supply chain. NG is gathered from crude oil wells; separated from water, oil, and other contaminants; and pumped into NG pipelines. It is compressed and pumped along vast pipeline networks at roughly 30 miles/h, typically traveling 10,000 miles or more in a matter of days before reaching storage facilities near consumer markets. It is compactly stored in liquid form (lNG) by cooling it to −260° and then thawed out before distributing to consumers. large-capacity lNG ships also take advantage of the compactness of gas while in transit between producers and consumers.

SCADA plays an important role in the safe transmission of NG by regulating the pressure that moves the gas along miles of pipeline. Compressor stations are located about every 50–100 miles to regulate speed and pressure. SCADA provides the necessary control information to regulate supplies as consumer demand varies. By compressing the gas, output can be sped up or slowed down.

The NG supply chain also mimics the oil supply chain in terms of vulnerabilities and risk. Clustering of pipelines in the same geographical location reduces operational and supply costs, but it also increases the risk of attack or acci- dental damage because of clustering, lack of redundancy, and high betweenness centrality in transmission networks. In addition, environmental regulations and Not In My Backard (NIMBy) play predominant roles in self- organization of NG networks and storage clusters.

Most NG comes from the Gulf of Mexico coast and Canada and heads toward the East Coast where the large metropolitan populations consume large quantities of NG to heat homes (80%) and generate electrical power (20%). The largest of these transmission networks is the Transcontinental Pipeline, also known as Transco. Transco runs from Houston to the New york Harbor terminal (see Fig. 12.14).

C o p yr

ig h t ©

2 0 1 4 . Jo

h n W

ile y

& S

o n s,

I n co

rp o ra

te d . A

ll ri g h ts

r e se

rv e d .

THE NG SUPPly CHAIN 233

Table 12.4 lists the largest NG pipeline networks along with their capacities and pipeline lengths. All of the pipelines are over 10,000 miles long, and taken together, they account for 35% of all NG in the United States. The Transco Pipeline pumps nearly 10 billion cubic feet (Bcf)/day, and the longest, Tennessee Gas Pipeline (14,700 miles), pumps nearly as much. Together, these seven networks account for 79,500 of the 212,000 miles (38%) of pipeline and collectively pump 46.8 MMbl/day of the nation’s 133 MMbl/day consumption.

Three networks listed in Table 12.4 are of particular interest because they provide 16% of the national supply and nearly all of the NG energy consumed by the populous PADD1 region (Eastern United States). Number 1 Transco, number 2 Columbia, and number 6 Dominion form a critical

node near Cove Point, Maryland. The Cove Point Intersection illustrates a typical risk in energy supply chains.

West of Washington, DC, across the Potomac River in the Virginia counties of Fairfax, Prince William, and loudoun lies the nexus of three major pipelines: Transco, Columbia, and Dominion (aka CNG). Dominion runs through leesburg Station; Transco runs through loudoun Station, Nokesville, and Dranesville. Pleasant Valley forms a network node through which 16% of the nation’s supply of NG flows. This node con- nects lNG pipelines from Dominion, Transco, and Columbia lines. Disruption in this geographic region could interrupt 20,000 MMbl/day of lNG and NG on its way to New york and points north. Cove Point Intersection is a critical node in the NG supply chain because of its potentially high consequence.

FigurE 12.14 The 10,600 mile Transco Pipeline carries liquid natural gas (lNG) from fields along the Gulf of Mexico and ports along the East Coast to markets in the Northeast.

tablE 12.4 the largest ng pipelines are more than 10,000 miles long and move over one-third of all ng in the nation

Rank 1 2 3 4 5 6 7

name Transcontinental Columbia Tennessee ANR Texas Eastern Dominion El Paso Owner Williams NiSource El Paso El Paso Duke Dominion El Paso Capacity (mmbl/day) 7,362 7,276 7,271 6,667 6,438 6,275 4,882 length (miles) 10,636 11,215 14,761 10,600 12,118 10,000 10,200

C o p yr

ig h t ©

2 0 1 4 . Jo

h n W

ile y

& S

o n s,

I n co

rp o ra

te d . A

ll ri g h ts

r e se

rv e d .

234 ENERGy

12.11 analysis

The energy supply chain is vast and complex. Coal resources are concentrated in just a few geographical locations in the United States and distributed through massive railway networks throughout the country. While coal is a cheap and plentiful fuel, its use is being restricted by environmental rules that limit CO

2 emissions. Regulation is the biggest threat

to this asset and will limit its future use in power generation. The other major sources of energy—gas and oil—depend

on critical supply chains that run all the way from domestic wells in the Gulf of Mexico and North Dakota and foreign wells in Canada and Saudi Arabia to the homes and offices of every American. These vast oil and NG networks are char- acterized by geographical clustering of refineries, single point of failure pipelines, and extremely large-capacity and concentrated storage terminals. These targets are extremely attractive because they are also extremely consequential.

Oil and NG supply chains are regulated much like the electric power grid—and are also undergoing radical transformation due to deregulation of their industries and environmental regulation. Existing pipelines are long and capable of hauling enormous quantities of energy. But they are subject to the competitive exclusion principle, which means the entire nation depends on only a few hub carriers like Colonial Pipeline, Transcontinental Pipeline, Dominion Pipeline, and Colombia Pipeline to carry most of the supply. The uniqueness of these common carriers translates into high-consequence targets for both accidents and terrorists. We have no choice but to protect these assets, because building redundant capacity is not economically feasible. Thus, oil and NG supply chains will continue to be highly structured, low-link-redundancy, high-risk networks.

Energy supply chains are not likely to be enhanced in any significant way for decades. According to the 2001 National Energy Policy:

There are over two million miles of oil pipelines in the United States and they are the principal mode for transporting oil and petroleum products. Virtually all natural gas in the United States is moved via pipeline. Pipelines are less flexible than other forms of transport, because they are fixed assets that cannot easily be adjusted to changes in supply and demand. Once built, they are an efficient way to move products. A modest sized pipeline carries the equivalent of 720 tanker truckloads a day—the equivalent of a truckload leaving every two minutes, 24 hours a day, 7 days a week.15

The topology of this CIKR sector has evolved for over a century. It is unlikely to restructure in less than another century. Even so, environmental regulation and growing interest in renewable sources of energy will gradually reshape this sector over the next 100 years. With so much at

stake, the only feasible strategy is to protect what we have and respond to accidents and attacks on critical nodes and links while transitioning to next-century supply chains.

12.12 ExErCisEs

1. Which of the following is a measure of energy? a. Horsepower b. kWh c. kW d. Watts e. Gallons

2. Most energy consumed in the United States comes from: Natural gas a. Petroleum b. Coal c. Hydroelectric d. Wind and solar

3. Gas and oil pipeline safety is monitored by: a. FERC b. NERC c. EPA d. OPS DHS

4. The energy sector is regulated by: a. FERC b. NERC c. EPA d. OPS e. DHS

5. The largest transmission pipeline, and the largest pri- vately financed project in the United States at the time, is: a. Colonial Pipeline b. Transco Pipeline c. Kinder Morgan Pipeline d. East Texas Gas Pipeline e. Dominion Pipeline

6. Most pipelines are monitored and controlled by: a. OPS b. SCADA c. E-ISAC d. FERC e. DHS

7. How many PADDs are there in the United States? a. 5 b. 2 c. 3 d. 1 e. 50—one per state

8. In the United States, the top 10 refineries produce: a. Nearly all of the gas and oil b. 20% of all petroleum products15http://www.whitehouse.gov/news/releases/2001/06/energyinit.html

C o p yr

ig h t ©

2 0 1 4 . Jo

h n W

ile y

& S

o n s,

I n co

rp o ra

te d . A

ll ri g h ts

r e se

rv e d .

REFERENCES 235

c. 5% of all petroleum products d. Over half of all gas and oil products e. Two-thirds of all gas and oil products

9. The energy sector began deregulation in: a. Energy Policy Act of 1992 b. PURPA in 1978 c. HlPSA of 1979 d. FERC Order 436 in 1985 e. DHS in 2003

10. The Gulf of Mexico oil supply chain is characterized by: a. Congestion b. High redundancy of transmission c. low link robustness d. low node robustness e. High spectral radius

11. Powder River Basin coal is vulnerable to: a. Earthquakes b. EPA c. Competition d. Power outages e. Railway disruptions

12. The largest NG transmission pipeline for the PADD1 area (Northeastern United States) is: a. Transco b. Columbia c. Dominion d. Tennessee Gas Pipeline e. Colonial Pipeline

13. The largest refined petroleum product pipeline in the United States is: a. Transco b. Columbia c. Dominion d. Tennessee Gas Pipeline e. Colonial Pipeline

14. The top refinery hazards in recent years have been: a. Terrorism b. Flooding c. Corrosion d. SCADA failure e. Equipment failure

15. The top oil pipeline hazards in recent years have been: a. Terrorism b. Flooding c. Corrosion d. SCADA e. Equipment failure

16. The top gas pipeline hazards in recent years have been: a. Corrosion b. Terrorism c. Equipment failure d. Human error e. Hurricanes

17. The top storage hazards in recent years have been: a. SCADA faults b. Equipment failure c. Hurricanes d. lightning/static electricity e. Regulation

18. Why is the Cove Point Intersection at risk? a. Watson is the source or entry point to the pipeline. b. Fault probabilities are too high. c. It is the intersection of three large pipelines. d. It is near the capital. e. Casualties would be high.

19. Why is linden Station at risk? a. Consequences of corrosion would be high. b. It is near New york City. c. It is near New Jersey. d. It has large storage tanks. e. It is near water.

20. The most likely factor shaping the energy sector in the future is: a. The United States will run out of coal. b. Environmental regulation will limit coal. c. Solar and wind will replace natural gas and oil. d. The United States will run out of gas and oil. e. Nuclear power will become number 1.

rEFErEnCEs

[1] Trench, C. J. How Pipelines Make the Oil Market Work—Their Networks, Operation and Regulation. A memorandum prepared for the Association of Oil Pipe lines and the American Petroleum Institute’s Pipeline Committee, December 2001. Available at www.aopl.org. Accessed June 30, 2014.

[2] Chang, J. I. and lin, C.-C. A Study of Storage Tank Accidents, Journal of Loss Prevention in the Process Industries, 19, 2006, pp. 51–59.

[3] Smith, P. K., Bennett, J. M., Darken, R. P., lewis, T. G., and larrañaga, M. D. Network-Based Risk Assessment of the U.S. Crude Pipeline Infrastructure, International Journal of Critical Infrastructures, 10, 2013, pp. 67.

C o p yr

ig h t ©

2 0 1 4 . Jo

h n W

ile y

& S

o n s,

I n co

rp o ra

te d . A

ll ri g h ts

r e se

rv e d .